EPITAXIAL GRAPHENE: SYNTHESIS, CHARACTERIZATION, AND …
Transcript of EPITAXIAL GRAPHENE: SYNTHESIS, CHARACTERIZATION, AND …
EPITAXIAL GRAPHENE: SYNTHESIS,
CHARACTERIZATION, AND DEVICES
RAM SEVAK SINGH
(M.Tech., IIT Kharagpur)
A THESIS SUBMITTED
FOR THE DEGREE OF DOCTOR OF
PHILOSOPHY
DEPARTMENT OF PHYSICS
NATIONAL UNIVERSITY OF SINGAPORE
(2012)
i
DECLARATION
ii
ACKNOWLEDGEMENT
A journey is easier when you travel together. Interdependence is certainly more
valuable than independence. This thesis is the result of four years of work whereby
I have been accompanied and supported by many people. It is a pleasant aspect that
I have now the opportunity to express my gratitude for all of them.
It is with a feeling of profound gratitude and regards that I acknowledge the
invaluable guidance and encouragement rendered to me by my deeply respected
supervisor Professor Andrew Wee Thye Shen of the Department of Physics,
National University of Singapore throughout the course of my Ph.D. work. Without
the meticulous care and attention with which he supervised me and constantly
reviewed the entire work it would not have been possible to carry out the work to
this level. He reviewed my manuscripts, whole thesis, and gave many precious
suggestions to my thesis. He also supported me by granting research assistantship
to me during the writing of thesis.
I would like to thank my co-supervisor, Assistant Professor Chen Wei, for his
positive support and help during my Ph.D. programme. He also reviewed my
manuscripts and thesis and gave me many valuable suggestions to my thesis.
I am also thankful to Professor Ji Wei for supporting me to initiate a work of
graphene-based photoconductive device. I gained fruitful experiences by discussing
with him.
No list would be complete without thanks to the most important people in my
life. My parents, my brother, my sisters can never be thanked enough for the
unconditional love.
iii
The chain of my gratitude would be definitely incomplete if I would forget to
thank the first cause of this chain, The Prime Mover who drives the entire world.
Winding up with a quote from a Robert Frost’s poem “…And miles to go before I
sleep”, I pray to Lord to bestow me with the power to dream and ability to achieve.
iv
LIST OF PUBLICATIONS
International Journals
1. Ram Sevak Singh, Venkatram Nalla, Wei Chen, Andrew Thye Shen Wee,
Wee Ji,“Laser Patterning ofEpitaxial Graphene for Schottky Junction
Photodetectors”, ACS NANO, 5, 5969 (2011).
2. Ram Sevak Singh, Venkatram Nalla, Wei Chen, Wei Ji, Andrew T. S. Wee,
“Photoresponse in epitaxial graphene with asymmetricmetal contacts”,
APPLIED PHYSICS LETTERS, 100, 093116 (2012). Selected for the
March 19, 2012 issue of virtual journal of nanoscale science and
technology.
3. Ram Sevak Singh, Xiao Wang,Wei Chen, Ariando,Andrew. T. S. Wee,
“Large room-temperature quantum linear magnetoresistance in
multilayered epitaxial graphene: Evidence for two-dimensional
magnetotransport.”
APPLIED PHYSICS LETTERS, 101, 183105 (2012).
4. Ram Sevak Singh, Qihua Xiong, Wei Chen, Andrew Thye Shen Wee,
“High-gain photodetectors based on oxygen plasma treated epitaxial
graphene.”
(In preparation)
Conference Presentation
1. Ram Sevak Singh, Venkatram Nalla, Wei Chen, Wei Ji, Andrew T. S. Wee.
(2012) Photoresponse in epitaxial graphene with asymmetricmetal
contacts. The 5th
MRS-S Conference on Advanced Materials. Nanyang
Executive Centre, NTU, Singapore.
2. Ram Sevak Singh) Venkatram Nalla, Wei Chen, Andrew Thye Shen Wee,
Wee Ji. Laser Patterning of Epitaxial Graphene for Schottky Junction
Photodetectors. (2011) Spintronics in Graphene Conference. University
Hall, National University of Singapore, Singapore.
v
Patent
Ram Sevak Singh,Venkatram Nalla, Chen Wei, Wee ATS, Ji Wei, “Laser
Patterning of Graphene/Graphene-oxide Schottky Juction forPhotoconductive
and Photovoltaic Applications” US, 61/515,380, Sep 2011 (Provisional).
vi
TABLE OF CONTENTS
Chapter 1. Literature background......................................................................1
1.1. Fundamentals of graphene..............................................................................1
1.2. Properties of graphene....................................................................................2
1.2.1. Crystal and electronic structures..............................................................2
1.2.2. Electronic transport..................................................................................4
1.2.3. Graphene optical properties.....................................................................6
1.3. Potential applications of graphene.................................................................8
1.3.1. Spin-valve devices...................................................................................9
1.3.2. Sensors...................................................................................................10
1.3.3. Graphene based optolectronic devices...................................................11
1.3.4. Other applications..................................................................................14
1.4. Motivation....................................................................................... ..............15
1.5. Objectives.......................................................................................................16
Chapter 2. Methodology and characterization techniques................................18
2.1. Methodology....................................................................................................18
2.2. Overview of ultra high vacuum (UHV) system and characterization
techniques………………………………………….......................................19
2.2.1. UHV systems........................................................................................19
2.2.2. STM......................................................................................................20
2.3. AFM............................................................................................................... 24
2.4. Raman spectroscopy........................................................................................29
2.4.1. Raman spectroscopy as a characterization tool for graphene
vii
materials................................................................. ................................32
2.5. Hall effect measurement and physical property measurement system
(PPMS)......................................................................................................... 34
Chapter 3. Synthesis and characterization of epitaxial graphene on
Si-face SiC.............. …........................................................................................40
3.1. Introduction..................................................................... ............................40
3.2. Temperature-dependent formation of EG on Si-face SiC substrate:
in-situ EG growth monitoring by STM ..........................................................42
3.3. EG film characterization using AFM and Raman spectroscopy
measurements..................................................................................................46
3.3.1. AFM measurements...............................................................................46
3.3.2. Raman singnatures and EG layers determination..................................47
3.3.3. Evidence of strain in EG........................................................................50
3.4. Hall measurements: electrical properties of EG..............................................51
3.4.1. Materials and devices fabrication..........................................................51
3.4.2. Hall measurements................................................................................52
3.4.3. Results and discussion...........................................................................53
3.5. Conclusion.....................................................................................................55
Chapter 4. Photoconductive devices based on pristine and laser modified
epitaxial graphene.................................................................................................57
4.1. Introduction......................................................................................................57
4.1.1. Laser patterning of epitaxial graphene for Schottky junction
photodetectors………………………………………..........................62
viii
4.1.1.1. Materials and characterization...............................................62
4.1.1.2. Laser patterning and characterization....................................64
4.1.1.3. EG-LEG-EG photodetector .................................................69
4.1.1.3.1. Position-dependent photoresponse observation...70
4.1.1.3.2. Temporal photocurrent profiles............................72
4.1.1.3.3. Spectral photoresponse investigation.................. 74
4.1.1.3.4. Interdigitated EG-LEG-EG device.......................76
4.1.1.4. Infrared absorption spectrum of LEG.................................78
4.1.1.5. EG-LEG-EG characteristics summary, comparison and
advantages..............................................................................79
4.1.2. Photoresponse in EG with asymmetric metal contacts…..................81
4.1.2.1. Materials and device fabrication......................................... 81
4.1.2.2. Photoconductive response of Au-EG-Al device, and influence
of EG active layers..............................................................................84
4.1.2.3. Au-EG-Al characteristics summary, comparison and
advantages............................................................................................89
4.1.2.4. Conclusion............................................................................90
Chapter 5. High-gain photodetectors based on oxygen plasma treated
epitaxial graphene............................................................................................92
5.1. Introduction....................................................................... .........................92
5.2. Materials and device fabrication........................................................ ..........94
5.3. Oxygen plasma treatments of EG device.....................................................95
5.4. Device measurements..................................................... .............................96
5.5. Results and discussion................................................................................97
ix
5.5.1. Photoconductive response ...................................................................97
5.5.2. Photoelectron spectroscopy (PES) investigation..................................99
5.5.3. Temperature-dependent resistance and Arrhenius plot........................101
5.5.4. Raman measurements..........................................................................102
5.5.5. Possible photoresponse mechanism.....................................................103
5.5.6. Light intensity-dependent photoresponse........................................... 105
5.5.7. Spectral photoresponse.......................................................................107
5.6. Applications................................................................................. ...............113
5.7. Conclusion.......................................................................................... .....114
Chapter 6. Magnetotransport studies of multilayer EG on C-face SiC....115
6.1. Introduction............................................................... ............................115
6.2. Material synthesis..................................................................................120
6.3. Device fabrication and characterization...................................... ...........121
6.3.1. Raman characterization......................................................................121
6.3.2. Surface morphology and thickness determination using AFM...........123
6.4. Magnetotransport measurements...................................................................125
6.5. Results and Discussion...................................................................................125
6.5.1. Room temperature large linear magnetoresistance...............................125
6.5.2. Evidence for two dimensional (2D) magnetotransport.........................127
6.5.3. Study of weak localization....................................................................133
6.5.4. Resistance (R)-temperature (T) characteristic......................................136
6.6. Conclusion......................................................................................................137
Chapter 7. Conclusion and future work....................................................139
x
Bibliography...............................................................................................143
xi
SUMMARY
The aim of this thesis is to prepare, characterize and explore epitaxial graphene
(EG) in devices. EG sample preparation on semi insulating Si-face 4H SiC was
carried out by thermal annealing of the SiC substrates in a home-built customized
UHV chamber. Preparation of ultrathin few layer graphene (1-2 layer and 4-6
layer) are directly monitored in situ using scanning tunnelling microscopy (STM).
It is observed that the first-layer of EG forms at a 0.25 nm step up from zero-layer
and second-layer film at 0.35 nm step up from the first-layer. Higher temperature
(>1400 oC) annealing results in the formation of multiple layers films, and the
surface morphology changes dramatically with larger-terrace/step features and
evolution of ridges that traverse over the steps. The as-prepared EG samples are
characterized using atomic force microscopy (AFM), Raman spectroscopy and Hall
measurements. The quality of the EG samples are found to be sufficient for
electrical/optical device applications.
The subsequent chapters report work related to photoconductive devices. A
simple laser irradiation method has been developed to fabricate Schottky junction
photodetectors on (~two-layer) EG. The photodetector shows an efficient, fast
(nanosecond) and wavelength-independent photoresponse in a broad-band
spectrum from ultraviolet through visible to infrared light, distinctively different
from conventional photon detectors. Furthermore, an asymmetric metallization
scheme is utilized with pristine EG to study the photoconductive characteristics of
devices. The effect of the number of EG layers on the photodetector performance
is also investigated. My findings show efficient photoconductivity by utilization of
xii
the asymmetric contacts with EG, making them useful for zero-gap graphene based
photodetectors. I have also found that the number of graphene layers affects the
photoconductivity and response time of the device. The device with four layer
graphene shows higher photoresponsivity (and slightly slower response time)
compared to the device with two layer EG.
In the context of my photoconductive device study, I developed a simple, dry,
effective, and environmentally safe method of mild oxygen plasma treatment in EG
devices to produce high-gain ultraviolet (UV) photodetectors. The photodetector
exhibits performances significantly better than previously reported graphene-based,
or other semiconductor-based photodetectors. Large photoconductive gain,
detectivity, and high selectivity of the photodetector in the UV spectral region,
underscore its potential over Si or several other organic, inorganic, or the hybrid
material based UV photodetectors. Potential applications of the photodetector
includes UV detection, imaging etc.
Finally, magnetotransport studies from room temperature to 2 K in EG prepared
on C-face SiC are carried out. Multilayered EG is prepared on C-face SiC
substrate. The as-prepared EG is investigated using Raman, AFM and
magnetotransport measurements. I have observed a large linear magnetoresistance,
which is sustained at room temperature, and is distinctively different from other
carbon based materials. At low temperatures (≤ 20 K), the appearance of weak
localization effects in the vicinity of zero magnetic field is observed. A two
dimensional (2D) magnetotransport mechanism is verified experimentally, that
gives rise to the linear magnetoresistance (LMR) in EG.
xiii
LIST OF TABLES
Table 3.1. Room temperature Hall measurement results....................................54
Table 4.1. EG-LEG-EG characteristics and its comparison with other carbon based
photodetectors....................................................................... ..............................80
Table 4.2. Summary of Au-EG-Al photodetector device characteristics and its
comparison with other carbon based photodetectors...........................................89
Table 5.1. Comparison of OPTEG photodetectors with previously reported
photodetectors....................................................................................................111
Table 6.1. Room temperature MR/LMR comparison.........................................127
xiv
LIST OF FIGURES
Figure 2.1. (a) The specific epitaxial graphene (EG) preparation lab equipped with
in-situ STM imaging technique. (b) Enlarged view of UHV-chamber dedicated for
EG preparation....................................................... ......................................... 23
Figure 2.2. Schematic of a tapping mode Veeco AFM.................................... 26
Figure 2.3. Configurations of the Veeco D3000 NS49 AFM system.............. 27
Figure 2.4. The Veeco D3000 NS49 AFM system used in my thesis work ... 27
Figure 2.5. Computer monitor with nanoscopeV613r1 software.................... 28
Figure 2.6. Energy level diagram showing different energy states involved in
Raman signals……………………………………………………….................. 29
Figure 2.7. Schematic of a Renishaw Raman spectroscopy set up showing the
basic components of the system..................................................................... 30
Figure 2.8. The Renishaw Raman spectroscopy system. (a) The full view. (b)
enlarged view showing the sample-putting-stage, objective lens and laser power
supply etc. (c) Computer monitor for spectrum recording............................. 31
Figure 2.9. A Hall bar geometry commonly used in thin film or graphene
materials.................................................................. ..........................................35
Figure 2.10. PPMS with controller (model 6000)........................................... 37
Figure 2.11. (a) An EG sample wire-bonded onto the resistance bridge sample
holder for transport measurements. (b) A wire-bonder for Cu wire-bonding....38
Figure 3.1. STM images (Vtip = 0.80 V) showing (a) first-layer EG (layer1, say
EGL1) at a 0.25 nm step up from zero-layer (EGL0), (b) second-layer (EGL2) at a
0.35 nm step up from first-layer. Atomically resolved STM images (Vtip = 0.18 V)
of EGL1 (single layer EG) and EGL2 (bilayer EG) are shown in insets. (c), (d) the
line profiles extracted from images (white dash lin es) in (a) and (b)
respectively............................................. ..........................................................44
Figure 3.2. AFM images of EG prepared on Si-face SiC at different annealing
temperatures. (a) few layer (~ two layer) EG at 1380 0C (annealing time: 5-15
min). (b), (c) Surface morphology at 1400 oC-1450
oC. We observe EG ~4-6 layer
(deduced using Raman measurements) at 1450 oC with ridges (indicated with white
arrows in c) crossing over the steps indicating the continuity of the film over entire
xv
SiC surface. The bottom panels shows the scales in Z-direction obtained across
white dash lines indicated in (a), (b) and (c).......................................................46
Figure 3.3. Raman spectra of EG prepared on high purity semi-insulating (HPSI)
SiC substrates. (a) (i) Single layer EG (ii) Two layer EG and (iii) Four-six layer
EG. (b) Lorentzian fit to 2D-band of single layer EG. (c) Lorentzian fit to 2D-band
(FWHM ~ 62 cm-1
) of two layer EG.................................................................48
Figure 3.4. Raman spectra acquired from several locations on EG sample grown at
temperature ~ 1450 °C. The width (or FWHM; determined using Lorentzian fitting
of 2D-bands) of 2D-bands ranging ~ 73 cm-1
-78 cm-1
that corresponds to EG layer
thickness ~ 4-6 layer........................................................................................ 49
Figure 3.5. (a) Schematic of Hall bar device and connections. (b) Optical picture
of as-fabricated Hall bar on EG sample............................................................53
Figure 3.6. Measured Hall resistance (left panel) and sheet resistance (right panel)
as a function of magnetic field for (a) 2-layer EG on Si-face SiC, (b) 4-6 layer EG
on Si-face SiC...................................................................... .............................54
Figure 4.1. STM images of 1-2 layer EG on semi-insulating Si-face 4H-SiC. (a)
40×40 nm2 STM image (Vtip=0.81 V) showing two regions (single layer and bi-
layer EG) separated by a step of 0.35 nm. Atomically resolved STM images (left
side) of single and bi-layer EG (Vtip = 0.16 V) taken from the highlighted area. (b)
A line profile across the step (indicated as white dash line) extracted from the STM
image in (a).......................................................................................................63
Figure 4.2. (a) 2D-bands intensities on different locations of 2 mm×2 mm area of
a sample. Raman spectra acquired on several locations (> 20 spots, with spot size
~1 m) show an uniformity in the film thickness (~ two layers) over a large area on
the substrate. (b) shows the width (~ 62 cm-1
) of 2D-band in 2-layer EG to confirm
the number of layers in EG sample.................................. ................................63
Figure 4.3. (a) Schematic diagram showing both evaporation of a rectangular part
of the Au film and conversion from EG to LEG. (b) In-situ monitoring of transient
photocurrent as a function of laser pulse number after the evaporation of Au film.
(c) Raman spectra recorded after the evaporation of Au film: (i) HOPG unexposed
and (ii) HOPG exposed to 100 laser pulses; and (iii) EG unexposed and (iv) EG
exposed to 100 laser pulses............................................. ............................... 65
Figure 4.4. Raman spectral evolution during laser patterning on HOPG: (i) HOPG
sample exposed to laser pulse energy 1.5 mJ (100 laser pulses) and (ii) HOPG
sample exposed toLaser pulse energy 4 mJ (100 laser pulses)........................66
Figure 4.5. Detailed analyses on the Raman spectra of LEG. (a) Broadening and
strengthening of G-band and D-band in LEG, as compared to them in EG. (b)
Gaussian fit to the G-band in LEG to reveal two SiC peaks and a G-peak of
graphene. (c) Gaussian fit to the D-band in LEG to reveal a SiC peak (green) and a
D-peak (blue) of graphene. Both Raman analyses and non-linear I-V characteristics
xvi
(Figure 4.8) suggest that LEG exhibits characteristics similar to reduced graphene
oxide..................................................................................................................67
Figure 4.6. Surface morphology using AFM. (a) an AFM image of graphene film
before exposure and (b) an AFM image after exposure. Surface similarity with no
significant change (or peeling of the film) supports that the overall structure of EG
remains intact....................................................................................................68
Figure 4.7. (a) Schematic of EG-LEG-EG photodetector. (b)Transient
photocurrents across the zero-biased single-line EG-LEG-EG junction device over
one second when one of the two junctions is excited by laser pulses of 7-ns
duration and 10-Hz repetition rate at 1064-nm wavelength and 2.4-mJ pulse
energy................................................................... ............................................69
Figure 4.8. I-V characteristics of the single-line EG-LEG-EG Schottky junction
device with white light illumination at 100 mWcm-2
and without light
illumination......................................................................................................70
Figure 4.9. (a) Optical image and Raman spectra measured from the un-converted
EG (black) and LEG (red) region. (b) Schematic of photocurrent line scanning
across the EG-LEG-EG junction. (c) Maximal transient photocurrent and (d)
photocurrent decay time as a function of the distance from the center when
illuminated by laser pulses at 50 J per pulse. The green line in (a) represents the
spatial profile of the laser pulse used for conversion from EG to LEG. The green
spot in (b) represents the area illuminated for the photocurrent measurement.71
Figure 4.10. Temporal photocurrent profiles of the EG-LEG-EG junction when
one junction is excited at (a) 1064-nm and (b) 532-nm laser pulses. The inset in (a)
shows the transient photocurrent profile of the HOPG sample at 1064 nm. The red
line is a single exponential fit. Photocurrent maxima (c) and decay times (d)
excited at 532nm and 1064nm as a function of laser pulse energy..................73
Figure 4.11. Internal spectral responsivity of the EG-LEG-EG junction device
under zero-bias............................................................... .................................75
Figure 4.12. (a) I-V characteristics (filled squares) of the interdigitated device with
(red) and without (black) white light illumination at 100 mW cm-2
. The solid
curves (blue) are the modeling, and schematic of the device-model is shown in (b).
The inset shows the top view of the device with the scale bar=1 mm.The
asymmetry with respect to the bias voltage is primarily due to the dissimilarity of
the two junctions. (c) Schematic of the energy band diagram corresponds to EG-
LEG-EG structure………………………………………………………………76
Figure 4.13. (a) Infrared absorption spectrum of LEG and EG and (b) plot of
(absorption coefficient ×photon energy)2 as a function of photon energy. The plot
in (b) is converted from (a) and the intercept between the straight line (green) and
the x-axis (band gap is measured by extrapolating the straight line to the = 0 axis)
indicates the band gap............................................... .....................................79
xvii
Figure 4.14. (a) Schematic view of zero-bias Au/EG/Al device connected to a
cathode ray oscilloscope (CRO). (b) Optical micrograph of the EG device
contacted with Au (yellow) and Al (white) metals..............................................82
Figure. 4.15. The width (FWHM) of 2D band in Raman spectrum has been used to
analyse the number of layers in EG on Si-face SiC (describe in chapter 3 in detail).
(a) The Raman 2D-band of EG samples having width (FWHM) ~ 60 cm-1
and 73
cm-1
that indicates the thickness ~ two and four layers. The inset shows the
corresponding STM images of two ( Vtip = 0.16 V) and four layers (Vtip = 0.21 V)
EG. (b) 2D-band spectra at different positions (left-middle-right) acquired on 0.05
mm ×2 mm area devices with (i) two layer EG (ii) four layer EG. The measured
2D-band at several points (> 20 points on each device ) showed no significance
variation in intensity or width indicating fair layer thickness uniformity (two and
four layers) in each devices.................................................................................83
Figure 4.16. (a) Photoresponse of theAu/EG/Al device at different positions
excited with laser pulses(photo scan from Au to Al electrode). (b) Plot of
maximum photocurrent (PC) near Au/EG and EG/Al contacts. (c) Photoresponses
(maximum photocurrent) in devices with two and four layers of EG. Laser pulse
duration: 7ns, wavelength: 532nm, and fluence: 0.94 mJ....................................85
Figure 4.17. (a) Schematic of biased Au/EG/Al device. (b) Device I-V
characteristics with and without light illumination at 632.8 nm wavelength. The
inset shows the measured external photoresponsivity (PR) as the function of the
bias voltage. The I-V curves in black and blue represent dark and photocurrent
respectively............................................................... ..........................................87
Figure 4.18. I-Vcharacteristics (EG ~ two-layer) acquired on (a) Al/EG/Al, (b)
Au/EG/Au and (c) Au/EG/Al devices with and without light illumination at 632.8
nm wavelength (Power = 0.96 mW). The top-left inset in (c) shows the external
photoresponsivity in the asymmetric Au/EG/Al device, and the bottom-right inset
shows the zoom-in view around Vbias = 0 V , which clearly indicates the
photoresponse at 0 V..................................................... .....................................88
Figure 5.1 (a) The EG device wire-bonded onto a sample holder (or “chip
carrier”). The two long wires soldered to the two electrodes on EG can easily be
used for device measurement connections. (b) Optical picture of Cr/Au contacts on
EG showing an active area ~ (170 μm ×300 μm).................................................94
Figure 5.2. (a) The specific plasma system (reactive ion etching (NTI RIE-2321)
system) (b) enlarged view of the control monitor................................................95
Figure 5.3. (a) Keithly 2400 current source/measurements unit used for I-V
measurements under dark condition. (b ) Agilent B2912A precision
source/measurements unit used for temporal photoresponse measurements under
illumination with white light...............................................................................96
Figure 5.4. (a) Photograph of the real device under white light. (b) A schematic of
the device showing pristine EG and OPTEG region as an active area. (c) I-V
xviii
characteristics, showing change in channel resistance of the device during
treatments for 20 s (OPTEG1) and 40 s (OPTEG2). The inset shows the optical
picture of Cr/Au contacted two electrode device. (d), (f) Temporal photocurrent Iph
obtained from devices with OPTEG1 and OPTEG2 active channels under white
light (light intensity ~ 30 mW cm-2
, bias voltage = 2 V). (e), (g) Rise (red) and
decay exponential (blue) fittings to the data in (d), (f), estimating response times of
OPTEG1(tr = 7 s, td = 12 s) and OPTEG2 (t r = 18 s, td = 37 s) devices
respectively.................................................... .....................................................97
Figure 5.5. Core-level PES of pristine EG and OPTEG. (a) A wide-scan spectrum
obtained from pristine EG and OPTEG. (b) Si 2p spectra of pristine EG and
OPTEG. (c) C 1s of pristine EG. (d) C 1s of OPTEG showing the content of
chemisorbed C-O and C=O species.....................................................................99
Figure 5.6. UPS of pristine EG and OPTEG)...................................................100
Figure 5.7. (a) Resistance vs. temperature acquired from pristine EG and OPTEG.
(b) Arrhenius plots obtained from data in (a). (c) Data of ≤ 50 K plotted on
appropriate axis for WL-type behaviour..............................................................101
Figure 5.8. Raman characterization in pristine EG and OPTEG devices for two
different treatment times: 20 s (OPTEG1) and 40 s (OPTEG2)........................103
Figure 5.9.(a) Upper panel shows the pristine n-type graphene having free
electrons in conduction band.The bottom panel shows the pristine EG device (b)
OPTEG showing a trap that have captured free electrons, and reduced the electron
density in conduction band. Bottom panel shows the scenario in the device. (c)
Under illumination, electron-hole pairs are generated. Migration of photogenerated
carriers, under the surface depletion electric field, results in a hole trapping or
neutralizing at negatively charged traps, leaving behind high density of unpaired
electrons that moves towards positive electrode and gives rise to a huge
photocurrent. Trapping or neutralization of holes hinder recombination of electron
hole pairs and hence prolong their lifetime……………………………………...104
Figure 5.10. (a) Light intensity F depenndent photocurrent Iph observation in a
OPTEG device. (b) A plot showing photocurrent following a power law
dependency with light intensity. The inset shows the plot in logarithmic
scales...................................................................... ...........................................105
Figure 5.11. Spectral photoresponse measurement system.................................107
Figure 5.12. (a) Photo of experimental set-up during spectral photoresponse
measurement of OPTEG device. (b) Zoom-in view near the device..................108
Figure 5.13. Spectral (300 nm-1000 nm) photoresponse observations in (a), (b)
OPTEG1 device and (c), (d) OPTEG2 device....................................................109
Figure 6.1. Crystal structure of SiC showing C-face (000ī)and Si-face (0001)..115
xix
Figure 6.2. Raman spectroscopy investigations in mulilayer EG on C-face SiC. (a)
A typical Raman spectrum obtained by measurement within the tested device area
(shown by white rectangle in inset). (b) 2D-band fitted with single Lorenzian
function with width (FWHM) ~ 42 cm-1
. (c) A typical single Lorenzian fitted 2D-
band of single layer EG on Si-face SiC, showing identical width ~ 42 cm-1
. (d)
Raman 2D-bands acquired from several location (> 20 points) within the device
area. All the spectra show single Lorenzian nature of 2D bands, indicating
decoupled graphene layers prepared on C-face SiC...........................................122
Figure 6.3. (a) AFM image ofEG film on C-face SiC. (b) Line profile taken across
white dash line showing the scale in Z-direction. (c) A typical AFM image
showing the EG film and SiC (etched part of EG) region. (d) A line profile taken
acrossthe white dash line indicated in (c). The average thickness of the EG film
was found to be ~ 6 nm±1 nm, that corresponds to ~ 20±3 layers in EG..........124
Figure 6.4. V+, V-, I+, I- are four-terminal voltage and current labels used for
measurement of MR in EG.................................................... ............................125
Figure 6.5. (a) MR (in %) at different temperatures T from 300 K to 2 K in
magnetic field B range of -9 T to 9 T. (b) The separated view of (a) from 300 K-50
K, exhibiting LMR sustained at room temperature (300 K) to below 50 K, with no
evidence of negative MR. The separated view of (a) from 20 K to 2 K, showing the
evolution of negative MR ( WL peak around B = 0 T).......................................126
Figure 6.6. Angular dependent magnetotransport in EG. (a) A schematic (b)
Resistance R vs. the tilt angle θ in a fixed magnetic field of 9 T and at T= 300 K
and 2 K. The solid lines (blue, at 300 K and red, at 2 K) are fitting of data with the
function |cos(θ)|................................................................................................128
Figure 6.7. MR as a function of magnetic field B2............................................130
Figure 6.8. (a-g) Linear fits to LMR (in 1 T≤B≤9 T range) data at different
temperatures 300 K, 100 K, 50 K, 20 K, 10 K, 5 K and 2K. These plots give the
slopes (dMR/dB(%)) at different temperatures..................................................131
Figure 6.9. Slope of LMR (dMR/dB) vs. temperature T (logarithmic scales). It is
evident that the curve does not follow a linear relationship, its is just a constant
function, which rules out a power law dependency ((dMR/dB)∝T n
), which is
expected for a classical LMR model...................................................................132
Figure 6.10. WL MR investigations in EG. (a) Magnetic field anisotropy effect on
WL MR, showing a little but non-vanishing negative MR when B is parallel (90°,
In-plane) to sample surface and current direction. (b) Magnetoconductance from 20
K to 2 K in low field regime are found to fit with WL model182
(dots are data while
solid lines are fits to the data using WL model, a scaling along y axis by a factor of
0.10 was adopted in order to fit the data). (e) Estimated Lϕ(phase-coherence length)
vs. T...................................................................................................................133
xx
Figure 6.11. Resistance (R) vs. temperature (T) at zero magnetic field B. The left
inset is the enlarged data in low temperatures (≤20 K) regime. The right inset
showing data of left inset plotted in appropriate axes for WL -type
behaviour............................................................... ............................................136
xxi
LIST OF ABBREVIATIONS
EG Epitaxial Graphene
SiC Silicon Carbide
UHV Ultra High Vacuum
STM Scanning Tunneling Microscopy
AFM Atomic Force Microscopy
HOPG Highly Oriented Pyrolytic Graphite
LEG Laser Modified Epitaxial Graphene
FWHM Full Width at Half Maximum
PR Photoresponsivity
OPTEG Oxygen Plasma Treated Epitaxial Graphene
DOS Density of States
XPS X-ray Photoelectron Spectroscopy
UPS Ultraviolet Photoelectron Spectroscopy
WL Weak Localization
MR Magnetoresistance
LMR Linear Magentoresistance
GO Graphene Oxide
RGO Reduced Graphene Oxide
GNR Graphene Nanoribbon
CNT Carbon Nanotube
IR Infrared
UV Ultraviolet
1
CHAPTER 1
Literature Background
1.1. Fundamentals of graphene
Carbon, one of the most intriguing elements, plays a unique role in nature. It
forms many allotropes such as diamond, graphite and recently discovered
fullerenes and nanotubes.
The elusive two-dimensional form of carbon calledgraphene is probably the
best-studied carbon allotrope theoretically. Because of its extraordinary electronic
properties such as high mobility (over 200000 cm2 V
-1 s-
1),
1, 2 high crystal
quality,3and room temperature anomalous quantum Hall effect (QHE),
4, 5 graphene
has attracted a lot of attention. Charge carriers in graphene can travel thousands of
interatomic distances without scattering.6 This is two-dimensional (2D) honeycomb
lattice which consists in one isolated plane of C atoms. The unit cell contains two
carbon atoms that generate two inequivalent sublattices. Elecrtrons in single layer
graphene, obeying a linear dispersion relation, behave like massless relativistic
particles which results in the observation of a number of peculiar electronic
properties mentioned above. Graphene also provides a bridge between condensed
matter physics and quantum electrodynamics, and opens new perspective for
carbon based electronics.
Graphene was first isolated in 2004 by a group of physicists from Manchester
University, UK, led by Andre Geim and Kostya Novoselov. The method they used
is called micromechanical cleavage5, 7
in which they took 3D graphite and extracted
a single sheet ( a monolayer of atoms) using Scotch tape.
2
Techniques of graphene preparation
So far a number of techniques have been develoved to synthesize the graphene.
Four major techniques are: (i) mechanical exfoliation from bulk graphite crystal,8
(ii) liquid-phase exfoliation (LPE) which consists of chemical wet dispersion
followed by ultrasonication, in both aqueous9 and non-aqueous solvents,
10 (iii)
thermally induced surface segregation from a carbon containing substrate such as
ruthenium11
or carbides (SiC),12
and (iv) chemical vapour deposition of
hydrocarbons on metal substrates.13, 14
Most experimental physics groups are
currently using samples obtained by micromechanical cleavage of bulk graphite,
the technique which provides high quality graphene crystallites up to 100 µm in
size which is sufficient for most research purposes. However, it is very time-
consuming and not suitable for practical applications because of its small size.15
The second method involves epitaxial growth of graphene on substrate and is the
first approach towards large area graphene growth. Due to large defect densities,
the sample quality is generally inferior and needs to be improved.16, 17
The
chemical vapour deposition has become the most used method to grow graphene on
a large scale. However, this technique is currently only viable on metal catalyst
surfaces and requires transferring of graphene onto a desired substrate.
1.2. Properties of graphene
1.2.1. Crystal and electronic structures
Graphene comprises a single or few layers of carbon tightly packed into a
hexagonal honeycomb lattice.6 Depending on the number of carbon layers
graphene has various names: single layer graphene (SLG), bilayer graphene (BLG)
3
and few layer graphene (FLG) and multilayer graphene (MLG).
SLG. The honeycomb lattice of SLG consists of two interpenetrating triangular
sublattices.18
This honeycomb lattice has two inequivalent carbon atoms
(designated as A and B) per unit cell separated by a distance 1.42 Å (lattice
constant). Each atom in the lattice has one s and three p orbitals. The in-plane sp2
bonding is strongly covalent and does not contribute to its conductivity. The
remaining p orbital is oriented perpendicular to the molecular plane and hybridizes
to form π (valence band) and π* (conduction band) bands. The bands form conical
valleys that touch at two of high-symmetry points, conventionally labeled K and K’
called Dirac points, in the Brillouin zone. As such, graphene has zero energy
bandgap, and hence called semimetal. Thus, the E-k relation is linear at low
energies near the zero effective mass for electrons and holes.19, 20
Due to this linear
dispersion relation at low energies, electrons and holes near these six points, which
are inequivalent (K, K’), behave like relativistic particles described by the Dirac
equation for spin 1/2 particles.21
Hence, the electrons and holes are called
Dirac fermions, and the six corners of the Brillouin zone are called the Dirac
points. The equation describing the E-k relation is 22
yxF kkvE ; where vf,
the Fermi velocity, is approximately 106m / s.
21
BLG. BLG is interesting because it shows some properties different from those
observed in single layer graphene. Unlike the linear energy band structure of single
layer graphene, bilayer graphene has a parabolic energy band structure.18
A gap can
open between the conduction band and valence band.22
BLG has been proven to be
less noisy for devices.23
4
FLG. FLG consists of more than two layers (≤ 6 layers) of carbon stacked on one
another by van der Waal weak forces. In graphite, the stacking order is ABAB..
(Bernal stacking).18
The stacking ordering in graphene can be complex. The
subband structure of a trilayer with Bernal stacking includes two touching
parabolic bands, and one with Dirac dispersion, combining the features of bilayer
and monolayer graphene. A gap can open in a stack with Bernal ordering and four
layers if the electronic charge at the two surface layers is different from that the
two inner ones.18
No gap is observed in graphene with higher number of layers,
even in the presence of charge inhomogeneity. Experimentally, it has been
observed that graphene becomes metallic with increasing number of layers due to a
gradual increase of charge carrier concentration at zero energy and they appear as
electron-like or hole-like carriers. An inhomogeneous charge distribution between
layers lead to 2D electrons and holes that occupy a few graphene layers near the
surface, and affect the transport properties of graphene stacks.
The graphene layers can be rotated with respect to each other due the weak
coupling between planes that allows for the presence of many different
orientational states that are quasi degenerate in energy.18
For example, a rotation
by a small angle in bilayer graphene leads to the decoupling between layers and
recovery of linear dispersion curve of the single layer with modified Fermi
velocity.
1.2.2. Electronic transport
Experimental results from transport measurements show that graphene has a
remarkably high electron mobility.1, 2
Additionally, the symmetry of the
5
experimentally measured conductance indicates that the mobilities for holes and
electrons should be nearly the same.20
The mobility is nearly independent of
temperature between 10 K and 100 K,7 which implies that the dominant scattering
mechanism is defect scattering. Scattering by the acoustic phonons of graphene
places intrinsic limits on the room temperature mobility2 to 200,000 cm
2 V
−1s
−1 at
a carrier density of 1012
cm−2
. However, for graphene on silicon dioxide substrates,
scattering of electrons by optical phonons of the substrate is a larger effect at room
temperature than scattering by graphene’s own phonons, and limits the mobility
to 40,000 cm2
V−1
s−1
.2
Despite the zero carrier density near the Dirac points, graphene exhibits a
minimum conductivity on the order of 4e2 / h.
24 The origin of this minimum
conductivity is still unclear. Rippling of the graphene sheet or ionized impurities in
the SiO2 substrate may lead to local puddles of carriers that allow conduction.20
Several theories suggest that the minimum conductivity should be4e2 / hπ;
however, most measurements are of order 4e2 / h or greater and depend on impurity
concentration.24
Recent experiments have probed the influence of chemical dopants on the
carrier mobility in graphene.24
Schedin et al.25
doped graphene with various
gaseous species (some acceptors, some donors), and found the initial undoped state
of a graphene structure can be recovered by gently heating graphene in vacuum.
Schedin et al. reported that even for chemical dopant concentrations in excess of
1012
cm−2
there is no observable change in the carrier mobility. Chen et al.24
doped
graphene with potassium in ultra high vacuum at low temperature. They found that
potassium ions act as expected for charged impurities in graphene, and can reduce
6
the mobility 20-fold. The mobility reduction is reversible on heating the graphene -
to remove the potassium.
Due to its 2D property, charge fractionalization (the fractioning of electrons
into anyons) is thought to occur in graphene. It is though that it may therefore be a
suitable material for the construction of quantum computers using anyonic
circuits.26
On the other hand, multilayer epitaxial graphene (EG) prepared on C-face SiC
has additional advantages over single layer graphene on SiO2. Multilayer EG
allows screening of substrate induced scattering ( underneath layer closer to the
substrate can protect the subsequent top layer from substrate scattering) and hence
spin-transport efficiency is enhanced.27
Furthermore, graphene layers in EG grown
on C-face SiC are electrically decoupled and behave like a single sheet of
graphene. This property of EG on C-face SiC is attributed to the presence of
rotational faults (graphene layers rotated by 30o or ± 2
o with respect to SiC
substrate, typically described as R30/R±2
fault pairs).28
Hence, these features of
multilayer films on C-face provide the motivation for the study of
magnetotransport in EG and its other electrical device properties. Details of EG on
C-face SiC can be found in chapter 6.
1.2.3. Graphene optical properties
Graphene has unexpectedly high opacity for an atomic monolayer, with a
startlingly simple value: each layer absorbs πα ≈ 2.3% of white light, where α is
the fine-structure constant.29, 30
This means SLG has 97.7% transparency.
Graphene only reflects <0.1% incident light in visible region.31
FLG absorbs light
7
additively with each layer having absorbance of 2.3%. Graphene shows an almost
flat absorption spectrum from 300 to 2,500 nm with a peak in the ultraviolet (270
nm) region. In the case of FLG, other absorption features can be seen at lower
energies, associated with interband transitions.32, 33
It is notable that observed long phase coherence lengths in graphene make it a
promising candidate for low power consumption and fast electronic devices.1
Carrier multiplication34
was also found to be a dominant phenomenon in graphene
which could be utilized to enhance the conversion efficiency of solar cells.
Luminescence is an important optical property having relevance in devices such
as light emitting diodes (LEDs). Zero-gap graphene lacks photoluminescence
property. Therefore, bandgap opening or defect incorporation to graphene would be
of much interest for these devices and several potential optoelectronic devices
where interband transitions play an important role. One route to opening the band
gap is cutting the graphene into ribbons or quantum dots. Another route involves
chemical or physical treatments, to reduce the connectivity of the π-electron
network. Graphene ribbon shows the photoluminescence indicating the existence of
bandgap.35
Reports based on the chemical treatments can also be found. For
example, oxygen plasma treated single layer graphene shows a bright
luminescence.36
Luminescent graphene based materials can now be routinely produced that
cover the infrared, visible and blue spectral ranges.36, 37
Photoluminescence in
graphene oxide, due to oxygen related defect states, is another example.36
Luminescent quantum dots are widely used for bio-labelling and bio-imaging.
However, their toxicity and potential environmental hazard limit widespread use
8
and applications. Fluorescent bio-compatible carbon-based nanomaterials may be a
more suitable alternative. Fluorescent species in the infrared and near-infrared are
useful for biological applications, because cells and tissues show little auto-
fluorescence in this region.38
Photoluminescent graphene oxide has been
demonstrated for live cell imaging in the near-infrared.39
It has been demonstrated
that the bandgap of graphene can be tuned from 0 to 0.25 eV (about 5 μm-
wavelength) by applying a voltage to a dual-gate bilayer graphene field-effect
transistor (FET) at room temperature.22
My review here indicates that fuctionalization, modification, surface treatment
or external electrostatic forces (For example, gate voltage) would be of much inte-
rest to developing graphene-based optoelectronics.
1.3. Potential applications of graphene
The unique electronic properties of graphene make it a possible candidate to
replace commercial silicon-based electronics. Even in its infancy, mobilities in
graphene are already an order of magnitude higher than that of modern silicon
transistors. The mobility remains high even at highest electric-field-induced
concentrations and is seemingly unaffected by chemical doping.25
Moreover, this
discovery comes at a perfect time as silicon-based technology is fast approaching
its fundamental limits. SLG has a fundamental problem which is that the
conducting channel in single-layer graphene cannot be pinched off due to minimum
conductivity. This severely hampers achievable on-off ratios for FETs (Field-Effect
Transistors). However, BLG shows much better promise in this area.40
A single p-n
junction (which acts as a barrier) in a BLG device would lead to a small tunneling
9
current. This can be used in the development of FETs. Furthermore, a bilayer FET
with a local gate inverting the sign of the charge carriers should yield a much larger
on-off ratio than single-layer graphene. It has also been observed that the electronic
gap of a graphene layer can be manipulated by applying a gate bias.41
This means
that a biased graphene bilayer is a tunable semiconductor where the electronic gap
can be manipulated by the electric field effect. These results have strong
implications for the development of carbon-based electronics.
Furthermore, designing graphene-based materials and devices often requires the
control of graphene properties. In particular, modification of graphene is desirable
so that its properties can be tuned specifically for an application. Strategies to
modify the properties of graphene include chemical doping.42
Details are described
in section 1.3.3. In order to meet the requirement of high sheet conductivity,
surface doping of graphene can be accomplished using organic molecules42, 43
or
high work function metal oxides.44
Zero-gap graphene is not suitable for
applications in switching devices or in optoelectronics. Adsorption of gases such as
hydrogen, oxygen on graphene surface has resulted in opening of a bandgap45
or
alteration of its optical properties.36
Laser induced heating was used to reduce
graphene oxide into graphene.46
1.3.1. Spin-valve devices
Spin-valve devices represent another route for possible graphene applications.
The magnetoresistance effect is the change in resistance in the presence of an
external magnetic field. It was found that soft magnetic electrodes can inject
polarized spins into graphene and a 10% change in resistance was observed during
10
the switch from parallel to anti-parallel state.47
With the presence of spin injection,
there is potential for research on spin current manipulation. Recently, highly
efficient (~ 75%) spin transport was observed in EG on C-face SiC.27
The
improved spin transport in EG is attributed due to two main factors: (i) multilayer
EG screens the substrate-induced scattering, leading to reduced spin relaxation
through the Elliot-Yafet mechanism,48
(ii) much flatter EG49
on SiC, compared to
the graphene on SiO2/Si.50
Furthermore, it was shown by a group in Delft
University51
that superconductivity can be induced in graphene using the proximity
effect and the magnitude of this supercurrent can be manipulated using an external
gate voltage. This could be used to create superconducting FETs and other devices.
1.3.2. Sensors
The group from Manchester University25
showed that graphene is capable of
absorbing gas molecules from its surrounding atmosphere. The absorbed molecules
change the local carrier concentration in graphene one electron at a time resulting
in step-like changes in the resistance. Hence, one can create a gas sensor by simply
monitoring the changes in the resistivity of graphene. Previous detectors have
failed in their quest to reach the ultimate goal of achieving a sensitivity that can
detect a single atom or molecule. Their failure can be attributed to the thermal
motion of charges and defects which lead to the creation of noise. The big
advantage with graphene is that it is an extremely low-noise material,23
electronically speaking. This makes it an ideal candidate for not only gas detectors
but also other probes which measure charge, magnetic fields and mechanical strain.
Graphene finds potential applications in biological fields as well. Extensive reports
11
on graphene based biosensors, such as DNA sensor,52-54
single cell detector,55
protein sensor54, 56
etc, can also be found.
1.3.3. Graphene based optolectronic devices
Graphene as a transparent conductor
Optolectronic devices (such as solar cells, light emitting diodes, displays, touch
screens etc) require a conductor (for example: anode or cathode in solar cell) with
high transparency and high conductivity or low sheet resistance Rs. ITO is the
dominant material used in these devices. ITO is a n-type semiconductor with
optical transparency ~ 80% and low sheet resistance of 10 Ω/□.
SLG having higher transparency (97.7% (graphene) > 80% (ITO)) could be an
alternative and better option as transparent conductor in optoelectronic devices.
However, graphene lacks low sheet resistance. An ideal SLG with 97.7 %
transparency has sheet resistance (Rs) ~ 6 k Ω/□.15
As such, graphene is superior to
ITO only in terms of transparency, not sheet resistance. Therefore, active work in
this area of research is still ongoing. Different strategies: spraying,57
dip and spin
coating,58
vacuum filtration and roll-to-roll processing59
have been explored to
prepare graphene based transparent conducting films.
A more effective strategy is chemical doping of graphene. The best result59
shows transparency ~ 90% with Rs ~ 30 Ω/□ using nitric acid treatment of CVD
graphene. This area of research is under progress with exploitation in devices such
as organic solar cells, light emitting diodes, touch screens etc.
Graphene/semiconducting material-hybrid system
Various reports can be found where graphene is mixed with semiconducting
12
materials (such as Si,43, 60
nanowires,61-65
quantum dots,66, 67
, TiO268
) in order to
obtain improved performance of the photoconductive or photovoltaic devices.
Miao et al.43
demonstrated chemically doped graphene/n-Si Schottky junction solar
cells with efficiency of ~ 8.6 %. The improved efficiency is attributed to reduction
of sheet resistance after doping, thereby increasing the built-in potential that
efficiently separates the electron-hole pairs generated by absorbed photons. Lee et
al.65
reported graphene-CdS NW hybrid structures for high-speed and large
photoconductivity. In the case of pristine graphene, the lifetime of its
photogenerated charge carriers is quite short (~ picoseconds)69, 70
due to the
ultrafast charge recombination process.71
Thus, superior mobility or conductivity
of graphene2, 72
combined with other semiconducting material can harvest the
photogenerated carriers more efficiently from semiconducting material to the metal
electrodes. As such, the key point is the utilization of graphene as an efficient
charge-carrier transport channel incorporated with other semiconducting materials.
Gerasimos et al.67
demonstrated a hybrid graphene-quantum dot based
phototransistor with a very high gain of ~ 108 electrons per photon. Photogenerated
carriers in the quantum dot are transferred to the graphene, where they recirculate
many times due to the high charge mobility of graphene and long trapped-charge
lifetimes in the quantum dot layer.
Photodetectors based on graphene materials
Photodetectors are devices that convert light energy into electrical current.
These photodetectors work on the general principle of the creation of electron-hole
pairs under the action of light. When a semiconductor material is illuminated by
photons of energy greater than or equal to its well-defined bandgap, the absorbed
13
photons promote electrons from the valence band into excited states in the
conduction band, where they behave as free electrons able to travel long distances
under the influence of an intrinsic or externally applied electric field. As such, the
spectral bandwidth is limited by materials absorption. Si or other III-V
semiconductors have limitations at long wavelengths,73
and hence light having
energy less than their band gap would not be detectable.73
Previous results show that graphene can absorb from ultraviolet to terahertz
range,30, 74, 75
hence it can be used as a photodetector working with a much broader
wavelength range. The response time, ruled by high carrier mobility of graphene,73
would be much faster, leading to ultrafast photodetectors. The highest
photoresponsivity ~ 6.1 mA W-1
was achieved by Xia and Mueller et al.76
in
pristine graphene prepared by mechanical exfoliation method, utilizing an
asymmetric metallization scheme. However, the efficiency of the device cannot
beat the current high performance photodetectors.77, 78
Furthermore, work on large
area processable graphene is needed to allow integration of devices over wafer
scales. As such, photodetector research on large area graphene79, 80
would be of
much interest. The reported low efficiency (6.1 mA W-1
) could be due to the
absorption by only a single layer of graphene, and the use of FLG or MLG with
high mobility graphene should further improve the performance. To commercialize
graphene based devices, high-gain photodetectors should be the next step for
applications such as UV detectors. This could be possible via two main routes:
modifying the graphene (for example opening of band gap, selective doping to
form p-n junctions, surface treatment etc) or forming a hybrid with highly optically
14
active material, where graphene can add value as a fast transport channel to the
photogenerated carriers.
1.3.4. Other applications
In the realm of composite materials, it has been shown that a graphene powder
of uncoagulated micrometer-size crystallites can be manufactured in a way suitable
for mass production.10, 81-83
One drawback is that graphene-based composites are
unlikely to be as mechanically strong as single-wall carbon nanotubes (SWNT)
owing to the latter’s stronger entanglement. However, SWNT has a disadvantage
in that it has a lower surface-to-volume ratio than graphene sheets due to the
inaccessibility of the inner nanotube surfaces. This makes graphene sheets more
favourable for altering properties17, 31
such as mechanical, rheological and
permeability. Coupled with much lower production costs, graphene-based
composite materials have a promising future.83
Graphene is also being potentially used as electrodes in Li ion batteries.84, 85
The large surface-to-volume ratio and high conductivity of graphene powder could
increase the efficiency of batteries.86-88
Graphene can serve as a binder material,
eliminating the use of binding polymer materials such as poly(vinylidene
fluoride).84, 88
Carbon nanotubes have also been considered for this purpose but
graphene powder has a commercial advantage as it is much cheaper to produce
as it is more readily prepared from a vast array of graphites.
Hydrogen storage is already a possible application for carbon nanotubes. It has
been suggested that graphene may have the ability to absorb large amounts of
15
hydrogen.89-92
Tozzini et al.92
proposes that controlled buckling of graphene layers
in a device can potentially reach a gravimetric capacity of ~ 8 %. Ao et al.89
claims,
based on the density functional theory, high-capacity hydrogen storage in Al-
adsorbed graphene. A graphene layer with Al adsorbed on both sides can store
hydrogen up to 13.79 wt% with average adsorption energy -0.193 eV/H2. Recently,
Kim et al.93
has demonstrated hydrogen storage with maximum storage capacity ~
4.8 wt% at 77 K and and at 9.0 MPa pressure in thermally modulated multilayered
graphene oxide (GO). However, issues related to release kinetics and loading need
to be further studied before this application can be be realized.
1.4. Motivation
The study of two-dimensional graphene is motivated by its unique properties
that make it an interesting and potentially useful material. Graphene has a wide
range of potential applications, as described above, which include spin valve
devices, chemical sensors, novel electronic and optoelectronic devices. A study of
graphene syntheses and characterization is important for its applications to become
viable. In particular, for practical device applications, large area high quality
graphene would be of significant relevance. Graphene can be used as an ultrafast
photodetector device in optical communication and holds great promise for future
photonic devices.156
Graphene-based materials and devices often require the
precise control of graphene properties. As such, modification of graphene is
desirable so that its properties can be tuned specifically for an application.
Different strategies to modify the properties of graphene have been described in the
literature review above. Nevertheless, multilayer EG on C-face SiC has many
16
advantages over single layer graphene on SiO2 as described in section 1.2.2. With
these motivations, synthesis, characterization, modification and the respective
applications of EG in devices are conducted in this thesis.
1.5. Objectives
In the present thesis, graphene on SiC substrates are synthesized and
characterized in detail using in-situ scanning tunnelling microscopy (STM), atomic
force microscopy (AFM), Raman spectroscopy, and magnetotransport
measurements. More importantly, photoconductive devices based on pristine or-
modified graphene have beenexplored in detail. These include the following:
1) Chapter 3. This chapter involves preparation of ultrathin few layers (~ two
layer or four-six layer) epitaxial graphene (EG) on Si-face 4H SiC
substrates and characterizations of the EG using in-situ STM and ex-situ
AFM, Raman and Hall measurements.
2) Chapter 4 (section 4.1.1). Here, a simple and large area processable method
oflaser irradiation is applied to fabricate EG-laser modified epitaxial
graphene (LEG)-EG Schottky junction photodetectors on Si-face 4H SiC.
The EG-LEG-EG photoconductive characteristics, advantages are
investigated.
3) Chapter 4 (section 4.12). This section involves asymmetric metallization
scheme utilization with pristine EG to study its photoconductive
characteristics and the effect of number of graphene layers on the device
performance.
17
4) Chapter 5. This chapter focusses on a study of mild oxygen plasma
treatment in EG devices to produce high-gain photodetectors.
Photoconductive characteristics, mechanism, and device-performance are
reported and discussed.
5) Chapter 6. Magnetotransport studies of multilayered EG prepared on C-face
4H SiC substrate are carried out. Significant two dimensional transport
induced linear MR (LMR) and weak localization observed experimentally
in EG are analysed and discussed.
18
CHAPTER 2
Methodology and Characterization Techniques
2.1. Methodology
EG samples were synthesized via thermal annealing of 4H SiC substrates in
ultra high vacuum (UHV). The preparation of FLG on Si-face SiC was directly
monitored in-situ using scanning tunneling microscopy (STM). Surface
morphology, roughness, thickness and electrical property studies were further
conducted using AFM, Raman, and Hall measurements. The details of synthesis
and characterization of EG on Si-face SiC is presented in chapter 3. These
characterization techniques including synchrotron-based photoemission
spectroscopy (PES) were used in whole thesis. Specifically, the laser modification
of graphene was conducted using nanosecond laser (1064 nm wavelength, 7-ns
duration, and 10 Hz repetition rate laser pulses were generated by a Nd: YAG laser
(Spectra Physics Quanta Ray)) irradiation method. Oxygen plasma treatments in
EG device were performed using a specific reactive ion etching (NTI RIE-2321)
system. The details of the electrical or photoconductive measurements are
described in the respective chapters itself. The subsequent sections quickly provide
an overview of different characterization techniques used in this thesis work.
19
2.2. Overview of UHV system and characterization techniques
2.2.1. UHV systems
The vacuum regime characterized by pressures lower than 10-9
mbar is called
UHV. To achieve an ultra high vacuum, we require the use of bakeable materials
in construction since we need to heat the entire system to 180°C for several hours
("baking") to remove water and other adsorbed trace gases on the surfaces of the
chamber. There is no single vacuum pump that can operate all the way from
atmospheric pressure to ultra-high vacuum. Instead, a series of different pumps is
used, according to the appropriate pressure range for each pump. Pumps commonly
used to achieve UHV include: turbomolecular pumps (especially compound and/or
magnetic bearing types), ion pumps titanium sublimation pumps etc. UHV
pressures are measured with a hot or cold cathode type ion gauge. Ultra-high
vacuum is necessary for many surface analytic techniques such as: STM, X-ray
photoelectron spectroscopy (XPS), Auger electron spectroscopy (AES), thin
film growth and preparation techniques with stringent requirements for purity, such
as molecular beam epitaxy (MBE), UHV chemical vapor deposition (CVD) and
UHV pulsed laser deposition (PLD), angle resolved photoemission
spectroscopy (ARPES) etc.
For STM experiments, UHV reduces oxidation and contamination, hence
enables imaging and the achievement of atomic resolution on clean metal and
semiconductor surfaces, e.g. imaging the surface reconstruction of the
unoxidized silicon surface.
20
2.2.2. STM
STM is a scanning probe microscope (SPM) instrument capable of resolving
surface detail down to the atomic level. The invention of STM microscopy by G.
Binnig, H. Rohrer at 1981 set a new milestone in the world of nanotechnology.94
With this technique, sophisticated manipulation, interaction and measurement
down to single atoms are achieved.95
. The STM is based on the concept of quantum
tunneling. When a conducting tip is brought very near to the surface to be
examined, a bias (voltage difference) applied between the two can allow electrons
to tunnel through the vacuum between them. The resulting tunneling current is a
function of tip position, applied voltage, and the local density of states (LDOS) of
the sample. Information is acquired by monitoring the current as the tip scans
across the surface, and is usually displayed in image form. As an example, if the
distance is decreased by 0.1 nm, the tunneling current will be increased by
approximately one order of magnitude. By scanning the sample or tip in a raster
pattern, an image can be generated representing the surface in real space. STM can
be a challenging technique, as it requires extremely clean and stable surfaces, sharp
tips, excellent vibration control, and sophisticated electronics.
The scanning of this STM is realized by the movement of tip. An XYZ piezo
tube is used in charge of precise movement. Peripheral electronics are connected to
the metal tip and conductive sample to apply bias, monitor tunneling current and
feedback to control the movement of the tip. The sample and tip are fixed to a stage
which is isolated from external vibrations by mechanical springs. An eddy current-
damping system is installed at the side of stage to achieve quick stabilization.
21
Two modes are used in STM scanning: constant current mode and constant
height mode. In the constant current mode, the distance of the tip from the surface
is kept constant during scanning. To achieve this, the tip is moved by feedback
control to maintain the constant tip-sample separation. In the constant height mode,
the tip height is fixed, while the recorded tunneling current changes exponentially
with changes of sample morphology. To avoid crashing of tip, a surface with less
than 1nm corrugations is required in this mode. It should be noted that STM does
not directly probe the real surface topography but maps the LDOS. Due to the
polarity of the bias, the images could either reflect the LDOS in filled state or in
empty state. If the LDOS equals to zero at certain energies, the STM could fail to
probe the topmost atoms on the surface if bias is set into such energy gaps. Thus,
STM measurements need to be performed at suitable biases to avoid undesired
images. Sometimes, when surfaces are composed by several atomic layers with
different energy gaps, the proper selection of bias in measurement could give
topographic images of underlying layers. For example, EG on Si-face SiC exhibits
such a property. At high bias (typically sample bias at -2V), the EG layer is
transparent, and images of the underlying interfacial layer are obtained. The
topography of graphene networks becomes visible at low bias (typically less than
0.5V).96
The specific STM system (Omicron) used in this work is shown in Figure 2.1.
The system has two main chambers (1) sample preparation chamber and (2) STM
analysis chamber. These two chambers are separated by a valve, so that in-situ
transferring of prepared sample can be done for STM imaging. The basic steps
involve (i) loading of sample through load lock to preparation chamber (ii)
22
transferring of as-prepared EG by transfer arm to STM analysis chamber (iii)
putting the sample from manipulator onto STM stage using wobble stick (see the
Figure 2.1). The CCD camera is mounted to see the tip movements (controlled by a
remote controller) during its approach to the sample up to the tunnelling distance.
Once these steps are achieved, we start scanning with cotrolling the tunnelling
current and tip bias using measurement control unit in SPM software. For
successful STM experiments clean and UHV compatible samples and sample
plates are essential. Sharpness of the tip is also a key factor to obtain an atomically
resolved STM images.
Our preparation chamber comprises of home-built silicon source with heating feed-
through as shown in Figure 2.1. As such, further cleaning of substrate under silicon
flux can be done in UHV chamber itself. X, Y, Z, and rotation moves components
are presented in Figure 2.1.
23
Figure 2.1. (a) The specific EG preparation lab equipped with in-situ STM
imaging technique. (b) Enlarged view of UHV-chamber dedicated for EG
preparation.
STM chamber EG preparation
chamber Valve
Silicon source
heating feed through
X-move
Y-move
Z-move
Rotation
Manipulator
CCD
Wobble stick
(b)
(a)
24
2.3. AFM
AFM is another type of scanning probe microscopy (SPM), with demonstrated
resolution on the order of fractions of a nanometer, more than 1000 times better
than the optical diffraction limit. The AFM is one of the foremost tools for
imaging, measuring, and manipulating matter at the nanoscale. In contrast to STM,
AFM can be used for conducting as well as non-conducting samples and large area
(several microns) surface features can be probed. The information is gathered by
feeling the surface with a mechanical probe. Piezoelectric elements that facilitate
tiny but accurate and precise movements on (electronic) command enable the very
precise scanning. In some variations, electric potentials can also be scanned using
conducting cantilevers.97
Basic principle of operation depends on the physical forces (such as mechanic
contact force, van der Waals forces, capillary forces, chemical bonding,
electrostatic force, magnetic forces etc depending on the situation of
measurements). The AFM consists of a cantilever with a sharp tip (probe) at its end
that is used to scan the specimen surface. The cantilever is typically silicon or
silicon nitride with a tip radius of curvature on the order of nanometers. When the
tip is brought into proximity of a sample surface, forces between the tip and the
sample lead to a deflection of the cantilever according to Hooke's law. Typically,
the deflection is measured using a laser spot reflected from the top surface of the
cantilever into an array of photodiodes. Other methods that are used include optical
interferometry, capacitive sensing or piezoresistive AFM cantilevers. These
cantilevers are fabricated with piezoresistive elements that act as a strain gauge.
Using a Wheatstone bridge, strain in the AFM cantilever due to deflection can be
25
measured, but this method is not as sensitive as laser deflection or interferometry.
If the tip was scanned at a constant height, a risk would exist that the tip collides
with the surface, causing damage. Hence, in most cases a feedback mechanism is
employed to adjust the tip-to-sample distance to maintain a constant force between
the tip and the sample. Traditionally, the sample is mounted on a piezoelectric tube,
that can move the sample in the Z direction for maintaining a constant force, and
the X and Y directions for scanning the sample. Alternatively a tripod configuration
of three piezo crystals may be employed, with each responsible for scanning in the
X, Y and Z directions (see Figure 2.2). This eliminates some of the distortion
effects seen with a tube scanner. In newer designs, the tip is mounted on a vertical
piezo scanner while the sample is being scanned in X and Y using another piezo
block. The resulting map of the area Z = f(X,Y) represents the topography of the
sample.
A number of modes are used to operate AFM, depending on the application. In
general, possible imaging modes are divided into static (also called contact) modes
and a variety of dynamic (non-contact or "tapping") modes where the cantilever is
vibrated. Schematic of a tapping mode Veeco AFM system, which was used in this
thesis works, is shown in Figure 2.2. It uses the silicon cantilever and tip and
different knobs to align laser onto the cantilver and detector mirror are shown in
Figures 2.3 and 2.4. The cantilever holder is removable and can be removed or
replaced again to change the tip. In tapping mode, the tip of the cantilever does not
contact the sample surface. The cantilever is instead oscillated at a frequency
slightly above its resonant frequency where the amplitude of oscillation is typically
a few nanometers (<10 nm). The van der Waals forces, which are strongest from
26
1 nm to 10 nm above the surface, or any other long range force which extends-
Figure 2.2. Schematic of a tapping mode Veeco AFM.
above the surface acts to decrease the resonance frequency of the cantilever. This
decrease in resonant frequency combined with the feedback loop system maintains
a constant oscillation amplitude or frequency by adjusting the average tip-to-
sample distance. Measuring the tip-to-sample distance at each (X,Y) data point
allows the scanning software to construct a topographic image of the sample
surface.98
Tapping mode AFM does not suffer from tip or sample degradation
effects that are sometimes observed after taking numerous scans with contact
AFM. This makes tapping mode AFM preferable to contact AFM for measuring-
27
Figure 2.3. Configurations of the Veeco D3000 NS49 AFM system.
Figure 2.4. (a) The Veeco D3000 NS49 AFM system used in my thesis work.
Detector Knob adjust Laser Sample
28
Figure 2.5. Computer monitor with nanoscope V613r1 software.
soft samples. In the case of rigid samples, contact and non-contact images may
look the same. However, if a few monolayers of adsorbed fluid are lying on the
surface of a rigid sample, the images may look quite different. An AFM operating
in contact mode will penetrate the liquid layer to image the underlying surface,
whereas in non-contact mode an AFM will oscillate above the adsorbed fluid layer
to image both the liquid and surface. Schemes for dynamic mode operation
include frequency modulation and the more common amplitude modulation. In
frequency modulation, changes in the oscillation frequency provide information
about tip-sample interactions. Frequency can be measured with very high
sensitivity and thus the frequency modulation mode allows for the use of very stiff
cantilevers. In amplitude modulation, changes in the oscillation amplitude or phase
provide the feedback signal for imaging. In amplitude modulation, changes in
the phase of oscillation can be used to discriminate between different types of
materials on the surface. Amplitude modulation can be operated either in the non-
contact (or tapping) or in the intermittent contact regime. In dynamic contact mode
29
, the cantilever is oscillated such that the separation distance between the cantilever
tip and the sample surface is modulated.
2.4. Raman spectroscopy
Basic principle and instrumentation
Raman spectroscopy is a spectroscopic technique that makes use of inelastic
scattering of monochromatic laser light with matter. The energy of the laser
photons shifts up or down due to interactions with molecular
vibrations, phonons or other excitations in the system.
Figure 2.6. Energy level diagram showing different energy states involved in
Raman signals.
The higher and lower energy shifts are called Stokes Raman scattering and anti-
Stokes Raman scattering respectively (depicted in Figure 2.6). If the emitted
30
photon has an energy equivalent to that of the excitation energy, this is known as
Rayleigh scattering. The Raman energy shift gives information about vibrational,
rotational and other low frequency transitions modes in molecules and hence the
technique can be used to study solid, liquid and gaseous samples.99
Schematic of a Renishaw Raman spectroscopy system is shown in Figure 2.7.
A Raman spectrometer for materials characterization normally uses a microscope
Figure 2.7. Schematic of a Renishaw Raman spectroscopy set up showing the
basic components of the system.
to focus the laser beam to a small spot (<1-100 µm diameter). Light from the
sample passes back through the microscope optics100
. This set of optics is called
spectrometer or spectrograph and comprises of components such as grating, fabry-
perot etalon, beam splitter etc as shown in Figure 2.7. Raman shifted radiation-
Microscope
Manual
mechanical stage Spectrometer unit
31
Figure 2.8. The Renishaw Raman spectroscopy system. (a) The full view. (b)
enlarged view showing the sample-putting-stage, objective lens and laser power
supply etc. (c) Computer monitor for spectrum recording.
Laser
power
supply
Objective
lens
Movable
sample
stage
Remote
control
(a)
(b)
(c)
32
(analysed by spectrometer) is detected with a charge-coupled device (CCD) and
finally spectra and image data from CCD is read, stored and processed by the
computer (Figure 2.8), which also controls all spectrometer functions.
2.4.1. Raman spectroscopy as a characterization tool for graphene materials
Introduction
Raman spectroscopy has become an important tool to characterise carbon based
materials101-109
such as graphite, carbon nanotubes and graphene. This section deals
the basic and reviews how the technique has been used to identify graphene
formation and for determination of the number of layers in mechanically
exfoliated graphene or graphene on SiC substrates.
The most prominent features in the Raman spectra of graphene are the so-called
G-band (appearing in range ~1582 cm-1
- 1600 cm-1
), 2D-band (in range ~ 2680 cm-
1-2740 cm
-1) and D-band (~ 1360 cm
-1).
103, 106 The G-band is associated with the
doubly degenerate (in-plane transverse optic: iTO and longitudinal optic: LO)
phonon mode (E2g symmetry) at the Brillouin zone center.108, 110
In fact, the G-band
is the only band coming from a normal first order Raman scattering process in
graphene. On the other hand, the 2D and D-bands originate from a second-order
process (2D-band frequency is around half the D-band frequency), involving two
iTO phonons near the K point for the2D-band or one iTO phonon and one defect in
the case of the D-band. The D-band is called disordered-induced band and it
appears when the sample has defects/disorders or at the edges of graphene.
Both the D and 2D bands exhibit a dispersive behavior since their frequencies in
the Raman spectra change as a function of the energy of the incident laser. The D-
33
band frequency upshifts linearly with increasing energy of the incident laser over a
wide laser energy range. The origin and the dispersive behaviour in the frequency
of the D and 2D bands originate from a double resonance (DR) Raman process.110,
111 In this DR process, the wave-vectors q of the phonons associated with the D and
2D bands (measured from the K point) would couple preferentially to the electronic
states with wave-vectors k (measured from the K point), such that q ≈ 2k. The
double-resonance (DR) process begins with an electron of wave-vector k around K
absorbing a photon of energy.112
The electron is inelastically scattered by a phonon
or a defect of wavevector q and energy Ephonon to a point belonging to a circle
around the K’ point, with wave-vector k+q, where the K’ point is related to K by
time reversal symmetry. The electron is then scattered back to a k state, and emits a
photon by recombining with a hole at a k state. In the case of the D band, the two
scattering processes consist of one elastic scattering event by defects of the crystal
and one inelastic scattering event by emitting or absorbing a phonon.113
Determination of number of layers in graphene sample using Raman spectroscopy
Two approaches have been adopted to determine the number of layers in
graphene materials using the Raman technique. The first approach involves the
determination of intensity-ratio (IG/I2D) of G- band and 2D-band.102
The proper
Gaussian fitting of G-band while Lorentzian fitting of 2D-band are carried out to
obtain the intensities of the two bands. Graf et al.102
has studied and extracted the
relation between number of layers and IG/I2D ratio. However, this method is not
suitable to be applied to EG directly due the SiC signals that appears with the G-
band in Raman spectrum.114, 115
34
The second approach101, 103, 116
involves the full-width-at-half-maximum
(FWHM) of the 2D-band which correlates with the number of layers, mobility and
stacking order in graphene materials. The Raman 2D-band of a single layer
graphene is, generally, found to fit with single Lorentzian while a bilayer graphene
is found to fit with four Lorentzian functions. Joshua et al.116
found a correlation
between 2D-band and mobility of the graphene, where sharper width of a 2D–band
indicates high mobility in EG. Lee et al.,103
by combining the observations with
various surface science techniques, standardized 2D-band having a relation with
number of layers in EG on Si-face SiC. As such, 2D-band investigation is the most
suitable for layer thickness or quality evaluation in EG.
2.5. Hall effect measurement and physical property measurement system
(PPMS)
Introduction
Hall measurements are carried out to characterise the electrical properties (such
as carrier type and its density, mobility and conductivity) of metallic or
semiconducting materials. The measurement is based on the principle of the Hall
effect in which a uniform current flows through the material in the presence of a
perpendicular applied magnetic field. The Lorentz force deflects moving charge
carriers to one side of the sample and generates an electric field perpendicular to
both the current density and the applied magnetic field. The Hall coefficient is the
ratio of the perpendicular electric field to the product of current density and
magnetic field, while the resistivity is the ratio of the parallel electric field to the
current density. Experimental determination of a real material's transport properties
35
requires some significant departures from the ideal model. To begin with, one
cannot directly measure the electric field or current density inside a sample.
Current density is determined from the total excitation current and the sample
geometry. Electric fields are determined by measuring voltage differences between
electrical contacts on the sample surface.
Hall effect measurements commonly use two sample geometries: (1) long,
narrow Hall bar geometries and (2) nearly square or circular van der Pauw
geometries. Here, I would only like to discuss the Hall bar geometry as this
geometry was only used in my works. A commonly used 6-contacts Hall bar
geometry for thin films or graphene is shown in Figure 2.9.
Figure 2.9. A Hall bar geometry commonly used in thin film or graphene
materials.
The Hall measurements and calculations for carrier density and mobility in EG
samples are described in the next chapter 3, section 3.4.3. The Hall bar channel
length L and width W are indicated in Figure 2.9. Measurements of longitudinal
(parallel) voltage Vxx and transverse voltage Vxy (or Hall voltage) at a fixed
magnitude of current flow, as shown in Figure 2.9, give the electrical property
36
informations such as carrier density, mobility, and sheet conductivity of the
material. The details for calculations of these parameters for graphene samples
have been presented in the next chapter 3.
PPMS
The PPMS is a standard instrument for heat capacity, magnetometry, electro-
transport and thermal transport measurements applications. The PPMS system used
in this thesis works is shown in Figure 2.10. It is configured to perform
magnetotransport measurements under transverse or longitudinal magnetic field up
to 9 T and temperatures from 400 K to 2 K. It consists of two main units: sample
chamber in which sample is inserted, and controller unit which controls all the
critical components of the instruments to provide direct communication with
application electronics for rapid data acquisition. The PPMS adds a configurable
resistance bridge board to the model 6000 PPMS controller (see Figure 2.11a). A
resistance bridge sample holder board with mounted epitaxial graphene sample is
shown in Figure 2.11a. It can measure two samples together at a time mounted for
four wire resistance measurements. The system is configured for both AC and DC
resistance options for four wire resistance, five wire Hall measurements, AC
resistivity, I-V curves critical currents etc. Furthermore, the PPMS system is
configured with horizontal and vertical rotator options. The horizontal rotator
option rotates the sample about the horizontal axis while the vertical rotation option
rotates the sample about vertical-axis. The basic steps for a transport measurement
involve: sample mounting onto as ample holder and wire bonding testing the
electrical contacts by inserting it into a testing unit, finally inserting the sample-
37
Figure 2.10. PPMS with controller (model 6000).
PPMS sample
chamber
PPMS controller
(model 6000)
38
Figure 2.11. (a) An EG sample wire-bonded onto the resistance bridge sample
holder for transport measurements. (b) A wire-bonder for Cu wire-bonding.
into the sample chamber at bottom, where a 12 pin- connector pre-wired to the
system electronics is available. The details of PPMS basics117, 118
can be found in
reference 48. The PPMS measures the resistance of the sample by passing tiny
currents through the sample from one lead to the next. Therefore, care must be
Cu wire-bonding (a)
(b)
39
taken to ensure that the different connecting leads are isolated (not in contact with
neighbouring contact).
40
CHAPTER 3
Synthesis and Characterization of Epitaxial Graphene on
Si-face SiC
3.1. Introduction
Graphene, a two-dimensional hexagonal network of carbon atoms, is a recently
discovered and emerging material for electrical and photonic device applications
due to its exotic physical properties such as room-temperature quantum Hall effect,
massless Dirac fermions, ballistic transport and optical behavior.79, 119
The size of
graphene flakes formed by exfoliation method is relatively small (less than 100
µm)15
and graphene preparation on SiC substrates, so called epitaxial graphene
(EG)119, 120
is the first approach in the context of large area production of the film.
The recent reports12, 120
on the large-area production of EG by annealing of silicon
carbide (SiC) stimulated a great deal of interest as it is compatible with industrial
large-scale processing techniques. Transport measurements on EG prepared on
semi-insulating SiC substrates showed high carrier mobilities12
(up to 27000 cm2
V-1
s-1
) and long phase coherence lengths, thereby making EG a promising
candidate for low-power-consumption and fast electronic devices. EG based
transistor devices in top gated configuration16, 17
have also been demonstrated. Very
few groups have succeeded to prepare high quality EG on SiC relevant to
electrical/optical devices. As such, a study of graphene syntheses and
characterization become important steps in order for its applications to become
viable. For example, for practical device applications, large area high quality
graphene would be of significant relevance. High quality homogeneous films of
41
graphene can be used in electrical as well as photonic devices115, 120
that show
superior and distinctively different performances compared to conventional devices
used in the semiconductor industry. As such preparation of graphene films on SiC
substrates and understanding of the film quality using standard method such as
STM, AFM, Raman spectroscopy, Hall measurements should be helpful for
commercial applications.
For device measurements and applications, the substrate underneath the film
should be highly insulating to avoid any substrate-conduction effect on device
performance. Therefore, synthesis of high quality graphene film on highly resistive
(semi-insulating SiC) SiC is an important challenge for researchers and engineers.
Some research groups/ technologists16, 121
have demonstrated the preparation of EG
on Si-and C-face of high-purity semi-insulating SiC substrates but the quality of
the graphene films are much inferior to graphene prepared by the mechanical
exfoliation method. Hence, more research on large scale high quality EG materials
synthesis is still required. The method of EG synthesis involves the controlled
annealing of the SiC substrates using direct current heating/resistive heating UHV
environments. The vacuum annealing of SiC for EG preparation requires an
addition pre-surface-treatment of SiC substrate using H2-etching at very high
temperature ~ 1600 °C. The process requires high temperature with H2 which is
dangerous as H2 gas is potentially explosive at high temperature. EG can directly
be prepared in a UHV environment without this rigorous substrate surface pre-
treatment process. The growth mechanism involves the sublimation of Si from SiC
(0001) (i.e. the so called Si-face of SiC) at high temperatures leaving the carbon
layers on top of the SiC surface.79, 122, 123
Therefore, the preparation conditions:
42
temperature, time of growth, pressure and rate of heating are crucial to produce
uniform and atomically thin film graphene.
This chapter presents details of EG preparation, in-situ monitoring of the EG
fims growth using STM and further ex-situ characterization using AFM, Raman
spectroscopy and Hall measurements. I succeeded to prepare ultrathin monolayer,
bilayer and few layer (~ 4-6) EG on Si-face surface of SiC. The morphological
changes, thickness and electrical properties of EG films prepared at different
annealing temperatures (described in subsequent sections) are studied. The
graphene materials prepared on semi-insulating SiC substrates are used to fabricate
photoconductive devices as described in chapters 4 and 5.114, 115
This shows that
EG materials synthesized in this study have potential not only in electrical but also
in photonic device applications.
3.2. Temperature-dependent formation of EG on Si-face SiC substrate: in-situ
EG growth monitoring by STM
Method
The following steps were carried out to prepare EG.
(i) SiC substrates (high purity semi-insulating 4H-SiC, resistivity≥105Ωcm) were
well cleaned using isopropanol (the substrates were put in isopropanol solution in a
beaker and then sonicated for 20 minutes), acetone (the same process for cleaning
with acetone solution was done for 20 minutes) followed by HF (10%) solution
(substrates were put for 1-2 minutes in the HF solution) before introducing them
into the preparation chamber.
43
(ii) The cleaned substrate was introduced through load lock to UHV (pressure ~
4.5x10-10
mbar) preparation chamber. Prior to the growth in UHV chamber, the
surface was further cleaned under Si-flux (for 5 min) at substrate temperature ~900
°C in the same UHV chamber to remove oxide contamination. The mechanism for
oxide removal is the formation of SiOx oxides on SiC surface and their desorption
at subsequent high temperature annealing.
(iii) Subsequently (10 minutes holding time at 900 oC after Si–flux is stopped), the
substrate was annealed at temperatures 1100 oC-1450
oC (heating rate 25
oC/min)
and surface graphitization was monitored in-situ using STM. The slow and steady
annealing process was adopted to maintain the ultra high vacuum in the chamber.
However, the pressures (slightly increased due to high temperature heating) were ~
9.5 × 10-10
mbar at growth temperature ~ 1380 °C and ~ 1.6 × 10-9
at growth
temperature ~ 1450 °C. An optical pyrometer device was used to measure the
temperatures of the sample surface. The annealing time at growth temperature is
also the key parameter which can affect the homogeneity and thickness of the EG
film. Therefore, optimised growth conditions for the formation of few layer (~ 2
layers and 4-6 layers) EG films were accomplished as described below.
44
Figure 3.1. STM images (Vtip = 0.80 V) showing (a) first-layer EG (epitaxial
graphene layer1, say EGL1) at a 0.25 nm step up from zero-layer (EGL0), (b)
second-layer (EGL2) at a 0.35 nm step up from first-layer. Atomically resolved
STM images (Vtip = 0.18 V) of EGL1 (single layer EG) and EGL2 (bilayer EG) are
shown in insets.(c), (d) the line profiles extracted from images (white dash lines) in
(a) and (b) respectively.
At lower temperature of around 1380 oC (annealing time: 5 min), the surface is
observed to be graphitised (70% of substrate surface) with one or two layers of
graphene. Formation of the first-layer of EG (say EGL1) was monitored by STM
(Figure 3.1). EGL1 is found to be at 0.25 nm step up from zero-layer (EGL0) while
second-layer EG (say EGL2) at 0.35 nm up from the first-layer (EGL1) as shown
EGL0
L1
EGL1
L1
EGL2
L1
EGL1
L1
0 1 2 3 4 5 6
0.00
0.05
0.10
0.15
0.20
0.25
Z (
nm
)
X (nm)
0 1 2 3 4 5 6-0.05
0.00
0.05
0.10
0.15
0.20
0.25
0.30
0.35
0.40Z
(nm
)
X (nm)
(c)
(a) (b)
(d)
45
by line profiles (Figures 3.1c and 3.1d). Annealing the substrate repeatedly (time:
5-15 minutes), the surface is observed to be fully graphitized (film thickness ~ two
layers) as confirmed by in-situ STM measurements. Furthermore, a temperature
increase from 1380 oC to 1400
oC results in a significant change in surface
morphology (with enhanced roughness) exhibiting larger-terrace/step features. The
further annealing of the sample with temperature (~ 1450 oC) results in contraction
of steps (step bunching) with reduced roughness. The evolution of ridges (white
lines) of height ~ 1-2 nm running across the steps is also observed at this
temperature. The surface shows more than two layers film at this temperature. It
has been observed that the graphene grows from the step edges9 and hence
formation of steps, its density and regularity have potential importance in terms of
homogeneity, crystallinity and thickness of the materials. The results suggest that
the larger area (on entire SiC surface) continuous with bigger domain sizes EG film
can be prepared at higher annealing temperatures. But higher temperature will
increase the Si desorption rate and hence the thickness of the film. This Si
desorption can be reduced if we could manage to increase the pressure around the
growth surface and that could result in slower desorption rate of Si and hence
graphene up to single layer would be more favourable. For large area (over entire
surface of SiC) single layer EG, careful and precise control of the parameters could
be a challenging task. The prepared EG film (~ 2 layers and 4-6 layer) in this work
are further correlated with Raman measurements in section 3.3.2.
The morphological changes during the annealing processes are described in
section 3.3.1.
46
3.3. EG film characterization using AFM and Raman spectroscopy
measurements
3.3.1. AFM measurements
The change in surface morphology of EG film prepared at different
temperatures was further investigated using ex-situ AFM. All the measurements
were performed in tapping (or non-contact) mode. Figure 3.2 shows the annealing
temperature dependent AFM images of EG films prepared on Si-face 4H-SiC
substrates.
Figure 3.2. AFM images of EG prepared on Si-face SiC at different annealing
temperatures. (a) Few layer (~ two layer) EG at 1380 0C (annealing time: 5-15
min). (b), (c) Surface morphology at 1400 oC-1450
oC. We observe EG ~4-6 layer
(deduced using Raman measurements) at 1450 oC with ridges (indicated with white
arrows in c) crossing over the steps indicating the continuity of the film over entire
SiC surface. The bottom panels shows the scales in Z-direction obtained across
white dash lines indicated in (a), (b) and (c) respectively.
(a) (b) (c)
0 200 400 600 800 1000-8
-6
-4
-2
0
2
4
6
8
Z (
nm
)
X (nm)0.0 0.2 0.4 0.6 0.8 1.0
-8
-6
-4
-2
0
2
4
6
8
Z (
nm
)
X (nm)0 1 2 3 4 5
-8
-6
-4
-2
0
2
4
6
8
Z (
nm
)
X (nm)
47
We observe the significant change in morphology of the film prepared at different
temperatures. The films grown at higher temperature (~1450 oC) show the larger
terraces with evolution of ridges (white line running across the steps as shown in
Figure 3.2c). These ridges are caused due to the thermal mismatch between film-
and substrate and are more pronounced at higher temperatures. Such ridges
(traversing over the steps, Figure 3.2c) are also an indication of continuous film
over large area on the substrate and can easily be monitored by AFM for
subsequent device processing and measurements.
3.3.2. Raman signatures and EG layers determination
The Raman spectra signatures acquired from EG samples are shown in Figure
3.3.The G-band in the Raman spectrum of graphene is related to phonon vibrations
in sp2 bonded carbon materials, while the D-band is the signature for disorder
induced by phonon.17, 106
The measurements were performed on EG using 514 nm
laser excitation wavelength, 10 mW power and wave number scan range: 1000 cm-
1 to 3000 cm
-1. All the measurements were performed at room temperature and
ambient environment. With increase in thickness of EG we observe the
strengthening of the G-band (at ~ 1593 cm-1
) with attenuation of the SiC peak (at ~
1520 cm-1
) as shown in Figure 3.3a. The 2D-band (at ~ 2710) of single layer EG
(Figure 3.3c) was fitted with a single Lorentzian (FWHM ~ 42 cm-1
) while for two
layer EG (Figure 3.3b) the 2D-band (at ~ 2732 cm-1
) could be fitted with four
Lorentzian (FWHM ~ 62 cm-1
). Relatively sharper 2D-band widths for
single/bilayer EG are a clear indication of high quality EG materials prepared in
our study. The width (FWHM ~ 73-78 cm-1
) of 2D-band for EG prepared at ~ 1450
48
0C was found to be relatively broader (the detail analyses are shown in Figure 3.4)
and we also observe the 2D-band slightly red-shifted. The observed weak D-band
(at ~ 1360 cm-1
) is related to defect/disorder in the material.
Figure. 3.3. Raman spectra of EG prepared on high purity semi-insulating (HPSI)
SiC substrates. (a) (i) Single layer EG (ii) Two layer EG and (iii) Four-six layer
EG. (b) Lorentzian fit to 2D-band of single layer EG. (c) Lorentzian fit to 2D-band
(FWHM ~ 62 cm-1
) of two layer EG.
The number of layers in graphene has been determined using the intensity IG/I2D
ratio in Raman spectra as described in previous chapter 2. However, this method
(a)
(b) (c)
1200 1500 1800 2100 2400 2700 3000
(iii)
(ii)
Ra
ma
n i
nte
sit
y
Wavenumber (cm-1)
Single layer EG
Two layer EG
Four-six layer EG
(i)
SiC
G
2D
D
2600 2650 2700 2750 2800 2850 2900
Ra
ma
n in
ten
sit
y
Wavenumber (cm-1)
Lorentzian fit to
2D-band
of two layer EG
FWHM ~ 62 cm-1
49
cannot be applied directly with EG as the graphene peaks overlap with the SiC-
related substrate signals (at ~ 1520 cm-1
and ~ 1690 cm-1
). Therefore, the more
convenient and accurate method is to probe the 2D bands and the correlated
change in its widths can provide the information about the number of layers in EG
samples. The single and bilayer graphene (directly measured with STM) have been
correlated with the 2D-bands as shown in Figures 3.3. The single layer graphene
2D-band is fitted with single Lorentzian (Figure 3.3b) with narrower width of ~ 42
cm-1
. On the other hand, bilayer EG 2D-band, fitted with ~ 4-Lorentzian (Figure-
Figure 3.4. Raman spectra acquired from several locations on EG sample grown at
temperature ~ 1450 °C. The width (or FWHM determined using Lorentzian fitting
of 2D-bands) of 2D-bands ranging ~ 73 cm-1
-78 cm-1
that corresponds to EG layer
thickness ~ 4-6 layer.
50
3.3c), with the width of ~ 62 cm-1
. As such this 2D-band FWHM trend as a
function of number of graphene layers can be standardized103
by the following
equation:
FWHM (2D) = (-45(1/N)) + 85) cm-1
[3.1]
where N is the number of EG layers.
Figure 3.4 shows the Raman 2D-bands acquired from several different locations on
EG sample prepared at temperature ~ 1450 °C. The widths of 2D-bands ranges
from ~ 73 cm-1
-78 cm-1
and this, as estimated with equation 3.1, corresponds to a
thickness ~ 4-6 layers in EG on Si-face surface of SiC substrate.
3.3.3. Evidence of strain in EG
G-band (or 2D) of single layer EG on Si-face SiC is found to be blue-shifted
with respect to the unstrained single layer exfoliated graphene on SiO2.
Furthermore, blue-shift can occur due to doping effect or strain in EG, but the 2D
band is almost unshifted for the electron doping range: 0-2.5 X103 cm
-2.124
As such,
we can attribute the simultaneous shift of G and 2D band to compressive strain
rather than doping.
Compressive strain is given as
Δω/ω0 = γ (strain) [3.2]
where γ is the Gruniesen parameter
For unstrained single layer graphene125
on SiO2
ω0(G) = 1585 cm-1
, γ (G) = 1.8
ω0(2D) = 2685 cm-1
, γ (2D) = 2.7
51
For single layer EG on Si-face SiC (Figure 3.3a)
ω(G) = 1593 cm-1
, ω(2D) = 2710 cm-1
From equation [3.2], we obtain the compressive strain of ~ 0.3% (using G-band
shift or 2D-band shift) in single layer EG on Si-face SiC. This strain can arise in
EG during the cooling down procedure due to the large difference in thermal
expansion coefficient between SiC and graphene (SiC contracts and graphene
expands).125
Furthermore, since the shift in G (or 2D band) is sensitive to strain, the
doping effect cannot be accurately determined. For example, the 2D and G bands
of our 4-6 layer EG are red shifted and this could be explained as relaxation of
strain rather than doping.
3.4. Hall measurements: electrical properties of EG
The electrical properties of EG prepared in my study were investigated by Hall
measurements in standard Hall bar devices. Only one-step lithography was used to
reduce the contamination from resists due to repeated processes of lithography. The
details of processes for device fabrication and measurements are presented in the
subsequent sections.
3.4.1. Materials and devices fabrication
The as-synthesized EG samples (~ two-layer and four-six layer) were used to
fabricate several devices in Hall bar geometry.
To fabricate devices, the following processes were carried out:
(1) Au (100 nm)/Cr (10 nm) metals (six-contacts) were thermally evaporated on all
pristine EG samples through shadow mask.
52
(2) The metals contacted EG was spin coated with PMMA resist (speed: 4000 rpm,
time: 80 sec) followed by baking at 180 oC for 4 minutes.
(3) Subsequently, electron beam lithography was used to pattern the EG film in
Hall bars (electron beam writing, then developing in MIBK+IPA(1:1) for 1 minute,
cleaning with IPA for 1 minute and then drying it with N2 gas blow). After
developing, the sample was put in oxygen plasma (flow rate: 20 sccm, power: 50
watt, chamber pressure: 125 mtorr) for 2 minutes to etch EG film out to give the
film a Hall bar geometry. Finally, the sample was put in acetone for two days to
remove the PMMA resist.
(4) The final sample with Hall bars devices was further annealed in vacuum for 40
minutes to improve the metal graphene contacts before the measurements.
3.4.2. Hall measurements
The EG samples Hall measurements were performed using PPMS. The details
about PPMS118
are given in section 2.4 of the previous chapter 2. At first, the EG
sample was placed and wire-bonded in a standard sample holder used in a PPMS
system.
For Hall measurements, a constant AC current of 1 µA was applied through the
sample and transverse voltage, i.e., Hall voltage (Vxy) and longitudinal voltage
(Vxx) were measured (Figure 3.5a) during a magnetic field sweep (0- 1 T) at 300 K.
The unit used for magnetic field in the experiment is Oersted, where 10000 Oe = 1
Tesla. The schematic of a EG Hall bar device and measurementconnections (with
optical picture of as-fabricated Hall bar on EG) is shown in Figure 3.5.
53
Figure 3.5. (a) Schematic of Hall bar device and connections. (b) Optical picture
of as-fabricated Hall bar on EG sample.
3.4.3. Results and discussion
The Hall measurements data (transverse resistance Rxy vs. magnetic field B and
longitudinal resistance Rxx vs. magnetic field B of a Hall bar device) can be used to
determine the electrical properties of EG such as carrier type and its density, sheet
resistivity, and carrier mobility. The sheet resistance Rs (or sheet resistivity in case
of graphene) for the Hall bar (370 µm x 150 µm) device in our study was extracted
as Rs= Rxx (W/L), where W is the Hall bar channel width ( 150 µm), L (370 μm) is
the distance between voltage leads 1 and 2 (Figure 3.5a). Figure 3.6 shows the plots
for the measured Hall resistance Rxy (left panel) and sheet resistance (right panel)
with varying magnetic field B (in tesla) of two-layer and four-six layer EG
samples. The 2D carrier density (ns) is derived by the equation: ns = -1/(eRH),
where RH is the Hall coefficient and can be obtained by taking the slope of the Rxy
(b) (a)
54
Figure 3.6. Measured Hall resistance (left panel) and sheet resistance (right panel)
as a function of magnetic field for (a) 2-layer EG on Si-face SiC, (b) 4-6 layer EG
on Si-face SiC.
Table 3.1. Room temperature Hall measurement results.
Sample Annealing
temperature (oC)
Maximum mobility
(cm2V
-1 s
-1)
Carrier density
(cm-2
) EG (2-layer) on Si-
face SiC
1380 1430 -4.6 1× 012 cm-2
EG (4-6 layer) on
Si-face SiC
1450 800 -3.1 × 1013 cm-2
(a)
(b)
0.0 0.2 0.4 0.6 0.8 1.0
-30
-25
-20
-15
-10
-5
0
Magnetic field B (T)
Hall r
esis
tan
ce R
xy (
)
255
270
285
300
315
330
345
Sh
eet
resis
tan
ce R
s (
)
0.0 0.2 0.4 0.6 0.8 1.0
-180
-160
-140
-120
-100
-80
-60
-40
-20
0
Hall r
esis
tan
ce R
xy (
)
Magnetic field B (T)
Sh
eet
resis
tan
ce R
s (
)
900
1000
1100
1200
1300
1400
1500
55
vs. B curve shown in Figure 3.6 and e is the electron charge. Furthermore, the Hall
mobility is calculated by µ = 1/(nseRs). The measurements were performed on
several 2-layer and 4-6 layer EG devices at room temperature and results
correspond to maximum mobility are tabulated in Table 3.1.
The EG (~ 2-layer) sample shows the maximum Hall mobility ~1430 cm2 V
-1s
-1
and corresponding carrier density~ -4.6 × 1012
cm-2
. While 4-6 layer EG shows
higher carrier concentration ~ -3.1 × 1013
cm-2
and corresponding mobility ~ 800
cm2 V
-1 s
-1. In general, the finding shows that the mobility increases with decrease
in carrier density and EG on SiC substrates have electrons as dominant carriers that
supports the previous reports.12, 16, 79
The reports say that because of work function
difference (0.3 eV) between SiC and graphene, the bottom interface graphene layer
is charged with electrons, and hence showing the n-type conductivity in few layer
graphene on Si-face SiC.
3.5. Conclusion
EG (~ two-layer and four-six layer) films are prepared on high purity semi-
insulating Si-face 4H-SiC substrates. The preparation of the graphene is directly
observed using STM. The observations show the first-layer of EG at a 0.25 nm step
up from zero-layer and second-layer film at 0.35 nm step up from the first-layer.
Higher temperature (>1400 oC) annealing results in formation of multiple layer
films and the surface morphology is changed dramatically with larger-terrace/step
features and evolution of ridges that traverses over the steps. The EG (~2-layer)
exhibits room temperature Hall mobility up to 1430 cm2V
-1s
-1, measured from large
(150 µm×370 µm) Hall bars, while the mobility is up to 800 cm2V
-1s
-1 for
56
relatively thicker (~ 4-6 layer ) EG. This approach to produce and understand
epitaxial graphene at different temperatures on semi-insulating SiC substrates has
potential relevance to electrical and photonic device applications.
57
CHAPTER 4
Photoconductive Devices Based on Pristine and Laser
Modified Epitaxial Graphene
4.1. Introduction
Photodetectors or photosensors are devices that are responsive to light energy. It
operates by absorption of photons and generating a flow of current in an external
circuit, proportional to the incident power. Variety of semiconductor materials such
as Si,77
Ge,78, 126
ZnO,127
GaN,128
GaAs,129
ZnS130
etc has been used as active
materials in devices as photodetectors that are utilized in diverse applications such
as spectroscopy, photography, analytical instrumentation, optical position sensors,
beam alignment, laser range finders, optical communications, environmental
science and medical imaging instruments etc. These photodetectors work on the
general principle of the creation of electron-hole pairs under the action of light.
When a semiconductor material is illuminated by photons of energy greater than or
equal to its well-defined bandgap, the absorbed photons promote electrons from the
valence band into excited states in the conduction band, where they behave like
free electrons able to travel under the influence of an intrinsic or externally applied
electric field.
The photodetecctors can be categorized into two types of devices: first one is the
“photoconductors”, which consists of a photo-active material (usually
semiconducting material) simply with two ohmic contacts, where the electric field
58
leading to the collection of charge carriers is provided by applying a bias voltage
between the contacts at either end; second type is the “photovoltaic
photodetectors”, which use the internal electric field of a p-n or Schottky (metal-
semiconductor) junction to achieve the charge separation.
Brief overview of graphene based photodetectors
Graphene materials are the subject of intense research for condensed matter
physicists, electronic engineers or chemists due to its exceptional electronic
properties.6, 131, 132
Though graphene is a zero-bandgap semiconductor, it has
impressive optical properties.30, 32
It can absorb photons from visible-infrared-
ultraviolet, and the strength of the interband transition in graphene is one of the
largest of all known materials. Multiple graphene layers absorb additively.32, 133
Observed long phase coherence lengths in graphene make it a promising candidate
for low power consumption and fast electronic devices1. Carrier multiplication
34, 134
was also found to be a dominant phenomenon in graphene which could be utilized
to enhance the conversion efficiency of solar cells. As such, graphene holds a great
promise for novel photonic devices.
Novel graphene based photonic devices researches have recently been a great
attraction to researchers. Few research groups76, 135-137
working on graphene
(produced by mechanical exfoliation of graphite) or solution processed graphene
oxide (so called reduced graphene oxide (RGO)) have demonstrated the
photodetectors devices previously.
Xia and Mueller et al.76, 135
report the mechanically exfoliated graphene based
photodetectors in which built-in electric field near metal electrode-graphene
interfaces is responsible to produce an ultrafast photocurrent in device. As such,
59
photo generation and carrier transport mechanism in graphene differ fundamentally
from those in photodetectors made from conventional semiconductors.
Furthermore, graphene-metal contacts76, 138
have strong effect for the formation of
p-n junctions depending on the difference in work function between the two
materials. Technically, asymmetric metallization scheme has been utilized with
carbon nanotube for efficient photonic devices.139
This scheme with graphene
enhances the overall photocurrent by breaking the mirror symmetry of the built-in
potential profile within the channel and hence improves the efficiency of the
devices.76
However, small size graphene flakes (up to 100 µm) used in study of above
described photodetectors limits the device compatibility with large area CMOS
process flow and hence its utilization for practical device applications in
semiconductor industry. Furthermore, maximum photoresponsivity (or efficiency)
6.1 mA W-1
, as reported by Mueller et al.,76
could be inferior to several existing
semiconductor materials based photodetectors.77, 78
As such, high efficiency
photodetectors based on graphene materials are still needed to be explored.
However, many technical challenges remain. One challenge is the large-area
patterning of EG via lithography which is compatible with current silicon-based
fabrication technology. The other challenge is the development of a simple, low-
cost technique for the making of p-n junctions or Schottky junctions, the key
component in electronic and optoelectronic devices. Hence, graphene based
photodetector device research and development in these directions are required for
high efficiency photonic devices compatible with practical semiconductor
processing.
60
Solution processed graphene or so called RGO based devices, showing the
photoconductive characteristics, have also been demonstrated. Basant et al.136, 137
has reported the infrared (IR) and ultraviolet (UV) photodetectors based on few
layers graphene (RGO) films with the detector responsivity 4 mA W-1
and 0.12 A
W-1
respectively. The reported response time (in the range of minutes) and
efficiency of these devices would be inferior to that in exfoliated graphene (as
described above) and to many other semiconducting photodetectors and hence
development of graphene based high performance photonic devices are still
ongoing.
This chapter is categorised into two sections: section 4.1.1. Laser patterning of
epitaxial graphene for Schottky junction photodetectors, and section 4.1.2.
Photoresponse in epitaxial graphene with asymmetric metal contacts.
In section 4.1.1, a simple laser irradiation method is developed to fabricate
Schottky junction photodetectors within few-layer EG. Graphene can be oxidized.
For example, by the addition of oxidizing agents, it was converted to GO via a
photo-chemical method,140
or simply heating graphene in ambient turned graphene
into GO.141
Conversely, laser direct writing/exposing methods were employed by
several research groups to reduce GO to graphene.46, 142, 143
Wu and his co-
workers144
demonstrated that a Schottky junction was created if both EG and e-
beam-modified GO were brought together in contact with each other. In this case,
a semimetal-semiconductor Schottky junction was formed as EG was semi-
metallic46, 142, 143
and e-beam-modified GO showed semiconducting behavior.144
But, their fabrication technique involves time-consuming processes (such as
chemical oxidation of film and then repeated process of lithography), with the
61
limited controllability and lack of scalability. The laser patterning method
presented here in section 4.1.1 enables one to modify EG locally to have
semiconductor properties, thereby achieving a large-area patterning of semimetal-
semiconductor Schottky junction. This method would also allow a scaling beyond
2-inch SiC wafers currently available as well as integration with CMOS
electronics.
In section 4.1.2, it is described how metal electrodes and number of active
layers in EG based devices can affect the device photoconductive characteristics.
The asymmetric metallization scheme (EG contacted with gold (Au) and
aluminium (Al) in planar Au/EG/Al fashion) with EG is found to have a great
impact for efficient device photoresponsivity. This scheme has also been utilized
with exfoliated graphene for efficient and ultrafast photodetection due to the
breaking of mirror symmetry of the internal electric field profile.76
Single carbon
nanotube based light emitting diodes with asymmetric contacts have also been
demonstrated to be efficient devices.139
The concept of utilizing EG with
asymmetric contacts has the additional advantage of being compatible with wafer-
scale CMOS electronics.
The transient photocurrent scan across our Au/EG/Al device shows a significant
change at Au/EG and EG/Al contacts. The device with Au/EG contact under
illumination shows positive photocurrent with higher magnitude while EG/Al
under illumination shows negative photocurrent with smaller magnitude. This is
the key feature for the generation of an overall (sum) efficient photocurrent in
device performance. Moreover, the number of active EG layers in device is found
to have significant influence on device performance. We observe that the device
62
with ~ four layers EG has higher photocurrent magnitude (and comparatively
slower response time) than the two layer EG device. The devices in this study offer
many advantages including faster response, high efficiency, low-power
consumption and compatibility with large scale processing techniques.
4.1.1. Laser patterning of epitaxial graphene for Schottky junction
photodetectors
4.1.1.1. Materials and characterization
To accomplish the laser patterning method, few layer EG (~ two layers) and
HOPG (highly oriented pyrolytic graphite) samples were used. First few-layer EG
on the Si-face of SiC substrates (high-purity, semi-insulating 4H-SiC (0001) with
resistivity≥105Ω cm and thickness=0.37 mm) was prepared by thermal desorption
of silicon at high temperatures in UHV (The details for EG preparation are
described in chapter 3). The preparation of few-layer EG was monitored by STM,
as shown in Figure 4.1. Subsequently, the film thickness and its uniformity were
investigated using Raman measurements on 2 mm × 2 mm sample area. The EG
film (~ two layers, determined by width of 2D-band as described in chapter 3) was
found to be continuous and uniform over large area on the surface of SiC substrate
(Figure 4.2). To compare the results in EG with graphite material, HOPG substrate
(purchased from Mateck GmbH Company) was also well cleaved to flatten the
surface prior to using it as a sample for device fabrication using laser patterning. Its
photoconductive measurements similar to the conditions with EG material based
devices were also performed.
63
Figure 4.1. STM images of 1-2 layer EG on semi-insulating Si-face 4H-SiC. (a)
40×40 nm2 STM image (Vtip=0.81 V) showing two regions (single layer and bi-
layer EG) separated by a step of 0.35 nm. Atomically resolved STM images (left
side) of single and bi-layer EG (Vtip = 0.16 V) taken from the highlighted area. (b)
A line profile across the step (indicated as white dash line) extracted from the STM
image in (a).
Figure 4.2. (a) 2D-bands intensities on different locations of 2 mm×2 mm area of
a sample. Raman spectra acquired on several locations (> 20 spots, with spot size
~1 m) show an uniformity in the film thickness (~ two layers) over a large area on
the substrate. (b) shows the width (~ 62 cm-1
) of 2D-band in 2-layer EG to confirm
the number of layers in EG sample.
0 2 4 6 8 10-0.05
0.00
0.05
0.10
0.15
0.20
0.25
0.30
0.35
Z (
nm
)
X (nm)
2650 2700 2750 2800 2850
Ra
ma
n in
ten
sit
y
Wavenumber (cm-1)
2D-bands for three different positions
on a 2mm x 2mm area of the sample
(a) (b)
(a) (b)
2600 2650 2700 2750 2800 2850 2900
Ra
ma
n in
ten
sit
y
Wavenumber (cm-1)
Lorentzian fit to
2D-band
of two layer EG
FWHM ~ 62 cm-1
64
4.1.1.2. Laser patterning and characterization
The laser patterning for single-line structures, as illustrated in Figure 4.3a, was
carried out by the following two steps. Firstly, the Au-coated sample was exposed
to cylindrically focused laser pulses (1 mm×3 mm). To evaporate a part of the Au
film in the focused area, the exposure to ten laser pulses (1.5 mJ per pulse, or
fluence of 0.05 J cm-2
) was required. After the evaporation, the Au coating was
split to two parts, used as two electrodes in photocurrent measurements described
later. Secondly, it was irradiated with more laser pulses at the same energy, turning
EG to laser-modified epitaxial graphene (LEG) with the degree of conversion
depending on the number of laser pulses. The two steps were monitored in-situ
with a digital oscilloscope by collecting transient photocurrents in a short circuit of
the two Au electrodes. Conversion from EG to LEG and formation of EG-LEG
junction were confirmed by the transient photocurrent measurement in the second
step, where the laser-induced photocurrent was enhanced drastically (within 60
laser pulses) and then clamped at a saturated level after completing the conversion
of EG. With this method, a 1 mm×3 mm rectangular area (Figure 4.3a) was created
and two junctions (EG-LEG and LEG-EG) were found at the left and right edge,
respectively. The same method was used to pattern the HOPG sample for
comparison, but no photocurrent was observed with the laser pulse energy up to 4
mJ (or fluence up to 0.2 J cm-2
).
Both EG (unexposed to laser radiation) and LEG were investigated by Raman
spectroscopy. The G-band in the Raman spectra of graphene is related to phonon
vibrations in sp2
bonded carbon materials, while the D-band is the signature for
disorder induced by phonon scattering at defect sites and impurity.42, 106
The
65
spectra in Figure 4.3c were obtained from the patterned areas using 514-nm
excitation wavelength at room temperature and ambient environment. The
spectrum in Figure 4.3c (iv) shows a broader G-band and intense D-band, similar to
the two characteristic peaks of GO (or reduced GO).145, 146
Therefore, it is believed
that LEG consists of GO or reduced GO. It was also observed that once the-
Figure 4.3. (a) Schematic diagram showing both evaporation of a rectangular part
of the Au film and conversion from EG to LEG. (b) In-situ monitoring of transient
photocurrent as a function of laser pulse number after the evaporation of Au film.
(c) Raman spectra recorded after the evaporation of Au film: (i) HOPG unexposed
and (ii) HOPG exposed to 100 laser pulses; and (iii) EG unexposed and (iv) EG
exposed to 100 laser pulses.
20 40 60 80 100
0.4
0.5
0.6
I (m
A)
Number of pulses
1200 1500 1800 2100 2600 2800 3000
(iii)
(iv)
(ii)
Exposed EG
EG
Exposed HOPG
HOPG
Ra
ma
n in
ten
sit
y
Wavenumber (cm-1)
(i)
(a)
(c)
(b)
66
conversion was complete, there was no increase in the transient photocurrent even
after exposure to higher laser pulse energy in the range from 1.5 mJ to 4 mJ. As
comparison, Figure 4.3c(ii) displays the spectrum recorded from the HOPG
exposed to 4-mJ laser pulses, the HOPG sample exposed to 1.5 mJ pulse energy
showed no photo response. At pulse energies greater than 4.5 mJ, the occurrence of
laser-induced damage was observed on the EG or HOPG. Figure 4.4 below shows
the Raman spectra evolution in HOPG exposed to different energy of laser pulses.
Figure 4.4. Raman spectral evolution during laser patterning on HOPG: (i) HOPG
sample exposed to laser pulse energy 1.5 mJ (100 laser pulses) and (ii) HOPG
sample exposed to Laser pulse energy 4 mJ (100 laser pulses).
The D and G bands of LEG in Raman spectrum was further investigated in
detail as shown in Figure 4.5. We found the broadening of these two peaks and SiC
peaks were found to be merged with them, as can be seen by peak fittings in
Figures 4.5b and 4.5c. We also found that the D-band is strengthened when
1200 1500 1800 2100 2600 2800 3000
1.5 mJ
4.0 mJ
Ra
ma
n in
ten
sit
y
Wavenumber (cm-1)
D
G 2D
(i)
(ii)
67
compared to pristine EG. The D-bands indicates the strength of disorder due to
laser induced oxidation in EG.
Figure 4.5. Detailed analyses on the Raman spectra of LEG. (a) Broadening and
strengthening of G-band and D-band in LEG, as compared to them in EG. (b)
Gaussian fit to the G-band in LEG to reveal two SiC peaks and a G-peak of
graphene. (c) Gaussian fit to the D-band in LEG to reveal a SiC peak (green) and a
D-peak (blue) of graphene. Both Raman analyses and non-linear I-V characteristics
(Figure 4.8) suggest that LEG exhibits characteristics similar to reduced graphene
oxide.
The surface morphology after exposure was also recorded using AFM (the details
about instrument and modes of operation are described in chapter 2) as shown in
1200 1400 1600 1800 2000
LEG
EG
SiC (1520)
SiC(1690)
SiC(1410)G(1595)
Ra
ma
n in
ten
sit
y
Wavenumber (cm-1)
EG
LEG
D(1360)
(a)
1250 1300 1350 1400 1450
Ra
ma
n in
ten
sit
y
Wavenumber (cm-1)
Gaussian fit of LEG spectrum (c)
SiC peak
D peak Envelop
1500 1600 1700 1800 1900 2000
SiC peak 2
SiC peak 1
Ra
ma
n i
nte
ns
ity
Wavenumber (cm-1)
Guassian fit of LEG spectrum (b)
G peak
Envelop
68
Figure 4.6. The image shows the surface looking almost similar with no significant
change (or peeling of the film).
200 nm 200 nm
(a) (b)
Figure 4.6. Surface morphology using AFM.(a) an AFM image of graphene film
before exposure and (b) an AFM image after exposure. Surface similarity with no
significant change (or peeling of the film) supports that the overall structure of EG
remains intact.
Mechanism of EG conversion to LEG
The mechanism of converting EG to LEG (GO or RGO) can be understood as a
laser-induced oxidation/reduction process. The absorption of intense laser energy
by EG gives rise to an increase in the temperature in the exposed area/volume,
which in turn oxidizes EG itself in ambient environment by absorbing oxygen in
the air. The laser-induced rise in the temperature of a material is proportional to the
laser fluence (F), as given by the following equation,147
where B is the exposed area of the material, is the efficiency of conversion from
the absorbed light energy into heat, is the density of the material, C is the specific
heat of the material, d is the material thickness and V is the exposed volume. With
CV
dFBT
)]exp(1[
[4.1]
69
F = 0.05J cm-2
, an optical absorbance of bilayer EG = [1- exp(–αd)] ≈ 0.05 and an
assumption of = 100%, we calculate T in the range of 450-600C. Such high
temperatures were sufficient to the oxidation/reduction in air.148, 149
4.1.1.3. EG-LEG-EG photodetector
The schematic of EG-LEG-EG photodetector and its transient photoconductive
response with 1064 nm laser pulses are shown in Figure 4.7. The results show an
efficient and fast (nanosecond) photoresponses. Furthermore, it is found that the
device is also responsive to white light as shown in Figure 4.8 as I-V
characteristics.
Figure 4.7. (a) Schematic of EG-LEG-EG photodetector. (b) Transient
photocurrents across the zero-biased single-line EG-LEG-EG junction device over
one second when one of the two junctions is excited by laser pulses of 7-ns
duration and 10-Hz repetition rate at 1064-nm wavelength and 2.4-mJ pulse
energy.
To gain more insights, we conducted position-dependent photocurrent scans,
spectral response (wavelength vs. photoresponsivity) measurements with white
(a) (b)
0.0 0.2 0.4 0.6 0.8 1.0
0.3
0.6
0.9
Cu
rren
t (m
A)
Time (s)
70
light and more photoconductive characterizations of EG-LEG-EG photodetector,
and which are described in subsequent sections.
-15 -10 -5 0 5 10 15
-150
-100
-50
0
50
I (m
A)
Voltage (V)
Dark (Single line)
Light (Single line)
Figure 4.8. I-V characteristics of the single-line EG-LEG-EG Schottky junction
device with white light illumination at 100 mWcm-2
and without light illumination.
4.1.1.3.1. Position-dependent photoresponse observation
The 7-ns laser pulses were focused by a focusing lens onto the above-described
device with a circular spot diameter of ~200 m. Figure 4.9a displays both optical
image and Raman spectra obtained from different locations on the device.
Evidently, in the region near to the Au electrode (the dashed line in Figure 4.9a),
EG is un-converted due to lower laser irradiation resulting from the edge in the
Gaussian spatial profile of laser pulses; the central region (blue in Figure 4.9a) is
fully converted to LEG, supported by the Raman spectra (Figure 4.9a). Thus, two
Schottky junctions are produced in a back-to-back format, as shown by the
schematic in Figure 4.9b. Figure 4.9c shows that the photocurrent is maximized
71
(absolute value) at ±0.5 mm from the center, with polarity consistent with the
observation from graphene-metal junction devices.76, 150
Figure 4.9. (a) Optical image and Raman spectra measured from the un-converted
EG (black) and LEG (red) region. (b) Schematic of photocurrent line scanning
across the EG-LEG-EG junction. (c) Maximal transient photocurrent and (d)
photocurrent decay time as a function of the distance from the center when
illuminated by laser pulses at 50 J per pulse. The green line in (a) represents the
spatial profile of the laser pulse used for conversion from EG to LEG. The green
spot in (b) represents the area illuminated for the photocurrent measurement.
-1.5 -1.0 -0.5 0.0 0.5 1.0 1.5
-0.6
-0.4
-0.2
0.0
0.2
0.4
0.6
0.8
I (m
A)
Distance (mm)
-1.5 -1.0 -0.5 0.0 0.5 1.0 1.520
30
40
50
d
ec
ay
(ns)
Distance (mm)
(b) (a)
(c)
(d)
72
The formation of two Schottky barriers manifests itself in the measured signals
where the maximum indicates the maximum built-in field. If the effective internal
field width is defined as the full width of half maximum (absolute value) in the
photocurrent measurement, a width of ~400 m can be determined from Figure
4.9c. It is in this region (400 m× 3 mm) where the separation of photo-generated
pairs of charge carriers (electrons or holes) occurs and subsequently, separated
charge carriers are collected by the Au electrode as the measured photocurrent.
The junction formation, coupled with high mobility of graphene enables photo-
generated charge carriers to be separated and extracted efficiently by the Au
electrodes, thereby shortening the photocurrent recovery time, as supported by our
transient measurements. The measured photocurrent decay times depend on the
location, as shown in Figure 4.9d, the further away from the Au electrodes, the
longer the decay times. The discontinuity in the decay time implies that the
photodynamic mechanism in the LEG region be entirely different from the EG-
LEG junction. No photocurrent was observed when the measurement was
conducted on the central LEG region or the Au film.
4.1.1.3.2. Temporal photocurrent profiles
Figure 4.10 shows the temporal photocurrent profiles of the EG-LEG-EG
junction when one junction is excited at 1064 nm and 532 nm laser pulses.
Photoconductive responses of the EG-LEG-EG junction depend on the excitation
energy. As the excitation pulse energy grows, the photocurrent tends to saturate, as
shown in Figure 4.10c. The saturation could be a result of saturation of light
absorption.151
Photo-thermoelectric effect across the interface between two
73
different materials due to high photo-excited carrier density could also account for
the saturation of photocurrent.152
Figures 4.10a and 4.10b display the laser-pulse-
energy-dependent temporal profiles of photocurrent pulses excited at 1064 nm and-
Figure 4.10. Temporal photocurrent profiles of the EG-LEG-EG junction when
one junction is excited at (a) 1064-nm and (b) 532-nm laser pulses. The inset in (a)
shows the transient photocurrent profile of the HOPG sample at 1064 nm. The red
line is a single exponential fit. Photocurrent maxima (c) and decay times (d)
excited at 532nm and 1064nm as a function of laser pulse energy.
532 nm, respectively, which fit well with the single exponential decay time and the
rising time of ~7 ns. As shown in Figure 4.10d, the decay becomes slower with the
increase of excitation pulse energy. At higher laser excitation, more carriers are
photo-generated, resulting in slower decays due to carrier-phonon scattering in EG,
0 200 400 600 800 1000
0.0
0.5
1.0
0 2000 4000
0.00
0.09
0.18
I (m
A)
Time (ns)
4 mJ
0.15 mJ
0.35 mJ
0.75 mJ
1 mJ
2.4 mJ
I (m
A)
Time (ns)
0 1 2 30.0
0.4
0.8
1.2 1064 nm
532 nm
I (m
A)
laser pulse energy (mJ)
0 200 400 600 800 1000-0.3
0.0
0.3
0.6
0.9
0.26 mJ
0.98 mJ
1.85 mJ
I (m
A)
Time (ns)
0 1 2 3
10
20
30
40
50
1064 nm
532 nm
laser pulse energy (mJ)
d
eca
y (
ns)
(c) (a)
(d) (b)
74
LEG or substrates.151, 153
The photocurrent decay times are faster at 532 nm, as
compared with 1064 nm, see Figure 4.10d. This is consistent with avalanche photo
diodes whereby exciting with higher energy photons results in an increase in the
kinetic energy of photocarriers, giving rise to shorter decay times.154, 155
Figure 4.10c also shows that the same magnitude of photocurrent is observed for
the two wavelengths at the same pulse energy, indicating that the photo
responsivity is independent of light wavelength. Such wavelength independence
indicates that EG plays an important role in the photo-generation, consistent with
the light-wavelength-independent absorption characteristic of graphene.30, 33
This is
completely different from conventional semiconductor photon detectors whose
optical responses strongly depend on light wavelength.
4.1.1.3.3. Spectral photoresponse investigation
The wavelength independence was further validated by our DC photocurrent
measurement in the spectral range from 200 nm to 1000 nm. From the
measurement in Figure 4.11, the internal photoresponsivity (or efficiency) defined
below is found to be ~3.3 mA W-1
.
where Iph is the collected photocurrent, B is the illuminated EG-LEG surface area
(0.4 mm×2 mm), and I() is the intensity spectrum of the incident light.
The measured spectral responsivity shows a uniform response (within the error
of 10%) covering from ultraviolet (200 nm) through visible to infrared (1000 nm)
)()]exp(1[)(
BId
IR
ph
[4.2]
75
wavelengths. Figure 4.11 also shows the noise level measured under dark, the
origin of which is still unclear and more investigations are required. Such a unique
response makes the EG-LEG-EG photodetector versatile for many applications. It
should be mentioned that an external photoresponsivity of 6.1 mA W-1
in a biased
graphene photoconductor was observed at 1.55 m.76
As such, we conclude that
Figure 4.11. Internal spectral responsivity of the EG-LEG-EG junction device
under zero-bias.
graphene-based photoconductive devices are not only excellent visible-infrared
sensors, but they are also superior UV detectors.
200 400 600 800 1000
0.0
0.5
1.0
1.5
2.0
2.5
3.0
3.5
Re
sp
on
siv
ity
(m
A W
-1)
Wavelength (nm)
Dark
Light
76
4.1.1.3.4. Interdigitated EG-LEG-EG device
Interdigitated EG-LEG-EG device was also fabricated, and the top view of the
device is shown in the inset of Figure 4.12a. In this device, the structure consists of
-15 -10 -5 0 5 10 15
-6
-4
-2
0
2
Dark
Light
Cu
rren
t (m
A)
Voltage (V)
Figure 4.12. (a) I-V characteristics (filled squares) of the interdigitated device with
(red) and and without (black) white light illumination at 100 mW cm-2
. The solid
curves (blue) are the modeling, and schematic of the device-model is shown in (b).
The inset shows the top view of the device with the scale bar=1 mm.The
asymmetry with respect to the bias voltage is primarily due to the dissimilarity of
the two junctions. (c) Schematic of the energy band diagram corresponds to EG-
LEG-EG structure.
IP1 IP2
EG
LEG
EG
Au Au
(c)
(b)
(a)
77
several vertical and horizontal lines. The length of each vertical line was controlled
by the size of a rectangular aperture placed in front of the cylindrical lens. The
horizontal line was made by rotating the device 90° around the laser beam. The I-V
curve of the device under dark exhibits strong semiconducting behavior, as shown
in Figure 4.12a. The photoresponse was measured when the entire surface of the
inset in Figure 4.12a was illuminated with the same solar simulator. From Figure
4.12a, an external photoresponsivity (or efficiency) of 0.1 A W-1
(after the dark
current was deducted) is achieved at a bias voltage of -10V.
The I-V curves in Figure 4.12a can be explained by a commonly-used circuit
model in which two opposing Schottky diodes are connected in a series as shown
in Figure 4.12b. The corresponding band diagram and photo-excitation are
displayed in Figure 4.12c.
The non-linear I–V characteristic of each Schottky junction can be expressed by the
thermoionic emission model.156
The total current (I) and bias voltage (V) are given
by
where A* is the effective Richardson constant, n is the diode ideality factor, q is the
electron charge; k is the Boltzmann constant, and T (= 297 K) is the temperature.
IP1, IP2 are the photocurrents, A1, A2 are the contact areas between EG and LEG,
V1, V2 are the voltage drops, and φ1, φ2 are the Schottky barrier heights of diode 1
and diode 2, respectively. VLEG is the voltage drop across the LEG region. The
fittings to both dark and illuminated I-V curves were achieved at φ1 = 0.75 eV, φ2 =
LEG
nkTqV
kTP
nkTqV
kTP
VVVV
eeTAAIeeTAAII
21
2
22
2
11 112211
[4.4]
[4.3]
78
0.70eV, IP1 = IP2 = 0 for dark, and IP1 ≈ 13 mA, IP2 ≈ 16 mA under light
illumination. By considering A1 = A2 (≈ thickness of EG × length of Schottky
junction ≈ 910-8
cm2) and n = 2, the effective Richardson constant A* extracted
from the best fit is 8850 Acm−2
K−2
, comparable to the reported constant for the Pt-
Si Schottky junction.157
The two Schottky barrier heights (φ1 = 0.75 eV, φ2 =
0.70eV ) are close to the reported value of 0.7 eV by Wu, et al.,144
indicating that
LEG is a semiconductor. In the curve fitting, voltage drops were floated, the best fit
was achieved at V1 = 0.07V, V2 = 0.08V, revealing that 85% of the total bias
voltage is dropped on the LEG region as VLEG = 0.85V. This provides a practical
means to reduce the operating bias voltage by designing a narrower width of the
LEG region.
The I-V characteristics of the above device suggest that LEG should be a
semiconductor.
4.1.1.4. Infrared absorption spectrum of LEG
To further support that the LEG is a semiconductor, we carried out an additional
experiment: the total area of a 3 mm × 10 mm EG sample was modified to LEG.
Then, its infrared-absorption spectrum was measured as shown in Figure 4.13.
From Figure 4.13a, a plot shown in Figure 4.13b was generated. An extrapolation
of the straight line to = 0 axis indicates the bandgap. A bandgap of ~ 0.63 eV was
obtained by this method. This observation further indicates that LEG indeed shows
semiconducting properties, with a defined bandgap in infrared region of
electromagnetic spectrum.
79
Figure 4.13. (a) Infrared absorption spectrum of LEG and EG and (b) plot of
(absorption coefficient ×photon energy)2 as a function of photon energy. The plot
in (b) is converted from (a) and the intercept between the straight line (green) and
the x-axis (band gap is measured by extrapolating the straight line to the = 0 axis)
indicates the band gap.
4.1.1.5. EG-LEG-EG characteristics summary, comparison and advantages
The summarized photoconductive characteristics of EG-LEG-EG photodetector
and its comparison with previously reported carbon based photodetectors are
presented in Table 4.1. From the Table, we can infer EG-LEG-EG as one of
potential candidates among carbon based photodetectors.69, 76, 136, 137, 158, 159
In the case of graphene based photodetector, light absorption (material
thickness, absorption coefficient) property, charge carrier transport and
recombination in graphene would be the factors which can affect the
photoresponsivity and response speed in a device. For example, one or two layer
graphene show very little (6.1 mA W-1
) photoresponsivity and it can be understood
due to the weak absorption of light (2.3%-4.6%). Previous studies show ultrafast
800 1200 1600 2000 2400
0.06
0.12
0.18
Ab
so
rba
nc
e
Wavelength (nm)
(a)
LEG
EG
0.6 0.9 1.2 1.5-0.28
0.00
0.28
0.56
(h)2 x
10
14
(eV
cm
-1)2
Photon energy (eV)
Bandgap of LEG ~0.63 eV(b)
LEG
EG
80
(ps) carrier recombination in graphene.71
Faster the recombination can result in
lower photoresponsivity. In the case of our EG-LEG-EG device, LEG acts as the
active layer for light absorption (absorbance ~ 14%) while EG can efficiently
extract the photogenerated carriers. The combined effect causes the overall
efficient performance. Furthermore, Schottky junction EG-LEG-EG photodetectors
Table 4.1. EG-LEG-EG characteristics and its comparison with other carbon based
photodetectors
Active
material/device
External
photoresponsivity
Bias voltage
Response
time
References
EG-LEG-EG
(Single line
structure)
0.5 mA W-1 Zero-bias 50 ns
This work
EG-LEG-EG
(Interdigitated
Structure)
0.1 A W-1 -10 V This work
Pristine graphene
(exfoliated)
0.5 mA W-1 Zero-bias < ns 69
Pristine graphene
with asymmetric
contacts
6.1 mA W-1 0.4 V < ns 76
RGO 0.12 A W-1
1 V 35 min 136
RGO 4 mA W-1 2 V 29 s 137
GNR 1 A W-1
2 V 96 s 137
CNT 7.9 mA W-1 1 159
CNT/organic
semiconductor/C60
21 mA W-1
<1 (7.2±0.2) ns 158
81
in this study offer advantages as follows. (1) A simple method of fabrication. (2)
Large area processability (a 1 mm × 3mm device is working) of the devices.
Internal (built-in) electric fields, responsible for the separation of photo-generated
charge carriers (electrons or holes), were found to be present in narrow regions
(~0.2 m) adjacent to the electrode-graphene interfaces in graphene FET
photodetectors.76
In our device, the width of internal field is as large as ~ 400m,
thereby enhancing the capability of electron-hole separation in a considerable
amount and leading to a great efficiency of conversion from light to photocurrent.
Previously reported graphene- or GO-based devices160
showed microsecond
photoconductive decays. Our zero-bias devices exhibit nanosecond recovery times.
The most important feature is that the photoresponse of our devices is independent
of light wavelength from ultraviolet (200 nm) through visible to infrared (1064
nm), a unique optical phenomenon only manifested itself in graphene as an active
photosensor.
Studies involving device structures with reduced length scales, the influence of
number of active graphene layers, as well as graphene-metal contact effects could
help further improve the device performance.
4.1.2. Photoresponse in EG with asymmetric metal contacts
4.1.2.1. Materials and device fabrication
In this study, EG samples (~ two layers and four layers) prepared on high purity
semi-insulating Si-face of 4H-SiC substrates were used to fabricate the devices
with asymmetric (Au, Al) metals contacts (the details for materials synthesis can be
found in chapter 3). To fabricate the devices, Au and Al metals (asymmetric
82
contact) were thermally evaporated on the left and right sides of pristine EG
samples through a shadow mask. The device schematic and optical micrograph of a
fabricated EG with asymmetric contacts are shown in Figure 4.14.
Figure 4.14. (a) Schematic view of zero-bias Au/EG/Al device connected to a
cathode ray oscilloscope (CRO). (b) Optical micrograph of the EG device
contacted with Au (yellow) and Al (white) metals.
The graphene thicknesses were two and four layers, Raman measurements showed
FWHM (2D) of EG samples are 60 and 73 cm-1
, indicating that two and four layers
were grown on SiC surface, as shown in Figure 4.15.
83
Figure. 4.15. The width (FWHM) of 2D band in Raman spectrum has been used to
analyse the number of layers in EG on Si-face SiC (describe in chapter 3 in detail).
(a) The Raman 2D-band of EG samples having width (FWHM) ~ 60 cm-1
and 73
cm-1
that indicates the thickness ~ two and four layers. The inset shows the
corresponding STM images of two ( Vtip = 0.16 V) and four layers (Vtip = 0.21 V)
EG. (b) 2D-band spectra at different positions (left-middle-right) acquired on 0.05
mm × 2 mm area devices with (i) two layer EG (ii) four layer EG. The measured
2D-band at several points (> 20 points on each device ) showed no significance
variation in intensity or width indicating fair layer thickness uniformity (two and
four layers) in each devices.
(a)
(b)
84
4.1.2.2. Photoconductive response of Au-EG-Al device, and influence of EG
active layers over the device performance
Transient photocurrent measurements were performed on the device (with two
layer EG as the active layer) by scanning the laser beam (total scan size ~ 100 µm)
from Au to Al electrodes. The laser beam was provided by a Nd:YAG laser
(Spectra Physics Quanta Ray) with 532-nm wavelength, 7-ns duration, and 10-Hz
repetition rate laser pulses and was focused onto the device with a spot size of ~20
m. The temporal photocurrent profiles (as shown in Figure 4.16a) were collected
with a digital oscilloscope(Tektronix TDS 380, 50 ohm terminated, 400 MHz
bandwidth); and the positive and negative terminals of the oscilloscope were
directly connected to the Au and Al electrodes, respectively. All the measurements
were carried out at room temperature and ambient condition. The device shows the
maximum photoresponse near the metal/EG contacts and the photocurrent
decreases going away from the Au/EG or EG/Al interface as shown in Figure
4.16a. Interestingly, we observe that photoresponses near the two contacts (Figure
4.16b) are different in polarity and magnitude. The significant difference in
photocurrent magnitudes near the two asymmetric contacts is responsible for the
overall (sum) higher device photoresponse.76, 115
The finding shows that the work
function change of the two metals contacted in planar geometry can dramatically
affect the photoresponse by breaking the mirror symmetry in graphene based planar
devices. The result shows that the photocurrent is pronounced near that
metal/graphene contact for which work function change between the two is larger,
resulting in p-n junction formation. Lowering the photocurrent magnitudes (by
contacting graphene with metal of matching work function) near one of the two-
85
Figure 4.16. (a) Photoresponse of the Au/EG/Al device at different positions
excited with laser pulses (photo scan from Au to Al electrode). (b) Plot of
maximum photocurrent (PC) near Au/EG and EG/Al contacts. (c) Photoresponses
(maximum photocurrent) in devices with two and four layers of EG. Laser
pulseduration: 7 ns, wavelength: 532 nm, and fluence: 0.94 mJ.
metal/EG contacts can, in principle, result in overall (sum) higher
photoconductivity in the device as supported by our experimental observations. As
such, for the given work functions42, 161
(øEG ~ 4.0 eV, øAu ~ 5.47 eV and øAl ~ 4.06
eV ), the Au/EG/Al device should result in formation of p-n-n junctions.
86
We observe that the number of active EG layers in the device also influences the
device photo-to-electricity conversion efficiency. Figure 4.16c shows the transient
photo current measurement (with laser pulses of 532-nm wavelength, 0.94-mJ
fluence) performed on the devices having two and four layers of EG as the active
layers. It is noted that the photocurrent is enhanced in the four EG layer device.
This could be attributed to the larger number of incident photons absorbed with an
increased number of EG layers. However, in general, the photoresponse time
(decay time td) of the device increases with increase in photocurrent. This may be
attributed to carrier-carrier scattering (that slows down the device photoresponse)
due to the larger number of photogenerated charge carriers responsible for higher
photocurrent.156, 162
Our finding suggests that the photoresponse time of the EG/Al
junction may even be shorter than 7 ns and our measurement is limited by the
excitation duration.
In order to determine the photon-to-electron conversion efficiency, further
photocurrent measurements were performed on biased devices (Figure 4.17a, active
detector area of 0.05 mm × 2.0 mm) under continuous illumination of He-Ne laser
(wavelength ~ 632.8 nm, spot size ~ 2.0 mm). The voltage(V)-current(I)
characteristics of the 4-layer EG device with and without light illumination are
shown in above Figure 4.17b. The inset shows the bias dependence of external
photoresponsivity of the device. From the measurements, the maximum external
photoresponsivity of ~ 31.3 mA W-1
is estimated at a bias voltage of ~ 0.7 V. The
extracted photoresponsivity ~ 4.9 mA W-1
even at 0 V (Figure 4.17b) further
supports our photoresponse observations in the zero-bias device as shown in Figure
4.16. However, we found ~ 60% lower photoresponsivity in the device with two-
87
Figure 4.17. (a) Schematic of biased Au/EG/Al device. (b) Device I-V
characteristics with and without light illumination at 632.8 nm wavelength. The
inset shows the measured external photoresponsivity (PR) as the function of the
bias voltage. The I-V curves in black and blue represent dark and photocurrent
respectively.
layers of EG (as shown in Figure 4.18).
The device photoresponsivity observed in this study is comparably higher (~ 5-
times) than a previous report (6.1 mAW-1
) on pristine graphene.76
Furthermore, we
observed that EG devices with symmetric metal contacts (Au/EG/Al or Al/EG/Al
devices) show no significance photoresponses (Figures 4.18a, 4.18b) under
illumination and this highlights the importance of asymmetric metallization scheme
in graphene-based photodetectors. This scheme can easily be processed using-
88
Figure 4.18. I-V characteristics (EG ~ two layers) acquired on (a) Al/EG/Al, (b)
Au/EG/Au and (c) Au/EG/Al devices with and without light illumination at 632.8
nm wavelength (Power = 0.96 mW). The top-left inset in (c) shows the external
photoresponsivity in the asymmetric Au/EG/Al device, and the bottom-right inset
shows the zoom-in view around Vbias = 0 V , which clearly indicates the
photoresponse at 0 V.
standard CMOS lithography on 2-inch SiC wafers (unlike for exfoliated graphene
due to its smaller flake sizes). It offers the large detection area that is demanded for
many practical applications.
(c)
(a) (b)
89
\4.1.2.3. Au-EG-Al characteristics summary, comparison and advantages
The summary of Au-EG-Al photodetector characteristics and its comparison
with other carbon based photodetectors76, 136, 137, 159
are presented in Table 4.2.
From the Table 4.2 it is evident that the Au-EG-Al photodetectors in this study are
having significant relevance among many carbon based photodetectors.
Through this study we find that method of asymmetric metallization can be
accomplished with pristine graphene to produce relatively (compared to single line
LEG photodetector) efficient photodetectors. However, Au-EG-Al devices require
narrow channel length (gap between two -electrodes, Figure 4.14b) when compared
to the LEG (1 mm) device described in the previous section. As such, mobility of
EG sample, device structure (narrower channel length) are the important factors,
that may further improve the efficiency as well as response time of the devices,
which have potential relevance in high speed optical communication applications.
Table 4.2 Summary of Au-EG-Al photodetector device characteristics and its
comparison with other carbon based photodetectors.
Active
material/device
Photoresponsivity Response
time
Bias voltage References
Au-EG-Al
EG: two layers
12.5 mA W-1
60 ns
0.3 V
This work
Au-EG-Al
EG: four layers
31.3 mA W-1 75 ns 0.7 V This work
Pd-G-Ti
(G: exfoliated
graphene)
6.1 mA W-1 < ns 0.4 V 76
RGO 0.12 A W-1 35 min 1 V 136
RGO 4.0 mA W-1 29 s 2 V 137
GNR 1.0 A W-1
96 s 2 V 137
CNT 7.9 A W-1 < 1 V 159
90
4.1.2.4. Conclusion
In summary, this chapter describes the photoconductors based on laser modified
EG on SiC (section: 4.1.1), and pristine EG with asymmetric metals (Au, Al)
contacts (section: 4.1.2).
Section 4.1.1 presents fabrication of a large-area EG-LEG-EG Schottky junction
device on SiC substrates with a simple laser irradiation method. These zero-biased
junction devices show nanosecond photo responses, making them promising for
high-speed applications. More importantly, the devices exhibit a uniform
broadband (200 nm-1064 nm) photoresponse, demonstrating that the EG-LEG-EG
Schottky junction devices are not only excellent visible and infrared sensors, but
they are also superior UV detectors. The uniform photoresponse in EG-LEG-EG is
distinctively different from conventional semiconductor photon detectors whose
photoresponse strongly depends on light wavelength. An efficient external
photoresponsivity (or efficiency) of ~0.1 A W-1
is achieved with a biased
interdigitated EG-LEG-EG photodetector. The fabrication method studied here
opens a viable route to carbon optoelectronics for fast and highly-efficient
photoconductive detectors.
Section 4.1.2 demonstrates the photoresponse in EG devices with asymmetric
Au and Al metallization scheme. The zero-bias device photoresponse directly
observed in digital oscilloscope shows significance differences in photocurrent
magnitudes near the Au/EG and EG/Al contacts, resulting in significantly higher
device photoresponse efficiencies. Specifically, it is observed that the number of
layers in EG further influences the photocurrent magnitude and response time
regardless of incident photon energy or intensity. The maximum external
91
photoresponsivity ~ 31.3 mA W-1
at a bias voltage ~ 0.7 V under illumination with
632.8 nm wavelength is estimated. The finding shows that graphene devices with
asymmetric metal contacts can be used as photodetectors that are fast, efficient and
with low power consumption.
92
CHAPTER 5
High-Gain Photodetectors Based on Oxygen Plasma
Treated Epitaxial Graphene
5.1. Introduction
IR and UV photoconductors based on thin films of RGO or graphene
nanoribbons (GNR) have also been demonstrated by other groups with device
photoresponsivity up to 4 mA W-1
for IR light and up to 0.12 A W-1
for UV light.136
RGO is a promising candidate in organic or plastic electronics.163, 164
However,
compared to pristine graphene, the electrical property of RGO is poor, thus
limiting its ultimate utility in high-performance electronics.165, 166
In the previous
chapters, I have developed high-speed photodetectors based on EG materials. In
the context of carbon based materials, the carbon nanotube is the other promising
candidate. Various photoconductors have been demonstrated, either using pristine
carbon materials76, 138, 167
or a hybrid with other existing organic158
or inorganic
materials.63, 168
Furthermore, it has been observed that the attachment of quantum
dots on the graphene surface enhances the photoconductivity enormously in
devices.67, 169, 170
Oxygen plasma has been extensively utilized to tailor/improve the
organic,171
inorganic172
or carbon materials based devices.173, 174
In contrast to wet
chemical oxidation methods, it is a dry, and efficient CMOS-compatible technique.
Importantly, if the gas used is oxygen, the plasma is an effective, economical,
environmentally safe method for use in devices.
93
This chapter describes my successful development of a simple, and dry method
of surface modification in ultra-thin few layer EG devices using controlled mild
oxygen plasma treatment. My finding shows that this method can easily be
implemented to produce high-gain large area graphene based photoconductive
devices. Controlled low power plasmas, used in this study, do not produce a
substantial damage in EG. As such, depending on the annealing temperature, the
conversion back to pristine EG device is possible. The method also offers further
optimization and improvement by technically altering the device length scale,
structure etc, leading to commercial devices.
The trapping of charge carriers in nanostructured devices has produced high
photoresponsivity by prolonging the carrier life time.67, 175
We have observed large-
photoconductive gain in oxygen plasma treated EG devices in the UV spectral
range. A carrier trapping mechanism, that results in high photoconductivity in the
device, is proposed after combined investigations using XPS, Raman, electrical I-V
characteristics, and photoconductive characteristics. The device high-gain (~ 4-5
orders of magnitude higher than previously reported graphene based devices76, 136
),
high detectivity (~ 1012
Jones) and large selectivity in UV spectral range, highlight
its potential advantages over Si and other organic, inorganic or hybrid materials
based UV photodetectors.2,5,40,43
Nevertheless, the chapter also proposes potential
applications (described in section 5.6) of the photoconductor device developed in
this study.
94
5.2. Materials and device fabrication
In this study, few layer (~ 4 layer) EG were used to fabricate two-terminal
devices. Materials (EG) thickness and its uniformity were checked with Raman
measurements as previously described section 3.3.2.
A thermal evaporator was used to deposit Cr(10 nm)/Au(100 nm) metals
through a shadow mask onto pristine EG to fabricate two-terminal devices.
Subsequently, the EG device with two metal electrodes was mounted in a specific
sample holder for wire-bonding. Two copper wires (long enough to facilitate for
device measurement, easy handing without affecting the EG material and its wire-
bonding connections), were soldered to the electrodes. The photograph of the wire-
bonded EG device is shown in Figure 5.1.
Figure 5.1 (a) The EG device wire-bonded onto a sample holder (or “chip
carrier”). The two long wires soldered to the two electrodes on EG can easily be
used for device measurement connections. (b) Optical picture of Cr/Au contacts on
EG showing an active area ~ (170 μm ×300 μm).
(a)
(b)
95
5.3. Oxygen plasma treatments of EG device
Prior to the device treatment with oxygen plasma, I-V characteristics, Raman
signature and photoconductive characteristics of the pristine EG device were
recorded.
Figure 5.2. (a) The specific plasma system (reactive ion etching (NTI RIE-2321)
system) (b) Enlarged view of the control monitor.
96
The device was exposed with oxygen plasma (flow rate: 20 sccm, chamber
pressure ~ 50mTorr, rf frequency plasma power = 6 W) for two different time
durations: 20 s (OPTEG1 device) and 40 s (OPTEG2 device). Each time, the
device photoconductive characteristics were investigated. The specific plasma
system (NTI RIE-2321) used in this work is shown in Figure 5.2.
5.4. Device measurements
Device I-V characteristics (under dark condition) were conducted using Keithley
2400 (Figure 5.3a). Temporal photocurrent measurements were performed with an
Agilent B2912A (Figure 5.3b), with a xenon arc lamp as a white light source. The
ON and OFF light switching was performed at 30 and 40 second intervals in
OPTEG1and OPTEG2 devices respectively.
Figure 5.3. (a) Keithley 2400 current source/measurements unit used for I-V
measurements under dark condition. (b) Agilent B2912A precision
source/measurements unit used for temporal photoresponse measurements.
97
5.5. Results and discussion
5.5.1. Photoconductive response
Figure 5.4. (a) Photograph of the real device under white light. (b) A schematic of
the device showing pristine EG and oxygen plasma treated EG (OPTEG) region as
an active area. (c) I-V characteristics, showing change in channel resistance of the
device during treatments for 20 s (OPTEG1) and 40 s (OPTEG2). The inset shows
the optical picture of Cr/Au contacted two electrode device. (d), (f) Temporal
photocurrent Iph obtained from devices with OPTEG1 and OPTEG2 active
channels under white light (light intensity ~ 30 mW cm-2
, bias voltage = 2 V). (e),
(g) Rise (red) and decay exponential (blue) fittings to the data in (d), (f), estimating
response times of OPTEG1(tr = 7 s, td = 12 s) and OPTEG2 (tr = 18 s, td = 37 s)
devices respectively.
Figure 5.4a shows the real device under white light. A schematic of the
photoconductor device is shown in Figure 5.4b. Figure 5.4c shows the I-V (under
dark) characteristics monitored from pristine EG device and the device after
treatment processes: 20 s (OPTEG1) and 40 s (OPTEG2). A modulation of channel
98
resistance (resistance increases ~3-5 times in treated devices with respect to the
resistance of pristine EG channel) after oxygen plasma treatments for 20 s and 40 s
is observed, as shown by I-V in Figure 5.4c . This gives a hint to the formation of
defect traps on the surface which restricts the flow of free carriers, leading to the
increase in resistance of the channel. Figures 5.4d and 5.4f show the temporal
photoresponse obtained from devices (at fixed bias 2 V) illuminated with white
light (intensity ~ 30 mW cm-2
). Compared to the pristine EG device (which is
nearly insensitive to light), an enhanced photoresponse (Figure 5.4d) can be seen in
the device treated with oxygen plasma for 20 s (OPTEG1). Noticeably, an increase
in treatment time to 40 s (OPTEG2), gives an enhancement in photocurrent
magnitude and response times (tr = 18 s td = 37 s), as presented in Figures 5.4f,
5.4g. Figures 5.4e and 5.4g show the exponential rise and decay fits176
to the
temporal profiles obtained from OPTEG1 and OPTEG2 devices respectively. We
find that the treated devices show a relatively slower response times compared to
response time of ≤ ns in pristine graphene devices76, 114, 115
and its value is found to
increase with plasma treatment time. The enhanced large photoconductivity and
slower response times suggest the formation of localised charge traps, with its
density proportional to the treatment time. These charge traps could be formed due
to the adsorption of oxygen-related species/creation of carbon vacancies on the
surface. To obtain further insight, surface of OPTEG was investigated using core-
level photoelectron spectroscopy (PES), ultraviolet photoelectron spectroscopy
(UPS), and Raman measurements as described in sections 5.5.2, 5.5.4 respectively.
99
5.5.2. Photoelectron spectroscopy (PES) investigation
Changes in the surface chemistry of oxygen plasma treated EG were
investigated by synchrotron-based high-resolution core level PES (energy hν = 650
eV ) and UPS (energy hν = 60 eV) measurements. Figure 5.5a shows the wide-scan
core-level PES spectra obtained from pristine EG and OPTEG.
Figure 5.5. Core-level PES of pristine EG and OPTEG. (a) A wide-scan spectrum
obtained from pristine EG and OPTEG. (b) Si 2p spectra of pristine EG and
OPTEG. (c) C 1s of pristine EG. (d) C 1s of OPTEG showing the content of
chemisorbed C-O and C=O species.
We can see the appearance of O 1s peak (at 532.18 eV) in OPTEG and it
indicates the population of oxygen species on the surface. The Si 2p spectra (Figure
5.5b) show no significant chemical change, suggesting underlying SiC is intact
after treatment of EG surface. Furthermore, the careful analysis of C 1s peak (at ~
0 100 200 300 400 500 600 700
C 1s
O 1s
Inte
nsit
y (
a.u
.)
Binding Energy (eV)
Pristine EG
OPTEG
C 1sh = 650 eV
98 99 100 101 102 103 104 105
OPTEG
Inte
ns
ity
(a
.u.)
Binding energy (eV)
Pristine EG
Si 2p
281 282 283 284 285 286 287 288 289
Inte
ns
ity
(a
.u.)
Binding Energy (eV)
Pristine EGC1s
281 282 283 284 285 286 287 288 289
C=O
C-O
Inte
ns
ity
(a
.u.)
Binding Energy (eV)
C1s
C=C
OPTEG
(b)
(c)
(a)
(d)
100
284.6 eV), as shown in Figure 5.5d, shows chemisorbed C-O (at~ 286.0 eV ) and
C=O (at~ 287.3 eV) species173, 177
on OPTEG surface.
2.5 2.0 1.5 1.0 0.5 0.0
Inte
nsit
y (
a.u
.)
Binding Energy (eV)
Pristine EG
OPTEG
h = 60 eV
Figure 5.6. UPS of pristine EG and OPTEG.
Figure 5.6 shows the UPS spectra of pristine EG and OPTEG. The spectra look
similar except for a slight attenuation of electronic density of states (π-states of
graphene surface45
in OPTEG), that can occur due to chemisorptions or disruption
of sp2 graphene network. However, we have no evidence of bandgap opening at
the Fermi level using this observation. A flat band below the Fermi level is
identified as a band gap in hydrogenated graphene178
; we do not observe such a flat
band (Figure 5.6). This could be due to the overall effect that we see from 4-layer
graphene.
101
5.5.3. Temperature-dependent resistance and Arrhenius plot
By performing 4-contact temperature-dependent resistance (R(T)) measurements
as shown in Figure 5.7, we determined the activation energies involved in OPTEG.
The Figure 5.7a shows a slightly non-metallic behaviour in OPTEG. To examine
this non-metallicity, we generate an Arrhenius plot (ln(R) vs. 1/T) presented in
Figure 5.7. (a) Resistance vs. temperature acquired from pristine EG and OPTEG.
(b) Arrhenius plots obtained from data in (a). (c) Data of ≤ 50 K plotted on
appropriate axis for WL-type behaviour.
Figure 5.7b. Evidently, this analysis (Figure 5.7b) does not result in a linear
relationship expected in an intrinsic semiconductor with well-defined energy gap.
Instead, the plot shows a variable-slope with a temperature-dependent activation
energy, which changes from 0.4 meV in the vicinity of room temperature to 0.018
(a) (b)
(c)
0 50 100 150 200 250 300
0.5
1.0
1.5
2.0
Pristine EG
OPTEG
T (K)
R (
k
)
0.0 0.1 0.2 0.3 0.4 0.55.70
5.72
5.74
5.76
5.78
5.807.05
7.20
7.35
7.50
Pristine EG
OPTEG
1/T (K-1)
ln(R
)0.08 meV
0.4 meV
0.001 meV
0.018 meV
0 1 2 3 4 5 6
1.50
1.55
1.60
1.65
1.70
1.75
1.80
1.85
1.90
R (
k
)
ln(T)
102
meV at helium temperatures, which is much less than room temperature thermal
energy (kBT ~ 26 meV, for T = 300 K). Furthermore, we also see almost similar
behaviour of R(T) plots in pristine EG and OPTEG. The plot R vs. ln(T) as shown
in Figure 5.7c, is almost linear and satisfies the R∝ ln(T) relation. This type of
logarithmic temperature dependence has been observed in a number of two
dimensional electron gas (2DEG) system179, 180
or disordered EG on C-face SiC as
discussed in in chapter 6 and, and has been theoretically, recognised as weak
localization181, 182
or electron-electron interaction183, 184
in 2DEG system with
disorder. As such, non-metallic behaviour (Figure 5.7c) originates from scattering
in a 2D electronic system with some degree of disorder. The possible mechanism
of photoconductivity in our device is described in section 5.5.5. However, optical
property investigations in OPTEG may further be explored to obtain more insights.
5.5.4. Raman measurements
Raman measurements in pristine EG and OPTEG were performed on a Raman
microscope (Renishaw) with excitation wavelength of 515 nm. All spectra were
recorded using a 50× microscope objective (laser spot size ~ 1 µm) to focus the
laser excitation onto the sample as well as collect the scattered light. The
measurements were performed at room temperature and ambient environment.
Figure 5.8 shows the Raman spectra obtained from pristine EG, and after treatment
for 20 s (OPTEG1) and 40 s (OPTEG2) with oxygen plasma. The main features in
these Raman spectra are the G-band (at ~ 1581 cm-1
), 2D-band (2710 cm-1
) and D-
band (at ~ 1360 cm-1
).106, 107
Compared to the pristine EG, a slight enhancement in
103
in D-band is observed after oxygen plasma exposures. This D-band enhancement is
caused by increased concentration of sp3 carbon centres in the conjugated sp
2
Figure 5.8. Raman characterization in pristine EG and OPTEG devices for two
different treatment times: 20 s (OPTEG1) and 40 s (OPTEG2).
network, and could be a result of chemical C-O and C=O bonds (Figure 5.5d) or
carbon vacancies.
5.5.5. Possible photoresponse mechanism
We attribute the device high-photoconductivity (high-gain) to the formation of
localised charge carrier traps in OPTEG. The temporal photoresponse
1200 1500 1800 2100 2400 2700 3000
OPTEG1 (40 s)
G
D
Ra
ma
n in
tes
ity (
a.u
.)
Wavenumber (cm-1)
2D
Pristine EG
OPTEG1 (20 s)
104
measurements show relatively slower response times (tr and td as shown in Figures
5.4e and 5.4g, compared to much faster (≤ nanosecond) responses observed in pri-
stine graphene based devices.76, 114, 115
This result suggests the trapping phenomena
Figure 5.9. (a) Upper panel shows the pristine n-type graphene having free
electrons in conduction band.The bottom panel shows the pristine EG device (b)
OPTEG showing a trap that have captured free electrons, and reduced the electron
density in conduction band. Bottom panel shows the scenario in the device. (c)
Under illumination, electron-hole pairs are generated. Migration of photogenerated
carriers, under the surface depletion electric field, results in a hole trapping or
neutralizing at negatively charged traps, leaving behind high density of unpaired
electrons that moves towards positive electrode and gives rise to a huge
photocurrent. Trapping or neutralization of holes hinder recombination of electron
hole pairs and hence prolong their lifetime.
in OPTEG. As such, a schematic for the possible mechanism is presented in Figure
5.9. Pristine EG is n-type, having free electrons in the conduction band (Figure
5.9a). In case of oxygen plasma-treated device (Figure 5.9b), free electrons are
localized or trapped at defect centres induced by plasma exposure. Formation of
105
traps can occur due to the local disruption of graphene electronic system via
creation of carbon vacancy or formation of C-O or C=O bonds. As such, free
electrons are localized in traps. Under illumination, electron-hole pairs are
generated, and migration of holes on the surface causes them to be trapped or
neutralized at negatively charged traps (called hole traps), leaving a huge number
of unpaired electrons that move towards the positive gold electrode (Figure 5.9c).
Hence, unlike the ultrafast charge carrier recombination in pristine graphene
device,71
carrier recombination is slowed down due to the presence of traps in the
OPTEG device, and it gives rise to a huge photocurrent via prolonging the life time
of the photogenerated charge carriers.
5.5.6. Light intensity-dependent photoresponse
To get more insight for the above mechanism, light intensity-dependent
photocurrent measurements were carried out. Figure 5.10 shows the photogenera-
Figure 5.10. (a) Light intensity F depenndent photocurrent Iph observation in a
OPTEG device. (b) A plot showing photocurrent following a power law
dependency with light intensity. The inset shows the plot in logarithmic scales.
106
-ted current Iph at 2 V under different light intensities. The temporal photocurrent
responses shown in Figures 5.10a, 5.10b show that the photocurrent grows with
light intensity and saturates at higher intensity. It is also noted from the shape of
the temporal profile in Figure 5.10a that response times are faster at higher
intensities. The photocurrent Iph and light intensity F can be expressed by a power
law, Iph∝ , where n is an exponent,185
and ~ 0.49 (Figure 5.10b) for a OPTEG
device. Such a power law dependency (sublinear increase in photocurrent with
light intensity) is typically observed when traps of distributed energy are present.185
Photocurrent saturation and relatively faster response time can be understood; if we
assume that traps get saturated (filled traps) at higher intensity, hence increasing
the recombination rate. These photoconductive characteristics observations are
consistent with previous studies of photoconductors175, 185, 186
containing defect-
traps, and hence supporting the mechanism given above.
107
5.5.7. Spectral photoresponse
In order to determine the electron to photon conversion efficiency
(photoconductive gain or photoresponsivity) and its wavelength dependency
(spectral photoresponse), a specific photoconductivity measurement set up was
used. The schematic of the set up is shown in Figure 5.11. It consists of a-
Figure 5.11. Spectral photoresponse measurement system.
monochromator (iHR320, HORIBA) capable of producing light with wavelengths
ranging from 300 nm to 1000 nm. The light intensity was calibrated with a high
sensitivity and spectrally flat pyroelectric detector (Newport, model 70362).
Spectral photoresponses from OPTEG1 and OPTEG2 devices were measured at
fixed 2 V bias. All the measurements were performed at room-temperature and
ambient environment.
108
Figure 5.12. (a) Photo of experimental set-up during spectral photoresponse
measurement of OPTEG device. (b) Zoom-in view near the device.
The photoconductive gain (also called external quantum efficiency or quantum
efficiency) or photoresponsivity of a photodetector device can be estimated using
the following equations:
Gain = (Iph/P)(hν/q) = [1240/λ (nm)](Iph/P) = [1240/λ (nm)](PR) [5.1]
where, Iph is the photocurrent, P is the effective light power incident on the active
area A (here 170 μm × 300 μm) of a photodetector device, hν is the energy of an
incident photon, q is the electron charge, λ is wavelength of light, PR is the
photoresponsivity.
109
Figure 5.13. Spectral (300 nm-1000 nm) photoresponse observations in (a), (b)
OPTEG1 device and (c), (d) OPTEG2 device.
The spectral photoresponse of OPTEG1 and OPTEG2 photodetector devices
are shown in Figure 5.13. We see that OPTEG1 (treated with 20 s) device shows a
gain of 10000 in UV region (300 nm), 380 in visible blue light (470 nm) and 1.8 in
the IR region (1000 nm). Furthermore, the device (OPTEG2) treated with relatively
longer (40 s) oxygen plasma duration shows an enhancement in UV-detection-gain
by a factor of ~ 2.4. The results are summarised in Table 5.1. The results suggest
that the oxygen plasma treated device is most sensitive to UV light. An OPTEG
300 400 500 600 700 800 900 1000
0
2000
4000
6000
8000
10000
12000
Gain
Wavelength (nm)
OPTEG1
400 600 800 1000
0
100
200
300
400
500
Ga
in
Wavelength (nm)
200 400 600 800 1000
0
500
1000
1500
2000
2500
Ph
oto
res
po
ns
ivit
y (
A W
-1)
Wavelength ( nm)
OPTEG1
300 400 500 600 700 800 900 1000
0
5000
10000
15000
20000
25000
Gain
Wavelength (nm)
OPTEG2
200 400 600 800 1000
0
1000
2000
3000
4000
5000
6000
Ph
oto
res
po
nsiv
ity (
A W
-1)
Wavelength (nm)
OPTEG2
(d) (c)
(a) (b)
110
device could further be improved via reducing the length-scale of the device, using
asymmetric metallization scheme, or coupling the pristine graphene as transport
channel to produce fast and ultra high-gain UV photodetector. High mobility
pristine graphene coupled115
with active OPTEG should further accelerate the
carrier transport in devices. The gain (or photoresponsivity) of our OPTEG devices
are ~ 4-5 order of magnitude higher than previously reported photodetectors based
on pristine graphene,76
RGO materials,136
GO/TiO2 hybrid materials,68
and many
other organic/inorganic semiconducting materials.158, 187-195
A comparative
summary is given in Table 5.1.
An important figure of merit of a photodetector is detectivity (or sensitivity),
which refers to the minimum impinging optical signal, that is distinguishable above
noise. It is denoted as D*, and is measured in unit of Jones (cm Hz1/2
W-1
).
If noise from the dark current is the dominant contribution, the detectivity can be
expressed as
D* = (A1/2
PR)/(2qId)1/2
[5.2]
where PR is photoresponsivity, Id is the dark current, q is the electron charge.
Under illumination at 300 nm (UV light, Figure 5.13), the calculated detectivities
of OPTEG1 and OPTEG2 devices (with fixed 2 V bias) are ~ 1.3 × 1012
Jones and
3.2 × 1012
Jones respectively. Detectivity of our OPTEG device is much higher
compared to the detectivity of previously reported photodetectors based on organic
(5TmDCA)/inorganic(ZnO)193
hybrid system (1010
Jones), many other
organic/inorganic,68, 158, 196
and comparable to a commercial InGaAs (1012
Jones)197
photodetector or high performance AlGaN (3 × 1012
Jones) UV detector.198
111
Table 5.1.Comparison of OPTEG photodetectors with previously reported
photodetectors.
Active
materials
Gain Photoresponsivity
(A W-1
)
Detectivity
(Jones)
Bias
voltage
(V)
Referenc
es
OPTEG1
(20 s)
104
(300 nm)
3.80 × 102
(470 nm )
1.8
(1000 nm)
2.420 × 103
(300 nm)
1.44× 102
( 470 nm)
1.5
(1000 nm)
1.3 × 1012
(300nm)
2 This work
OPTEG2
(40 s)
2.23 × 104
(300 nm)
3.92 × 102
(470 nm)
0
(1000 nm)
5.397× 103
(300 nm)
1.48×102
( 470 nm)
0
(1000 nm)
3.2×1012
(300nm)
2 This work
Pristine
graphene
(Exfoliated)
4.88 × 10-3
(1550 nm)
6.1 × 10-3 < 1 76
RGO 0.40
( 360 nm)
0.12
( 360 nm)
1 136
RGO 3× 10-3
(1550 nm)
4× 10-3
(1550 nm)
2 137
Graphene
nanoribbon
(GNR)
0.8
(1550 nm)
1.0
(1550 nm)
2 137
RGO-TiO2
Hybrid
0.75
(300 nm)
0.18 (300 nm) 6.82×1011
(300 nm)
5 68
CNT/organics
emiconductor/
C60
2.3 × 10-
2(1155 nm)
2.1 × 10-2
(1155 nm)
1010
(1155 nm)
< 1 158
TiO2 thin film 9.49 × 102 1.99 × 102
(260 nm)
5 188
112
Active
materials
Gain Photoresponsivity
(A W-1
)
Detectivity
(Jones)
Bias
voltage
(V)
Referenc
es
ZnO thin film 1.6 × 103
( 300 nm)
4.0 × 102
(300 nm)
5 189
ZnO/NiO: Li
heterojuction
1.0
(360 nm)
0.3
(360 nm)
6V 190
n-ZnO/p-Si
heterojuction
2.0
(360 nm)
0.5
(310 nm)
30 V 191
AlGaN/GaN
heterostructure
7.0 × 104
(312 nm)
7.0 × 104
(312 nm)
100 V 192
5TmDCA/Zn
O hybrid
1.2 × 102
(500 nm)
0.48
(500 nm)
2.0 × 1010
(500 nm)
2 V 193
Commercial
SiC
0.53
(300 nm)
0.13
(300 nm)
20 V 194
Commerial
GaP
0.41
(300 nm)
0.1
(300 nm)
5 V 195
GaN 0.15
(360 nm)
7.245 × 109
(360 nm)
2 V 196
AlGaN 1.2 × 10-2
(280 nm)
1.2 × 10-2
(280 nm)
3 × 1012 (280
nm)
4 V 198
CNT/GaAs
Schottky
0.31
(633 nm)
0.16
(633 nm)
10 187
113
5.6. Applications
From the above Table 5.1 we conclude that photodetector devices in this work
of study are promising. Potential applications of our OPTEG photodetector include
UV detection for
i. astronomy: intra- and intersatellite-secured communications.
ii. biological and chemical sensors: ozone determination, determination of
pollution levels in air, biological agent detection, etc.
iii. flame sensors (fire alarm systems, missile-plume detection, combustion
engine control, UV imaging, including solar cell measurements etc.
Importantly, large photoconductive gain, detectivity and high selectivity in the
UV spectral range observed in our OPTEG photodetector would have an advantage
over the less spectrally selective Si UV detector. Wide band gap semiconductors
such as GaN, AlGaN, diamond, SiC, ZnO are some of the candidates that show UV
detection in devices. Comparing to these materials (Table 5.1), the OPTEG UV
photodetector would be advantageous in many aspects.
The UV photodetector for hyperspectroscopy199
is another important
application. Hyperspectroscopy is the new method of surface image aquisition. It
provides high position and spectral resolution that allows one to determine surface
chemical composition. The key elements in hyperspectroscopy are imaging
photodetectors placed in the focal plane of the specially designed spectrographs.
The main requirements to these detectors are: high sensitivity, high position
resolutions (below 100 μm), high signal to noise ratio, low power consumption.
Our OPTEG photodetector could fulfill most of the requirements for this
application.
114
5.7. Conclusion
In summary, this chapter describes the mild oxygen plasma treatment in EG
based devices to produce high-gain photodetectors. OPTEG surface is investigated
using Raman spectroscopy, synchrotron-based high resolution PES, and electrical
I-V measurements. Photoconductive characteristics of OPTEG devices are studied
in detail under illumination of white light, and the spectral response from IR (1000
nm) to UV (300 nm) range of light is examined. The spectral photoresponse
measurement of a OPTEG device ( treated for 20 s with oxygen plasma) exhibits
photoconductive gains of ~ 104 (or photoresponsivity ~ 2.5 ×10
3 A W
-1), 3.80 ×10
2
and 1.8 with UV (300 nm), visible (470 nm, blue light) and IR (1000 nm) lights
respectively. The device treated for a relative longer time (40 s, OPTEG2) shows
enhanced UV photoconductive gain of ~ 2.23×104. The carrier trapping
mechanism, which gives rise to a huge photoconductivity in the device, is
discussed and a mechanism based on the formation of localised charge carrier traps
is proposed. The high selectivity, large photoconductive gain, and detectivity in
range of 1.3 × 1012
to 3.2 × 1012
Jones for the OPTEG devices in the UV spectral
range highlight their potential advantages over Si and many other semiconductor
UV photodetectors. The method’s compatibility with CMOS process flow makes it
an efficient technique to fabricate graphene based high-gain photodetectors, as well
as potential applications in UV photodetection, imaging etc.
115
CHAPTER 6
Magnetotransport Studies of Multilayer Epitaxial
Graphene on C-face SiC
6.1. Introduction
The single crystalline 4 H-SiC material comprises two polar faces: Si-face
(epilayer in 0001direction) and C-face (epilayer in [000 ̅ direction (Figure 6.1).
Figure 6.1. Crystal structure of SiC showing C-face (000ī) and Si-face (0001).
Besides the preparation of graphene on Si-face, it is also possible to grow
graphene on C-face SiC substrates.28, 79
The formation process of graphene, its
C
Si
(000ī) C-face
(0001) Si-face
116
structure and properties on C-face SiC is different from that on the Si-face SiC
substrate. In contrast to the growth of graphene on the Si-face, graphene can be
grown at relatively lower temperatures on the C-face SiC. The faster growth (or
graphitization) on the C-face, often results in the formation of multilayered
graphene.
The structure, surface reconstruction, and electronic properties of EG on C-face
SiC have been investigated by several groups.28, 200
According to the reports, the
LEED pattern shows different surface reconstructions with rotationally disordered
diffraction rings in EG on C-face SiC. The initial graphene layer develops in
coexistence with the intrinsic surface reconstructions without the presence of an
interface layer or buffer layer, in contrast to the presence of a buffer layer
(identified as 0303636 R reconstruction)123
in EG on Si-face SiC. Previous
STM measurements show the graphene structure directly on top of the (3×3)
reconstruction with a Moir’e type modulation with a large superperiodicity. The
rotational disorder and the absence of an interface layer indicate a rather weak
coupling between graphene and the C-face SiC substrate.201, 202
The difference in growth mode in case of C-face SiC also results in quite
different electronic properties compared to graphene grown on Si-face SiC. It has
been found that, unlike graphene on Si-face SiC or HOPG which adopts the AB or
Bernal stacking,203
graphene on C-face SiC has unusual rotational fault stacking.28
These rotated planes are interleaved in the film and not part of isolated Bernal
stacked films with multiple rotated domains like those in HOPG. This unique
stacking leads to a symmetry change in adjacent non-Bernal layers that makes
these graphene sheets act electronically like a stack of isolated graphene sheets.204
117
Because these non-Bernal stacked layers make up the majority of the film, nearly
the entire graphene film (multilayer on C-face SiC) behaves electronically like
single layer graphene.205, 206
Previous studies in graphite (multilayer graphene film)
show very high MR (>1000% at 1T)207
unlike almost negligible MR in very thin
single/few layer graphene.6, 208
The reason is found to be large lateral size of the
graphite sample. In addition, multilayer EG on C-face SiC possesses rotational
stacking faults (graphene layers rotated by 30o or ± 2
o with respect to SiC substrate
(typically described as R30/R±2
fault pairs).28
With the rotational fault parabolic
dispersion is recovered near the K point. As such, the transport properties of C-
face EG have received much interest from graphene researchers.79, 116
Multilayer
layer graphene film on C- face SiC is of interest for magnetotrasport properties
where high magnetoresistance or long phase coherence length would be additional
advantages. Compared with SiO2 supported graphene, multilayer EG on C-face SiC
screens (layer closer to the substrate can protect the subsequent top layer from the
substract induced scattering) the substrate induced scattering leading to reduced
spin relaxation.27
Highly efficient (~75%) spin transport was recently reported by
Bruno et al.27
It has been observed that EG on C-face SiC has higher charge carrier
mobility compared to that of graphene on Si-face SiC and this mobility correlates
with the sharper 2D bands in Raman measurements.116
Depending on the
homogeneity or quality of multilayer C-face EG, the 2D-band in the Raman
spectrum shows a single Lorentzian shape, similar to the observation in single layer
graphene. Electronically decoupled multilayer films on C-face SiC, unlike that on
the Si-face SiC with strong substrate-film interaction, shows superior carrier
transport in the material.79
118
Magnetoresistace and weak localization
Magnetoresistance (MR) studies are important for the development of magneto-
electronic or spintronic devices. MR is the property of a material whereby a change
in its electrical resistance occurs when an external magnetic field is applied to it
and this effect was first discovered by William Thomson (more commonly known
as Lord Kelvin) in 1856.
A number of MR studies can be found in high mobility semiconductor materials
or 2DEG (two dimensional electron gas) systems.209, 210
In particular, phenomenon
of weak (-anti) localization, which often evolves as negative MR at low
temperatures (< 50 K) in the vicinity of zero magnetic field, is the fundamental
characteristic feature observed in a number of disordered semiconducting
materials or 2DEG systems.181, 211
In disordered or 2DEG electronic systems,
constructive quantum interference between elastically backscattered partial carrier
waves is attributed to the origin of weak localization. Weak localization is a low
temperature effect because inelastic scattering processes are due to electron-phonon
or electron-electron interactions that can be masked by thermal effects. This
coherence of the back scattered waves tends to be suppressed when a magnetic
field is applied.
Fundamental theoretical and experimental studies of MR in graphene have been
reported by a few research groups.182
Experimental MR observations in single or
few layer graphene show the presence of weak (-anti) localization in the
material.212
119
Linear MR
It has been found that the resistance of a conductor213
in an applied magnetic
field increases quadratically (MR∝B2) with field and then saturates at a relatively
low value; this behaviour shows its inability to detect the polarity of a magnetic
field. Therefore, linear MR (LMR) in a device is of great interest for magnetic
sensing applications. Numerous studies involving the fabrication of geometrical
devices,214
nanoscale manipulation of structure215
or introduction of
defects/impurities,216, 217
have been found to produce LMR.
Room temperature operation is needed for practical device applications in the
semiconductor industry. It is important to understand the fundamental physics
along with the exploration of respective applications. Unlike single layer
graphene,148
multilayer EG would be more interesting and favourable in many
aspects114, 115
for practical uses. Multilayer graphene structures in a device are more
robust, the interior layers are protected from the environment, and the layered
structure allows intercalation.
However, many technical challenges remain. One challenge is the production of
large area, high quality multilayer graphene films on the C-face SiC. Another
challenge is the careful fundamental transport studies in the material, underscoring
the potential of exploiting EG on C-face SiC nanomaterials for room temperature
magneto-electronic applications. The presence of large LMR and its sustainability
at room temperature has potential for many magneto-electronic device applications
such as magnetic sensors.215, 218
In this chapter, I explore the MR properties in multilayer EG on C-face SiC
substrate from room temperature to temperatures below 2 K in a large magnetic
120
field B range from 9 T to –9 T. The finding shows LMR, that is positive, large, and
stable at room temperature. We observe direct experimental evidence of two
dimensional magnetotransport effect, that results in LMR in EG. In graphite, such
linear MR has been observed only at low temperatures,219
suggesting EG is a
distinctively different material. Furthermore, the evolution of negative MR in the
low field regime at temperatures ≤ 20 K, is also found, and the negative MR is
experimentally as well as theoretically recognised as weak localization, which is
unique in graphene, rather than in graphite. The room temperature stable and
relatively high LMR, unambiguously highlights the superiority of multilayer EG
over single layer graphene220
(MR < 1%) and several organic or inorganic
semiconductor materials.221-223
This study demonstrates the potential of exploiting
the material for further development of graphene based magneto-electronics.
6.2. Material synthesis
EG was synthesised on C-face semi insulating 4H-SiC substrates (substrate
dimension 2 mm ×10 mm, purchased from CREE company). The following steps
were carried out in synthesis process:
(i) The SiC substrate was cleaned using isopropanol (substrate was put in a beaker
having isopropanol solution and then sonicated for 20 minutes), acetone (the same
process was repeated with acetone solution) and HF ( a dilute HF(10%).
(ii) The cleaned substrate was mounted in a specific sample holder to facilitate the
heating of the sample at desired high temperatures.
(iii) The mounted sample in sample holder was introduced in the load-lock, and
load in load lock before transfer to main chamber for annealing.
121
(iv) The sample holder was transferred from load-lock to main UHV (pressure
≤10-10
mbar) chamber and inserted into the manipulator for heating. The resistance
of the sample was checked with a multimeter before applying voltage and current
(using power supply) through the manipulator to the sample holder.
(v) The current was applied slowly in order to maintain a constant heating rate ~ 25
°C/min. The temperature was monitored using an optical pyrometer.When the
temperature reaches ~ 900 °C, the sample was exposed to a silicon flux (from a a
silicon source within the UHV chamber) for 10 minutes for further cleaning.
After10 minutes holding time (after silicon flux is stopped), annealing continued
with the same constant heating rate up to ~ 1400 °C and sample was held at 1400
°C for 40 minutes. Subsequently, the sample was cooled down slowly at the same
rate (25 °C/min) to room temperature. After half an hour at room temperature, the
sample was taken out from chamber for further characterization using Raman and
AFM measurements to confirm the formation of graphene on the SiC surface.
6.3. Device fabrication and characterization
After confirming the formation of graphene film on SiC surface (using Raman
signatures), several four (Cr(10)/Au(100 nm metal contacts))-electrodes devices
were fabricated.
6.3.1. Raman characterization
Raman spectroscopy measurements were performed several times to identify the
growth of graphene on C- face SiC and film quality over surface (> 20 points over
2mm ×2 mm area) of the substrate. All the measurements were carried out on a-
122
Figure 6.2. Raman spectroscopy investigations in mulilayer EG on C-face SiC. (a)
A typical Raman spectrum obtained by measurement within the tested device area
(shown by white rectangle in inset). (b) 2D-band fitted with single Lorenzian
function (the dots represent data, solid blue line is fitting) with width (FWHM) ~
42 cm-1
. All the 2D-bands obtained from several locations (> 20 points) within the
device area were found to be of single Lorentzian nature, suggesting electronically
decoupled graphene layers on C-face SiC. (c) A typical single Lorenzian fitted 2D-
band of single layer EG on Si-face SiC, showing identical width ~ 42 cm-1
Raman microscope (Renishaw) with 514 nm excitation wavelength at room
temperature and ambient environment.
1200 1500 1800 2100 2400 2700 3000
SiC
2D
D
Ra
ma
n i
nte
ns
ity
(a
.u.)
Wavenumber (cm-1)
G
(a)
(c) (b)
2590 2660 2730 2800 2870
Wavenumber (cm-1)
Multilayer EG on C-face SiC
FWHM ~ 42 cm-1
Ra
ma
n i
nte
ns
ity
(a
.u.)
123
A typical Raman spectrum (measured within the area highlighted by the white
rectangle in inset) is shown in Figure 6.2a. The spectrum in Figure 6.2a shows a G
(at ~ 1590 cm-1
), 2D (at ~ 2716 cm-1) and less intense D (at ~ 1360 cm
-1) bands,
which are characteristic peaks expected in carbon based materials.106, 107
In the case
of graphene, the 2D band width has strong correlation with the number of layers or
quality of the film.116
The D-band usually indicates the presence of defect/disorder
in graphene. We find that the 2D-band is well fitted with a single Lorentzian
(Figure 6.2b), with a sharper (FWHM) ~ 42 cm-1
) width. Interestingly, the width is
found to be identical to that of single layer graphene on Si-face SiC, prepared
previously in our lab (Figure 6.2c). The less intense D-band and sharper 2D-band
are indicators of good quality EG used in this study.
6.3.2. Surface morphology and thickness determination using AFM
The surface morphology (Figure 6.3a) and film thickness (Figures 6.3b, 6.3c)
were examined using AFM measurements (performed with a Veeco D3000 NS49
system). All the measurements were performed in tapping mode at room
temperature and ambient condition. Figure 6.3a shows the AFM image of EG film
on C- face SiC. The graphene over SiC substrate has a measured rms roughness of
~ 2.5 nm, which is much reduced value compared to a previous report224
on
multilayer EG. With uniform heating to ≥ 1400 ºC, a multilayer film with reduced
roughness can be produced over large area on SiC substrate. Furthermore, the 2D
band in spectrum cannot be applied to determine the thickness of multilayer EG
film on C-face SiC as we find sharpening of the 2D band in the multilayer film-
124
Figure 6.3. (a) AFM image of EG film on C-face SiC. (b) Line profile taken across
white dash line showing the scale in Z-direction. (c) A typical AFM image
showing the EG film and SiC (etched part of EG) region. (d) A line profile taken
acrossthe white dash line indicated in (c). The average thickness of the EG film
was found to be ~ 6 nm±1 nm, that corresponds to ~ 20±3 layers in EG.
(unlike EG on Si-face SiC). To perform an AFM height profile measurement, parts
of the EG film were etched using oxygen plasma. Figure 6.3b shows a typical AFM
image exhibiting the EG and etched region (SiC). The height profile measurements
in such process shows the average film thickness of ~ 6 nm±1 nm corresponding to
~ 20±3 EG layers.
0 200 400 600 800 1000
0
2
4
6
8
10
Z(n
m)
X (nm)
(a)
(c)
0.0 0.2 0.4 0.6 0.8 1.0-8
-6
-4
-2
0
2
4
6
8
Z (
nm
)
X (nm)
(d)
(b)
125
6.4. Magnetotransport measurements
Figure 6.4. V+, V-, I+, I- are four-terminal voltage and current labels used for
measurement of MR in EG.
The MR measurements in EG sample were performed using a PPMS system
(PPMS, model: 6000, Quantum Design). A constant AC current of 1µA was
applied through the sample ( from I+ to I-, Figure 6.4) and voltage (or resistance)
was measured between V+, and V- in a perpendicular magnetic field B with
magnitude varying from -9 to 9 T. The MR was estimated with the equation:
MR (%) = ((R(B)-R(B=0))/R(B= 0))×100. [6.1]
6.5. Results and Discussion
6.5.1. Room temperature large linear magnetoresistance
Figure 6.5a shows the MR (%) from 300 K (room temperature) to 2 K in a
variable magnetic field from -9 to 9 T. We can see a LMR with its magnitude ~
50% at 300 K and at 9 T. This room temperature LMR observed in EG is much
higher than several previously reported MR or LMR in single layer graphene,208
carbon nanotube,225, 226
organic materials221, 222
or many other inorganic
semiconducting materials223, 227
and comparable to recently reported MR
magnitude (50%) in graphene nanoribbon.228
Graphene
Cr/Au electrode
126
Figure 6.5. (a) MR (in %) at different temperatures T from 300 K to 2 K in
magnetic field B range of -9 T to 9 T. (b) The separated view of (a) from 300 K-50
K, exhibiting LMR sustained at room temperature (300 K) to below 50 K, with no
evidence of negative MR. The separated view of (a) from 20 K to 2 K, showing the
evolution of negative MR (WL peak around B = 0 T).
Table 6.1 shows some previously reported room temperature LMR (or MR) values.
Furthermore, by lowering the temperature from 300 K to 2 K, a slight decrease in
MR is found, while maintaining the LMR from 300 K to 50 K (Figures 6.5a, 6.5b).
Specifically, for temperatures ≤ 20 K, evolution of negative MR (the MR peak
around B = 0, Figures 6.5a, 6.5c) is observed. This negative MR is found only at
very low field (<0.5 T) regime in the vicinity of zero magnetic field, and in
temperature range 20 K to 2 K.
127
Table 6.1. Room temperature MR/LMR comparison
Materials LMR/MR
(%)
References
EG on C-face SiC 50 This work
Single layer graphene < 1 208
Carbon nanotube < 3 225, 226
V(TCNE)2 ~ 2 221, 222
MnAs-GaAs
composite
< 14 223
Graphene nanoribbon -50 228
6.5.2. Evidence for two dimensional (2D) magnetotransport
To obtain more insight on the observed LMR results, we performed the angular
dependent magnetoresistance measurements in EG. The direction of B was changed
by rotating the sample (along the current direction) such that the direction of B
changes with respect to both sample plane and current flow direction (Figure 6.6a).
Three dimensional (3D) transport involves conduction from bulk carriers which
can occur in thicker or bulk 3D materials. In 3D materials, interlayer tunneling of
charge carriers (movement of carrier in z-direction) can occur. To test the transport,
we examine the oscillation of resistance by varying B direction. For 2D transport,
the perpendicular component of the magnetic field should follow the relation B⊥=
Bcos(θ),229
and any deviation from this indicates the contribution due to 3D
transport. For example, existence of in-plane (B direction parallel to the sample-
128
Figure 6.6. Angular dependent magnetotransport in EG. (a) A schematic (b)
Resistance R vs. the tilt angle θ in a fixed magnetic field of 9 T and at T= 300 K
and 2 K. The solid lines (blue, at 300 K and red, at 2 K) are fitting of data with the
function |cos(θ)|.
sample surface (θ = 90°, Figure 6.6a) or along the direction of current flow) MR is
the indicative of contribution from 3D transport. Figure 6.6b shows the resistance
129
of the sample as a function of rotation angle when the sample was rotated (from 0°
to 360°) in a fixed 9 T magnetic field, and at T=300 K and 2K respectively. We see
that the R(θ) has a maxima at 0° and minima at 90°. The experimental results
(Figure 6.6b) were found to fit with the functional form |cosθ| (solid blue and red
lines at 300 K and 2 K respectively in Figure 6.6b), indicating a 2D
magnetotransport in EG.
Several mechanisms have been proposed to characterize the phenomenon of
LMR from classical and quantum perspectives. The first model to be applied to
LMR was quantum description by Abrikosov,230
which examines a small bandgap
(or zero-gap like graphene) material subjected to a magnetic field strong enough to
reach the quantum limit where all the electrons occupy the lowest Landau level,
and called quantum LMR. Quantum LMR also considers linear energy dispersion
relation (effective mass m* = 0) as in graphene. As such, it indicates that the
transverse MR has a quadratic field dependence in the classical low-field case
where ħωc kT but a linear dependence in the quantum limit.230, 231
The quantum
regime usually happens at the strong-field region where the temperature condition
is ħ(eB/m* c) k T.
230 For our graphene samples, the effective mass m
* = 0 so the
left hand side of inequality diverges making this condition of LMR obtainable at
room temperature and higher and this supports our experimental observation of
LMR obtained at room temperature. As shown in Figure 6.5b, MR curve exhibits
highly linear field dependence, with the linear MR starts from a small field (≥ 1 T).
For a classical quadratic dependency of MR,213
the data of low field (≤ 1 T) MR (at
300 K) vs. B2
(Figure 6.7) is plotted. The slope of a linear fit to this data results in
mobility ~ 1862 cm2V
-1s
-1, which is consistent with the mobility obtained using
130
Hall measurement. Moreover, the other criterion (
)
should also be
fulfilled for the usual quantum LMR as observed in high mobility InSb,209, 232
Ge233
and graphite234
and also recently in the topological insulator Bi2Se3.235
In our case,
the LMR starts from a very small field of ~ 1 T, which corresponds to a critical
carrier density of 5.95 ×1016
cm-3
for the quantum LMR. The carrier density of our
samples at 300 K is ~3.0 ×1013
cm-2
which is smaller than the critical carrier
density.
Figure 6.7. MR as a function of magnetic field B2.
Alternatively, a classical model has been developed by Parish and Littlewood.236
The model is applied to a strongly inhomogeneous conductor (unlike Abrikosov
model which only deals with weak disorder). The simplest model assumes a 2D
square lattice constructed of four-terminal resistors, with an external magnetic field
applied perpendicular to the network. The details can be found in Ref. 236. In this
classical model the linear dependence of the magnetoresistance is linked in a
0.0 0.2 0.4 0.6 0.8 1.00.000
0.005
0.010
0.015
0.020
0.025
0.030
0.035
MR
B2 (T2)
data
Linear fit
131
simple manner to electrical disorder. A strong magnetic field forces a significant
portion of the current to flow in a direction perpendicular to the applied voltage and
thereby provides a linear-in-B Hall resistance contribution to the effective MR. The
classical LMR finds a strong temperature dependency which follows the power
law. Johnson et al223
demonstrated LMR in strongly disordered MnAs-GaAs
composite semiconductors, with LMR strongly temperature dependent and
following a T1.85
power law dependency, which is attributed to the prediction of
classical LMR.236
In the case of EG on C-face SiC, we find that the LMR is temperature
independent. To test this behaviour, we have analysed the plot dMR/dB vs. T (log-
Figure 6.8. (a-g) Linear fits to LMR (in 1 T≤B≤9 T range) data at different
temperatures 300 K, 100 K, 50 K, 20 K, 10 K, 5 K and 2K. These plots give the
slopes (dMR/dB(%)) at different temperatures.
132
1 3 7 20 55 148 4031
3
7
20
dM
R/d
B (
%/T
)
T (K)
Figure 6.9. Slope of LMR (dMR/dB) vs. temperature T (logarithmic scales). It is
evident that the curve does not follow a linear relationship, it is invariant with T,
which rules out a power law dependency ((dMR/dB)∝T n
), which is expected for a
classical LMR model.
lot) as shown in Figure 6.9, by plotting the temperature dependence of the slope of
the linear portion of MR response over the range 1 T≤B≤9 T (Figure 6.8). We find
that this plot is not linear (which is expected for classical LMR behaviour), rather it
is invariant with T. Hence, this proves that LMR observed in our EG is temperature
independent.
In general, we conclude that LMR exists over the whole temperature range from
300 to 2 K, in the magnetic field range 1 ≤ B ≤ 9 T, suggesting that the observed
LMR in our study is a 2D magnetotransport effect. Moreover, the observed LMR
could be specific to the intrinsic properties (such as small disorder in material) in
EG on C-face SiC. Here we understand that for LMR behaviour, a minimum
disorder could play an important role to change conduction path of carriers from
quadratic to linear conduction behaviour. The low temperature LMR behaviour has
133
also been observed previously in several other materials such as InSb,209
topological insulator Bi2Se3235
and graphite.234
The room temperature LMR in EG
on C-face SiC as we observe here therefore has a distinct advantage over a graphite
material.
6.5.3. Study of weak localization
The observed negative MR from 20 K to 2 K were further investigated by
angular dependent magnetotransport experiments and the weak localization
model.182
Figure 6.10. WL MR investigations in EG. (a) Magnetic field anisotropy effect on
WL MR, showing a little but non-vanishing negative MR when B is parallel (90°,
-3 -2 -1 0 1 2 3
230
235
240
245
Re
sis
tan
ce
R (
)
Magnetic field B (T)
00 (Out of plane)
300
600
900 (In-plane)
2 K, Tilt angle
0.0 0.1 0.2 0.3 0.4 0.5-2
-1
0
1
2
3
4
5
6
7
2 K
5 K
10 K
(
S)
Magnetic field B (T)
20 K
0 5 10 15 20
0.5
L(
m)
T (K)
(a)
(b) (c)
134
In-plane) to sample surface and current direction. (b) Magnetoconductance from 20
K to 2 K in low field regime are found to fit with WL model182
(dots are data while
solid lines are fits to the data using WL model, a scaling along y axis by a factor of
0.10 was adopted in order to fit the data). (c) Estimated Lϕ(phase-coherence length)
vs. T.
Figure 6.10a shows the resistance R of the EG sample as function of magnetic field
B at temperature T=2 K, with B direction θ varying from 0° (perpendicular to the
sample surface) to 90° (B direction parallel to sample surface). We can see the
magnetic field anisotropic effect on negative MR (in low field <0.5 T regime), with
its residual small but non-vanished value even at 90° (B direction parallel to the
sample surface and direction of current). Such in-plane non-vanished negative MR
infers WL MR.237
The WL MR results were further analysed using WL theory
given by MacCann et al.182
WL arises from the constructive interference between
time-reversed multiple scattering trajectories of phase coherent carriers, leading to
coherent back-scattering of carriers and increasing the electrical resistance. When a
perpendicular magnetic field is introduced to break time-reversal symmetry or the
temperature is raised to destroy phase coherence, this interference is suppressed.
The model gives the correction in magnetoconductivity as182
( )
[ (
) (
) (
)] [6.2]
with ( ) ( ) (
) and
where ( ) [ ( ) ( ) [ ( ) ( )
If a scaling factor α is introduced, the phase coherence length Lϕ is given as
135
(
)
[6.3]
Further experiments were carried out (in order to get sufficient data points for
fitting) to obtain the negative MR from 2 K to higher temperatures until the MR
peak (around B = 0) disappears. Ψ is the digamma function, and ( ) is the
magnetoconductivity at sufficiently high temperature (approximated using data at
T= 30 K) for the WL features to disappear. and represent the intervalley and
intravalley (elastic) scattering lengths respectively. The equation [6.2] is well fitted
to the MR data at 2 K, 5 K, 10 K and 20 K with a scaling factor α = 0.10 as shown
in Figure 6.8b. Subsequently, the phase-coherence length (inelastic scattering
length) (Figure 6.10c) can be estimated (using equation [6.3]) and it is temperature
dependent, which can be approximated (from Figure 6.8c) as Lϕ∝ √ . This
would result in a phase coherence time proportional to inverse temperature, which
in 2D indicates that dominant phase-breaking mechanism is e-e scattering and it
supporting to the previous reports in graphene.238
The model has given a phase
coherence length of ~ 515 nm at 2 K (Figure 6.10c), that is ~ 1.72 times higher than
the reported phase coherence length (300 nm) in EG on Si-face SiC.79
This WL
behaviour distinguishes EG from graphite material, and also sheds light on its
potential utilization in applications such as determination of interface roughness,239
and in studying various scattering phenomena240
in solids such Kondo effect, single
electron tunnelling effect etc.
136
6.5.4. Resistance (R)-temperature (T) characteristic
The electron transport mechanism of materials is often revealed by temperature
dependence of the resistance.241
In Figure 6.11, we present R(T) data in the absence
a magnetic field B from room temperature (300 K) down to 2 K. It is evident that
the resistance does not change much with temperature, inferring a nearly metal-like
behaviour in EG. To get more insight near low temperatures regimes, we enlarged
the data for temperatures T≤ 20 K as shown in left inset of Figure 6.11. A larger-
Figure 6.11. R vs. T at zero B. The left inset is the enlarged data in low
temperatures (≤20 K) regime. The right inset showing data of left inset plotted in
appropriate axes for WL-type behaviour.
continuous area of graphene can allow a longer electronic mean free path, the
product of which with the Fermi wavevector now exceeds unity. The increased
0 50 100 150 200 250 300
500
1000
1500
Resis
tan
ce R
(
)
Temperature T (K)
2 4 6 8 10121416182022
175
180
185
190
195
200
R (
)
T (K)1.0 1.5 2.0 2.5 3.0
180
190
200
R (
)
ln(T)
R∝ln(T)
137
structural coherence results in less frequent inelastic scattering and elastic
collisions dominate, increasing the possibility of electrons tracing a closed path
while remaining phase-coherent and leading to weak localization.242
Therefore, a WL-type conduction can be described by following equation
R= R0+ Pln(T) [6.4]
where R0 and P are constants.
To investigate this behaviour in EG, the data shown in the left inset of Figure 6.11
is plotted on appropriate axes as shown in right inset of Figure 6.11. The plot
shows the linear dependency, inferring R∝ln(T). Therefore, R(T) data exhibits a
WL-type conduction behaviour well below 20 K, which is consistent with our
observations of weak localization in magentoresistance measurements described
above in section 6.5.3. This type of logarithmic temperature dependence was
observed in a number of two dimensional electron gas (2DEG) systems first at low
temperatures <1 K in thin metal (AuPd) films180
and silicon inversion layers179
and
later up to temperatures of 100 K in ultrathin Pt films243
and InO films.244
Such
WL-type electrical conductivity originates from scattering in a 2D electronic
system with some degree of disorder.211211, 245
6.6. Conclusion
In summary, this chapter describes the preparation of EG on semi insulating C-
face SiC substrate, characterization of EG film using Raman and AFM and the
detailed magnetotransport properties in EG with four contact devices. We have
found a large (~ 50 %) and linear magnetoresistance in multilayer (~ 20 layers) EG,
138
which is stable at room temperature, distinguishing EG from other carbon
materials. Through the angular dependent MR measurements, it is shown that the
MR follows a |cosθ| (vertical component of magnetic field) functional form, which
is an indicative of 2D magnetotransport that displayes LMR in EG. We also
observe negative MR behaviour in the low field regime for temperatures ≤ 20 K,
recognising this as weak localization. The theoretical model of weak localization is
applied to our multilayer graphene, resulting in a phase coherence length of ~ 515
nm at 2 K. This chapter provides an understanding of magnetotransport in EG,
which is promising in exploring the potential of EG materials as electronically
decoupled sheets of graphene. The EG material, in this study, would have
advantages over single layer graphene and several other organic or inorganic
semiconducting materials for room temperature magneto-electronic device
applications.
139
CHAPTER 7
Conclusion and Future Work
This thesis comprises of my work on the preparation and characterization of
epitaxial graphene (EG), and subsequently exploitation of EG or modified EG in
devices. The preparation of EG was performed via thermal annealing of the SiC
substrates in our UHV chamber. The as-prepared ultrathin few layer (1-2 layer and
4-6 layer) EG are carefully investigated using STM, AFM, Raman and Hall
measurements. Hall measurements of EG (~two layer) exhibiting room temperature
Hall mobility up to 1430 cm2V
-1s
-1, measured from large (150 µm x 370 µm) Hall
bars, and up to 800 cm2V
-1s
-1 for relatively thicker (~ 4- 6 layer ) EG.
Next, a simple laser irradiation method was used to fabricate EG-laser modified
epitaxial graphene (LEG)-EG Schottky junction device on SiC substrates. These
devices show nanosecond photo responses, making them promising for high-speed
applications. More importantly, the devices exhibit a uniform broadband (200 nm -
1064 nm) photoresponse, demonstrating that the EG-LEG-EG Schottky junction
devices are not only excellent visible and infrared sensors, but they are also
superior UV detectors. The uniform photoresponse in EG-LEG-EG is distinctively
different from conventional semiconductor photon detectors whose photoresponse
strongly depends on light wavelengths. An efficient external photoresponsivity (or
efficiency) of ~0.1 A W-1
is achieved by an interdigitated EG-LEG-EG device
biased at -10 V while the efficiency is ~ 0.5 mA W-1
with a single line
photodetector at zero-bias.
Furthermore, asymmetric Au and Al metallization scheme with pristine EG was
140
found to exhibit efficient and fast photodetection properties. Specifically, it is
observed that the number of layers in EG further influences the photocurrent
magnitude and response time regardless of incident photon energy or intensity. The
maximum external photoresponsivity ~ 31.3 mA W-1
at a bias voltage ~ 0.7 V
under illumination with 632.8 nm wavelength is estimated.
I have also developed a simple, dry, effective, economical, and environmentally
safe method of mild oxygen plasma treatment in EG devices to produce high-gain
ultraviolet (UV) photodetectors. An oxygen plasma treated epitaxial graphene
(OPTEG) device shows high selectivity, large photoconductive gain in order of
104, and detectivity of 1.3 × 10
12 Jones in the UV spectral range, and that highlight
its potential advantages over Si and other semiconductor UV photodetectors. The
method’s compatibility with CMOS process flow make it an efficient technique to
fabricate graphene based high-gain photodetectors, as well as potential applications
in UV photodetection, imaging etc.
Multilayer EG on semi insulating C-face SiC substrate, which behave as
electronically decoupled single sheets of graphene, shows a large (~ 50 %) linear
magnetoresistance (LMR), which is stable at room temperature, distinguishing EG
from other carbon materials. A two-dimensional (2D) magnetotransport effect is
observed, which gives rise to LMR in EG. The LMR property observed in EG on
C-face SiC suggests potential applications as magnetic sensors.
Future work
1. Preparation of EG on SiC substrates still needs further optimization to
produce single, or tri-layer graphene over large area on the substrate, up to
141
wafer scale. High quality EG on C-face SiC would be of much interest for
magnetotransport properties such as spin injection in spin valve devices, in
which high mobility EG would be a great advantage. Furthermore, as-
grown EG on C-face shows the LMR properties and this material may
further be explored in geometrical devices (For example, van der Pauw
geometry) to improve the LMR magnitude further.
2. Although the laser patterning is a simple method to fabricate large area
devices, it has limitations such as low device efficiency (~ 0.5 mA W-1
) at
zero-bias. The maximum photoresponsivity ~ 0.1 A W-1
is achieved with
the interdigitated device at bias voltage of -10 V. This further provides a
practical means to reduce the width of LEG in the device in order to
produce low-power consuming and high-efficiency photodetectors. The
asymmetric metallization scheme in graphene with reduced channel length
in an interdigitated device structure could be interesting to explore it on
wafer scale and hence leading to high perfomance photodetectors.
Nevertheless, graphene can also be incorporated in highly light absorbing
materials such as quantum dots (QDs), or other semiconductor materials to
produce high efficiency photoconductive or photovoltaic devices.
3. To investigate the effect of thermal annealing, reduction of channel length
and asymmetric metallization scheme in oxygen plasma treated epitaxial
graphene devices. These strategies can be implemented with the OPTEG
devices to further enhance its performance. For example, the device
response time (of the order of seconds) is slower and a device with
142
narrower channel length (reducing carrier transit time) may shorten the
response time.
4. Altering the electronic and optical properties of single layer/few layer
graphene via oxygen plasma treatments. For example, further in-depth
studies could be done to investigate adsorbed oxygen-related doping or
bandgap opening, many body effects (such as electron-electron, or electron-
hole interactions), and 2D saddle-point excitons. Future work can also be
done by varying the number of graphene layers for the photodetector
studies, and more in-depth studies of the mechanisms involved (e.g. using
optical ellipsometry, TEM) in OPTEG devices.
143
Bibliography
1. K. I. Bolotin, K. J. Sikes, Z. Jiang, M. Klima, G. Fudenberg, J. Hone, P. Kim and
H. L. Stormer, Solid State Commun., 2008, 146, 351.
2. S. V. Morozov, K. S. Novoselov, M. I. Katsnelson, F. Schedin, D. C. Elias, J. A.
Jaszczak and A. K. Geim, Phys. Rev. Lett., 2008, 100, 016602.
3. J. Hass, R. Feng, T. Li, X. Li, Z. Zong, W. A. de Heer, P. N. First, E. H. Conrad,
C. A. Jeffrey and C. Berger, Appl. Phys. Lett., 2006, 89, 143106.
4. Y. B. Zhang, Y. W. Tan, H. L. Stormer and P. Kim, Nature, 2005, 438, 201.
5. K. S. Novoselov, D. Jiang, F. Schedin, T. J. Booth, V. V. Khotkevich, S. V.
Morozov and A. K. Geim, P. Natl. Acad. Sci. USA, 2005, 102, 10451.
6. K. S. Novoselov, A. K. Geim, S. V. Morozov, D. Jiang, Y. Zhang, S. V. Dubonos,
I. V. Grigorieva and A. A. Firsov, Science, 2004, 306, 666.
7. K. S. Novoselov, A. K. Geim, S. V. Morozov, D. Jiang, M. I. Katsnelson, I. V.
Grigorieva, S. V. Dubonos and A. A. Firsov, Nature, 2005, 438, 197.
8. H. Shioyama, J. Mater. Sci. Lett., 2001, 20, 499.
9. M. Lotya, Y. Hernandez, P. J. King, R. J. Smith, V. Nicolosi, L. S. Karlsson, F. M.
Blighe, S. De, Z. M. Wang, I. T. McGovern, G. S. Duesberg and J. N. Coleman, J.
Am. Chem. Soc., 2009, 131, 3611.
10. Y. Hernandez, V. Nicolosi, M. Lotya, F. M. Blighe, Z. Y. Sun, S. De, I. T.
McGovern, B. Holland, M. Byrne, Y. K. Gun'ko, J. J. Boland, P. Niraj, G.
Duesberg, S. Krishnamurthy, R. Goodhue, J. Hutchison, V. Scardaci, A. C. Ferrari
and J. N. Coleman, Nat. Nanotechnol., 2008, 3, 563.
11. S. Marchini, S. Gunther and J. Wintterlin, Phys. Rev. B, 2007, 76, 075429.
12. C. Berger, Z. M. Song, X. B. Li, X. S. Wu, N. Brown, C. Naud, D. Mayou, T. B.
Li, J. Hass, A. N. Marchenkov, E. H. Conrad, P. N. First and W. A. de Heer,
Science, 2006, 312, 1191.
13. T. A. Land, T. Michely, R. J. Behm, J. C. Hemminger and G. Comsa, Surf. Sci.,
1992, 264, 261.
14. H. Ueta, M. Saida, C. Nakai, Y. Yamada, M. Sasaki and S. Yamamoto, Surf. Sci.,
2004, 560, 183.
15. A. K. Geim and K. S. Novoselov, Nat. Mater., 2007, 6, 183.
144
16. Luxmi, S. Nie, P. J. Fisher, R. M. Feenstra, G. Gu and Y. G. Sun, J. Electron.
Mater., 2009, 38, 718.
17. C. Faugeras, A. Nerriere, M. Potemski, A. Mahmood, E. Dujardin, C. Berger and
W. A. de Heer, Appl. Phys. Lett., 2008, 92, 011914.
18. A. H. Castro Neto, F. Guinea, N. M. R. Peres, K. S. Novoselov and A. K. Geim,
Rev. Mod. Phys., 2009, 81, 109.
19. P. R. Wallace, Phys. Rev., 1947, 71, 476.
20. J. C. Charlier, P. C. Eklund, J. Zhu and A. C. Ferrari, Top. Appl. Phys., 2008, 111,
673.
21. P. Avouris, Z. H. Chen and V. Perebeinos, Nat. Nanotechnol., 2007, 2, 605.
22. Y. B. Zhang, T. T. Tang, C. Girit, Z. Hao, M. C. Martin, A. Zettl, M. F. Crommie,
Y. R. Shen and F. Wang, Nature, 2009, 459, 820.
23. Y. M. Lin and P. Avouris, Nano Lett., 2008, 8, 2119.
24. J. H. Chen, C. Jang, S. Adam, M. S. Fuhrer, E. D. Williams and M. Ishigami, Nat.
Phys., 2008, 4, 377.
25. F. Schedin, A. K. Geim, S. V. Morozov, E. W. Hill, P. Blake, M. I. Katsnelson and
K. S. Novoselov, Nat. Mater., 2007, 6, 652.
26. B. Seradjeh and M. Franz, Phys. Rev. Lett., 2008, 101, 246404.
27. B. Dlubak, M. B. Martin, C. Deranlot, B. Servet, S. Xavier, R. Mattana, M.
Sprinkle, C. Berger, W. A. De Heer, F. Petroff, A. Anane, P. Seneor and A. Fert,
Nat. Phys., 2012, 8, 557.
28. J. Hass, F. Varchon, J. E. Millan-Otoya, M. Sprinkle, N. Sharma, W. A. De Heer,
C. Berger, P. N. First, L. Magaud and E. H. Conrad, Phys. Rev. Lett., 2008, 100,
125504.
29. A. B. Kuzmenko, E. van Heumen, F. Carbone and D. van der Marel, Phys. Rev.
Lett., 2008, 100, 117401.
30. R. R. Nair, P. Blake, A. N. Grigorenko, K. S. Novoselov, T. J. Booth, T. Stauber,
N. M. R. Peres and A. K. Geim, Science, 2008, 320, 1308.
31. C. Casiraghi, A. Hartschuh, E. Lidorikis, H. Qian, H. Harutyunyan, T. Gokus, K.
S. Novoselov and A. C. Ferrari, Nano Lett., 2007, 7, 2711.
32. F. Wang, Y. B. Zhang, C. S. Tian, C. Girit, A. Zettl, M. Crommie and Y. R. Shen,
Science, 2008, 320, 206.
33. K. F. Mak, C. H. Lui, J. Shan and T. F. Heinz, Phys. Rev. Lett., 2009, 102, 256405.
34. J. McClain and J. Schrier, J. Phys. Chem. C, 2010, 114, 14332.
145
35. L. Dossel, L. Gherghel, X. L. Feng and K. Mullen, Angew Chem. Int. Edit., 2011,
50, 2540.
36. T. Gokus, R. R. Nair, A. Bonetti, M. Bohmler, A. Lombardo, K. S. Novoselov, A.
K. Geim, A. C. Ferrari and A. Hartschuh, ACS Nano, 2009, 3, 3963.
37. Z. T. Luo, P. M. Vora, E. J. Mele, A. T. C. Johnson and J. M. Kikkawa, Appl.
Phys. Lett., 2009, 94, 111909.
38. J. V. Frangioni, Curr. Opin. Chem. Biol., 2003, 7, 626.
39. X. M. Sun, Z. Liu, K. Welsher, J. T. Robinson, A. Goodwin, S. Zaric and H. J.
Dai, Nano Res., 2008, 1, 203.
40. M. I. Katsnelson, K. S. Novoselov and A. K. Geim, Nat. Phys., 2006, 2, 620.
41. E. V. Castro, K. S. Novoselov, S. V. Morozov, N. M. R. Peres, J. M. B. L. Dos
Santos, J. Nilsson, F. Guinea, A. K. Geim and A. H. C. Neto, Phys. Rev. Lett.,
2007, 99, 216802.
42. W. Chen, S. Chen, D. C. Qi, X. Y. Gao and A. T. S. Wee, J. Am. Chem. Soc.,
2007, 129, 10418.
43. X. C. Miao, S. Tongay, M. K. Petterson, K. Berke, A. G. Rinzler, B. R. Appleton
and A. F. Hebard, Nano Lett., 2012, 12, 2745.
44. Z. Y. Chen, I. Santoso, R. Wang, L. F. Xie, H. Y. Mao, H. Huang, Y. Z. Wang, X.
Y. Gao, Z. K. Chen, D. G. Ma, A. T. S. Wee and W. Chen, Appl. Phys. Lett., 2010,
96, 193302.
45. D. C. Elias, R. R. Nair, T. M. G. Mohiuddin, S. V. Morozov, P. Blake, M. P.
Halsall, A. C. Ferrari, D. W. Boukhvalov, M. I. Katsnelson, A. K. Geim and K. S.
Novoselov, Science, 2009, 323, 610.
46. V. Abdelsayed, S. Moussa, H. M. Hassan, H. S. Aluri, M. M. Collinson and M. S.
El-Shall, J. Phys. Chem. Lett., 2010, 1, 2804.
47. E. W. Hill, A. K. Geim, K. Novoselov, F. Schedin and P. Blake, IEEE T. Magn.,
2006, 42, 2694.
48. N. Tombros, C. Jozsa, M. Popinciuc, H. T. Jonkman and B. J. van Wees, Nature,
2007, 448, 571.
49. W. A. de Heer, C. Berger, M. Ruan, M. Sprinkle, X. B. Li, Y. K. Hu, B. Q. Zhang,
J. Hankinson and E. Conrad, P. Natl. Acad. Sci. USA, 2011, 108, 16900.
50. A. Locatelli, K. R. Knox, D. Cvetko, T. O. Mentes, M. A. Nino, S. C. Wang, M.
B. Yilmaz, P. Kim, R. M. Osgood and A. Morgante, ACS Nano, 2010, 4, 4879.
51. H. B. Heersche, P. Jarillo-Herrero, J. B. Oostinga, L. M. K. Vandersypen and A. F.
Morpurgo, Nature, 2007, 446, 56.
146
52. X. Q. Liu, R. Aizen, R. Freeman, O. Yehezkeli and I. Willner, ACS Nano, 2012,
6, 3553.
53. L. M. Zhu, L. Q. Luo and Z. X. Wang, Biosens. Bioelectron., 2012, 35, 507.
54. B. M. Venkatesan, D. Estrada, S. Banerjee, X. Z. Jin, V. E. Dorgan, M. H. Bae, N.
R. Aluru, E. Pop and R. Bashir, ACS Nano, 2012, 6, 441.
55. P. K. Ang, A. Li, M. Jaiswal, Y. Wang, H. W. Hou, J. T. L. Thong, C. T. Lim and
K. P. Loh, Nano Lett., 2011, 11, 5240.
56. S. Mao, K. H. Yu, G. H. Lu and J. H. Chen, Nano Res., 2011, 4, 921.
57. S. Gilje, S. Han, M. Wang, K. L. Wang and R. B. Kaner, Nano Lett., 2007, 7,
3394.
58. X. Wang, L. J. Zhi and K. Mullen, Nano Lett., 2008, 8, 323.
59. S. Bae, H. Kim, Y. Lee, X. F. Xu, J. S. Park, Y. Zheng, J. Balakrishnan, T. Lei, H.
R. Kim, Y. I. Song, Y. J. Kim, K. S. Kim, B. Ozyilmaz, J. H. Ahn, B. H. Hong and
S. Iijima, Nat. Nanotechnol., 2010, 5, 574.
60. T. T. Feng, D. Xie, Y. X. Lin, Y. Y. Zang, T. L. Ren, R. Song, H. M. Zhao, H.
Tian, X. Li, H. W. Zhu and L. T. Liu, Appl. Phys. Lett., 2011, 99, 132904.
61. G. F. Fan, H. W. Zhu, K. L. Wang, J. Q. Wei, X. M. Li, Q. K. Shu, N. Guo and D.
H. Wu, ACS Appl. Mater. Inter., 2011, 3, 721.
62. C. Xie, J. S. Jie, B. A. Nie, T. X. Yan, Q. Li, P. Lv, F. Z. Li, M. Z. Wang, C. Y.
Wu, L. Wang and L. B. Luo, Appl. Phys. Lett., 2012, 100, 193103.
63. X. W. Fu, Z. M. Liao, Y. B. Zhou, H. C. Wu, Y. Q. Bie, J. Xu and D. P. Yu, Appl.
Phys. Lett., 2012, 100, 223114.
64. H. X. Chang, Z. H. Sun, K. Y. F. Ho, X. M. Tao, F. Yan, W. M. Kwok and Z. J.
Zheng, Nanoscale, 2011, 3, 258.
65. H. Lee, K. Heo, J. Park, Y. Park, S. Noh, K. S. Kim, C. Lee, B. H. Hong, J. Jian
and S. Hong, Journal of Materials Chemistry, 2012, 22, 8372.
66. H. Y. Yang, D. I. Son, T. W. Kim, J. M. Lee and W. I. Park, Org. Electron., 2010,
11, 1313.
67. G. Konstantatos, M. Badioli, L. Gaudreau, J. Osmond, M. Bernechea, F. P. G. de
Arquer, F. Gatti and F. H. L. Koppens, Nat. Nanotechnol., 2012, 7, 363.
68. K. K. Manga, S. Wang, M. Jaiswal, Q. L. Bao and K. P. Loh, Adv. Mater., 2010,
22, 5265.
69. F. N. Xia, T. Mueller, Y. M. Lin, A. Valdes-Garcia and P. Avouris, Nat.
Nanotechnol., 2009, 4, 839.
147
70. P. A. George, J. Strait, J. Dawlaty, S. Shivaraman, M. Chandrashekhar, F. Rana
and M. G. Spencer, Nano Lett., 2008, 8, 4248.
71. J. M. Dawlaty, S. Shivaraman, M. Chandrashekhar, F. Rana and M. G. Spencer,
Appl. Phys. Lett., 2008, 92, 042116.
72. A. N. Cao, Z. Liu, S. S. Chu, M. H. Wu, Z. M. Ye, Z. W. Cai, Y. L. Chang, S. F.
Wang, Q. H. Gong and Y. F. Liu, Adv. Mater., 2010, 22, 103.
73. B. E. A. T. Saleh, M. C. Fundamentals of Photonics Ch. 18, 784-803 (Willey,
2007).
74. A. R. Wright, J. C. Cao and C. Zhang, Phys. Rev. Lett., 2009, 103.
75. Z. P. Sun, T. Hasan, F. Torrisi, D. Popa, G. Privitera, F. Q. Wang, F. Bonaccorso,
D. M. Basko and A. C. Ferrari, ACS Nano, 2010, 4, 803.
76. T. Mueller, F. N. A. Xia and P. Avouris, Nat. Photonics, 2010, 4, 297.
77. C. S. Lin, R. H. Yeh, C. H. Liao and J. W. Hong, Solid State Electron., 2002, 46,
2027.
78. H. C. Luan, K. Wada, L. C. Kimerling, G. Masini, L. Colace and G. Assanto, Opt.
Mater., 2001, 17, 71.
79. C. Berger, Z. M. Song, T. B. Li, X. B. Li, A. Y. Ogbazghi, R. Feng, Z. T. Dai, A.
N. Marchenkov, E. H. Conrad, P. N. First and W. A. de Heer, J. Phys. Chem. B,
2004, 108, 19912.
80. Y. Takahashi, Y. Ono, A. Fujiwara and H. Inokawa, J. Phys.-Condens. Mat., 2002,
14, R995.
81. X. Z. Zhou, G. Lu, X. Y. Qi, S. X. Wu, H. Li, F. Boey and H. Zhang, J. Phys.
Chem. C, 2009, 113, 19119.
82. D. Li, M. B. Muller, S. Gilje, R. B. Kaner and G. G. Wallace, Nat. Nanotechnol.,
2008, 3, 101.
83. X. Huang, X. Y. Qi, F. Boey and H. Zhang, Chem. Soc. Rev., 2012, 41, 666.
84. A. Abouimrane, O. C. Compton, K. Amine and S. T. Nguyen, J. Phys. Chem. C,
2010, 114, 12800.
85. P. C. Lian, X. F. Zhu, H. F. Xiang, Z. Li, W. S. Yang and H. H. Wang,
Electrochim. Acta, 2010, 56, 834.
86. D. Y. Pan, S. Wang, B. Zhao, M. H. Wu, H. J. Zhang, Y. Wang and Z. Jiao, Chem.
Mater., 2009, 21, 3136.
87. C. Y. Wang, D. Li, C. O. Too and G. G. Wallace, Chem. Mater., 2009, 21, 2604.
88. G. X. Wang, X. P. Shen, J. Yao and J. Park, Carbon, 2009, 47, 2049.
89. Z. M. Ao and F. M. Peeters, Phys. Rev. B, 2010, 81, 205406.
148
90. C. Ataca, E. Akturk, S. Ciraci and H. Ustunel, Appl. Phys. Lett., 2008, 93, 043123.
91. Y. Chan and J. M. Hill, Nanotechnology, 2011, 22, 305403.
92. V. Tozzini and V. Pellegrini, J. Phys. Chem. C, 2011, 115, 25523.
93. B. H. Kim, W. G. Hong, H. Y. Yu, Y. K. Han, S. M. Lee, S. J. Chang, H. R.
Moon, Y. Jun and H. J. Kim, Phys. Chem. Chem. Phys., 2012, 14, 1480.
94. G. Binnig, H. Rohrer, C. Gerber and E. Weibel, Appl. Phys. Lett., 1982, 40, 178.
95. M. F. Crommie, C. P. Lutz and D. M. Eigler, Science, 1993, 262, 218.
96. V. W. Brar, Y. Zhang, Y. Yayon, T. Ohta, J. L. McChesney, A. Bostwick, E.
Rotenberg, K. Horn and M. F. Crommie, Appl. Phys. Lett., 2007, 91, 122102.
97. F. J. Giessibl, Rev. Mod. Phys., 2003, 75, 949.
98. N. A. Geisse, Mater. Today, 2009, 12, 40.
99. http://www.andor.com/learn/applications/?docID=64.
100. http://www.chm.bris.ac.uk/pt/diamond/stuthesis/chapter2.htm.
101. A. C. Ferrari, Solid State Commun., 2007, 143, 47.
102. D. Graf, F. Molitor, K. Ensslin, C. Stampfer, A. Jungen, C. Hierold and L. Wirtz,
Nano Lett., 2007, 7, 238.
103. D. S. Lee, C. Riedl, B. Krauss, K. von Klitzing, U. Starke and J. H. Smet, Nano
Lett., 2008, 8, 4320.
104. P. Lespade, R. Aljishi and M. S. Dresselhaus, Carbon, 1982, 20, 427.
105. P. Lespade, A. Marchand, M. Couzi and F. Cruege, Carbon, 1984, 22, 375.
106. Z. H. Ni, W. Chen, X. F. Fan, J. L. Kuo, T. Yu, A. T. S. Wee and Z. X. Shen,
Phys. Rev. B, 2008, 77, 115416.
107. Z. H. Ni, H. M. Fan, Y. P. Feng, Z. X. Shen, B. J. Yang and Y. H. Wu, J. Chem.
Phys., 2006, 124, 204703.
108. H. Wilhelm, M. Lelaurain, E. McRae and B. Humbert, J. Appl. Phys., 1998, 84,
6552.
109. M. A. Pimenta, G. Dresselhaus, M. S. Dresselhaus, L. G. Cancado, A. Jorio and R.
Saito, Phys. Chem. Chem. Phys., 2007, 9, 1276.
110. R. Saito, A. Jorio, A. G. Souza, G. Dresselhaus, M. S. Dresselhaus and M. A.
Pimenta, Phys. Rev. Lett., 2002, 88, 027401.
111. C. Thomsen and S. Reich, Phys. Rev. Lett., 2000, 85, 5214.
112. W. Kohn, Phys. Rev. Lett., 1959, 2, 393.
113. S. Piscanec, M. Lazzeri, F. Mauri, A. C. Ferrari and J. Robertson, Phys. Rev. Lett.,
2004, 93, 185503.
149
114. R. Sevak Singh, V. Nalla, W. Chen, W. Ji and A. T. S. Wee, Appl. Phys. Lett.,
2012, 100, 093116.
115. R. S. Singh, V. Nalla, W. Chen, A. T. S. Wee and W. Ji, ACS Nano, 2011, 5, 5969.
116. J. A. Robinson, M. Wetherington, J. L. Tedesco, P. M. Campbell, X. Weng, J.
Stitt, M. A. Fanton, E. Frantz, D. Snyder, B. L. VanMil, G. G. Jernigan, R. L.
Myers-Ward, C. R. Eddy and D. K. Gaskill, Nano Lett., 2009, 9, 2873.
117. http://www.mrl.ucsb.edu/mrl/centralfacilities/chemistry/resPPMS.pdf.
118. http://www.qdusa.com/products/ppms.html.
119. P. N. First, W. A. de Heer, T. Seyller, C. Berger, J. A. Stroscio and J. S. Moon,
Mrs Bull., 2010, 35, 296.
120. Y. M. Lin, C. Dimitrakopoulos, K. A. Jenkins, D. B. Farmer, H. Y. Chiu, A. Grill
and P. Avouris, Science, 2010, 327, 662.
121. A. A. Lebedev, A. M. Strel'chuk, D. V. Shamshur, G. A. Oganesyan, S. P.
Lebedev, M. G. Mynbaeva and A. V. Sadokhin, Semiconductors, 2010, 44, 1389.
122. A. Charrier, A. Coati, T. Argunova, F. Thibaudau, Y. Garreau, R. Pinchaux, I.
Forbeaux, J. M. Debever, M. Sauvage-Simkin and J. M. Themlin, J. Appl. Phys.,
2002, 92, 2479.
123. S. W. Poon, W. Chen, E. S. Tok and A. T. S. Wee, Appl. Phys. Lett., 2008, 92,
104102.
124. A. Das, S. Pisana, B. Chakraborty, S. Piscanec, S. K. Saha, U. V. Waghmare, K. S.
Novoselov, H. R. Krishnamurthy, A. K. Geim, A. C. Ferrari and A. K. Sood, Nat.
Nanotechnol., 2008, 3, 210.
125. N. Ferralis, R. Maboudian and C. Carraro, Phys. Rev. Lett., 2008, 101, 156801.
126. D. Z. Feng, S. R. Liao, P. Dong, N. N. Feng, H. Liang, D. W. Zheng, C. C. Kung,
J. Fong, R. Shafiiha, J. Cunningham, A. V. Krishnamoorthy and M. Asghari, Appl.
Phys. Lett., 2009, 95, 261105.
127. A. Ghasempour Ardakani, M. Pazoki, S. M. Mahdavi, A. R. Bahrampour and N.
Taghavinia, Appl. Surf. Sci., 2012, 258, 5405.
128. A. Muller, G. Konstantinidis, M. Androulidaki, A. Dinescu, A. Stefanescu, A.
Cismaru, D. Neculoiu, E. Pavelescu and A. Stavrinidis, Thin Solid Films, 2012,
520, 2158.
129. M. Currie, F. Quaranta, A. Cola, E. M. Gallo and B. Nabet, Appl Phys Lett, 2011,
99.
130. Y. Liang, H. Liang, X. Xiao and S. Hark, Journal of Materials Chemistry, 2012,
22, 1199.
150
131. A. K. Geim, Science, 2009, 324, 1530.
132. P. Blake, P. D. Brimicombe, R. R. Nair, T. J. Booth, D. Jiang, F. Schedin, L. A.
Ponomarenko, S. V. Morozov, H. F. Gleeson, E. W. Hill, A. K. Geim and K. S.
Novoselov, Nano Lett., 2008, 8, 1704.
133. Z. Q. Li, E. A. Henriksen, Z. Jiang, Z. Hao, M. C. Martin, P. Kim, H. L. Stormer
and D. N. Basov, Nat. Phys., 2008, 4, 532.
134. T. Winzer, A. Knorr and E. Malic, Nano Lett., 2010, 10, 4839.
135. F. N. Xia, T. Mueller, Y. M. Lin and P. Avouris, 2010 Conference on Lasers and
Electro-Optics (Cleo) and Quantum Electronics and Laser Science Conference
(Qels), 2010.
136. B. Chitara, S. B. Krupanidhi and C. N. R. Rao, Appl.Phys. Lett., 2011, 99, 113114.
137. B. Chitara, L. S. Panchakarla, S. B. Krupanidhi and C. N. R. Rao, Adv. Mater.,
2011, 23, 5419.
138. T. Mueller, F. Xia, M. Freitag, J. Tsang and P. Avouris, Phys Rev B, 2009, 79.
139. S. Wang, Q. S. Zeng, L. J. Yang, Z. Y. Zhang, Z. X. Wang, T. A. Pei, L. Ding, X.
L. Liang, M. Gao, Y. Li and L. M. Peng, Nano Lett., 2011, 11, 23.
140. H. T. Liu, S. M. Ryu, Z. Y. Chen, M. L. Steigerwald, C. Nuckolls and L. E. Brus,
J. Am. Chem. Soc., 2009, 131, 17099.
141. S. Ryu, L. Liu, S. Berciaud, Y. J. Yu, H. T. Liu, P. Kim, G. W. Flynn and L. E.
Brus, Nano Lett., 2010, 10, 4944.
142. Y. L. Zhang, L. Guo, S. Wei, Y. Y. He, H. Xia, Q. D. Chen, H. B. Sun and F. S.
Xiao, Nano Today, 2010, 5, 15.
143. D. A. Sokolov, K. R. Shepperd and T. M. Orlando, J. Phys. Chem. Lett., 2010, 1,
2633.
144. X. S. Wu, M. Sprinkle, X. B. Li, F. Ming, C. Berger and W. A. de Heer, Phys.
Rev. Lett., 2008, 101, 026801.
145. K. N. Kudin, B. Ozbas, H. C. Schniepp, R. K. Prud'homme, I. A. Aksay and R.
Car, Nano Lett., 2008, 8, 36.
146. D. Yang, A. Velamakanni, G. Bozoklu, S. Park, M. Stoller, R. D. Piner, S.
Stankovich, I. Jung, D. A. Field, C. A. Ventrice and R. S. Ruoff, Carbon, 2009,
47, 145.
147. N. Wang, B. D. Yao, Y. F. Chan and X. Y. Zhang, Nano Lett., 2003, 3, 475.
148. L. Liu, S. M. Ryu, M. R. Tomasik, E. Stolyarova, N. Jung, M. S. Hybertsen, M. L.
Steigerwald, L. E. Brus and G. W. Flynn, Nano Lett., 2008, 8, 1965.
151
149. S. Ryu, M. Y. Han, J. Maultzsch, T. F. Heinz, P. Kim, M. L. Steigerwald and L. E.
Brus, Nano Lett., 2008, 8, 4597.
150. F. N. Xia, T. Mueller, R. Golizadeh-Mojarad, M. Freitag, Y. M. Lin, J. Tsang, V.
Perebeinos and P. Avouris, Nano Lett., 2009, 9, 1039.
151. E. G. Mishchenko, Phys. Rev. Lett., 2009, 103, 246802.
152. X. D. Xu, N. M. Gabor, J. S. Alden, A. M. van der Zande and P. L. McEuen, Nano
Lett., 2010, 10, 562.
153. G. C. Xing, H. C. Guo, X. H. Zhang, T. C. Sum and C. H. A. Huan, Opt. Express,
2010, 18, 4564.
154. K. Berndt, H. Durr and D. Palme, Opt. Commun., 1983, 47, 321.
155. D. M. Rayner, A. E. Mckinnon, A. G. Szabo and P. A. Hackett, Can. J. Chem.,
1976, 54, 3246.
156. S. M. N. Sze, K. K. Physics of Semiconductor Devices, 3rd and I. N. Y. ed.; John
Wiley & Sons, 2007.
157. N. Toyama, J. Appl. Phys., 1988, 64, 2515.
158. M. S. Arnold, J. D. Zimmerman, C. K. Renshaw, X. Xu, R. R. Lunt, C. M. Austin
and S. R. Forrest, Nano Lett., 2009, 9, 3354.
159. M. Freitag, Y. Martin, J. A. Misewich, R. Martel and P. H. Avouris, Nano Lett.,
2003, 3, 1067.
160. X. Lv, Y. Huang, Z. B. Liu, J. G. Tian, Y. Wang, Y. F. Ma, J. J. Liang, S. P. Fu,
X. J. Wan and Y. S. Chen, Small, 2009, 5, 1682.
161. H. B. Michaelson, J. Appl. Phys., 1977, 48, 4729.
162. D. M. Rayner, A. E. Mckinnon and A. G. Szabo, Rev. Sci. Instrum., 1977, 48,
1050.
163. D. A. Dikin, S. Stankovich, E. J. Zimney, R. D. Piner, G. H. B. Dommett, G.
Evmenenko, S. T. Nguyen and R. S. Ruoff, Nature, 2007, 448, 457.
164. S. Park, N. Mohanty, J. W. Suk, A. Nagaraja, J. H. An, R. D. Piner, W. W. Cai, D.
R. Dreyer, V. Berry and R. S. Ruoff, Adv. Mater., 2010, 22, 1736.
165. H. A. Becerril, J. Mao, Z. Liu, R. M. Stoltenberg, Z. Bao and Y. Chen, ACS Nano,
2008, 2, 463.
166. A. Bagri, C. Mattevi, M. Acik, Y. J. Chabal, M. Chhowalla and V. B. Shenoy, Nat.
Chem., 2010, 2, 581.
167. J. B. Zhang, N. Xi, K. W. C. Lai, H. Z. Chen, Y. L. Luo and G. Y. Li, 2007 7th
IEEE Conference on Nanotechnology, Vol 1-3, 2007, 1164.
168. S. Liu, J. Ye, Y. Cao, Q. Shen, Z. Liu, L. Qi and X. Guo, Small, 2009, 5, 2371.
152
169. Y. Lin, K. Zhang, W. F. Chen, Y. D. Liu, Z. G. Geng, J. Zeng, N. Pan, L. F. Yan,
X. P. Wang and J. G. Hou, ACS Nano, 2010, 4, 3033.
170. X. M. Geng, L. Niu, Z. Y. Xing, R. S. Song, G. T. Liu, M. T. Sun, G. S. Cheng, H.
J. Zhong, Z. H. Liu, Z. J. Zhang, L. F. Sun, H. X. Xu, L. Lu and L. W. Liu, Adv.
Mater., 2010, 22, 638.
171. G. Kalita, S. Adhikari, H. R. Aryal, M. Umeno, R. Afre, T. Soga and M. Sharon,
Appl. Phys. Lett., 2008, 92, 063508.
172. T. Sugino, D. Hirata, I. Yamamura, K. Matsuda and J. Shirafuji, J. Electron.
Mater., 1996, 25, 733.
173. U. Cvelbar, B. Markoli, I. Poberaj, A. Zalar, L. Kosec and S. Spaic, Appl. Surf.
Sci., 2006, 253, 1861.
174. N. Xiao, X. C. Dong, L. Song, D. Y. Liu, Y. Tay, S. X. Wu, L. J. Li, Y. Zhao, T.
Yu, H. Zhang, W. Huang, H. H. Hng, P. M. Ajayan and Q. Y. Yan, ACS Nano,
2011, 5, 2749.
175. C. Soci, A. Zhang, B. Xiang, S. A. Dayeh, D. P. R. Aplin, J. Park, X. Y. Bao, Y.
H. Lo and D. Wang, Nano Lett., 2007, 7, 1003.
176. A. Bera and D. Basak, Appl. Phys. Lett., 2008, 93, 053102.
177. H. Y. Jeong, J. Y. Kim, J. W. Kim, J. O. Hwang, J. E. Kim, J. Y. Lee, T. H. Yoon,
B. J. Cho, S. O. Kim, R. S. Ruoff and S. Y. Choi, Nano Lett., 2010, 10, 4381.
178. Z. Q. Luo, J. Z. Shang, S. H. Lim, D. H. Li, Q. H. Xiong, Z. X. Shen, J. Y. Lin and
T. Yu, Appl. Phys. Lett., 2010, 97, 233111.
179. D. J. Bishop, D. C. Tsui and R. C. Dynes, Phys. Rev. Lett., 1980, 44, 1153.
180. G. J. Dolan and D. D. Osheroff, Phys. Rev. Lett., 1979, 43, 721.
181. D. J. Thouless, Phys. Rev. Lett., 1977, 39, 1167.
182. E. McCann, K. Kechedzhi, V. I. Fal'ko, H. Suzuura, T. Ando and B. L. Altshuler,
Phys. Rev. Lett., 2006, 97, 146805.
183. B. L. Altshuler, A. G. Aronov and P. A. Lee, Phys. Rev. Lett., 1980, 44, 1288.
184. X. S. Wu, X. B. Li, Z. M. Song, C. Berger and W. A. de Heer, Phys. Rev. Lett.,
2007, 98, 136801.
185. P. A. N. V. Joshi, Science, and Technology ( Marcel Dekker, New Yark, 1990).
186. K. W. Liu, M. Sakurai and M. Aono, Sensors-Basel, 2010, 10, 8604.
187. A. Behnam, J. Johnson, Y. Choi, L. Noriega, M. G. Ertosun, Z. C. Wu, A. G.
Rinzler, P. Kapur, K. C. Saraswat and A. Ural, J. Appl. Phys., 2008, 103, 114315.
188. H. L. Xue, X. Z. Kong, Z. R. Liu, C. X. Liu, J. R. Zhou, W. Y. Chen, S. P. Ruan
and Q. Xu, Appl. Phys. Lett., 2007, 90, 201118.
153
189. Y. Liu, C. R. Gorla, S. Liang, N. Emanetoglu, Y. Lu, H. Shen and M. Wraback, J
Electron Mater., 2000, 29, 69.
190. H. Ohta, M. Hirano, K. Nakahara, H. Maruta, T. Tanabe, M. Kamiya, T. Kamiya
and H. Hosono, Appl. Phys. Lett., 2003, 83, 1029.
191. I. S. Jeong, J. H. Kim and S. Im, Appl. Phys. Lett., 2003, 83, 2946.
192. M. Martens, J. Schlegel, P. Vogt, F. Brunner, R. Lossy, J. Wurfl, M. Weyers and
M. Kneissl, Appl. Phys. Lett., 2011, 98, 211114.
193. M. Sofos, J. Goldberger, D. A. Stone, J. E. Allen, Q. Ma, D. J. Herman, W. W.
Tsai, L. J. Lauhon and S. I. Stupp, Nat. Mater., 2009, 8, 68.
194. http://www.eoc-inc.com/UV_detectors_silicon_carbide_photodiodes.htm.
195. http://www.eoc-inc.com/ifw/JEP%205-365%200408.pdf.
196. C. K. Wang, S. J. Chang, Y. K. Su, Y. Z. Chiou, C. S. Chang, T. K. Lin, H. L. Liu
and J. J. Tang, Semicond. Sci. Tech., 2005, 20, 485.
197. Y. Yamamoto, T. Fukushima, Y. Suna, N. Ishii, A. Saeki, S. Seki, S. Tagawa, M.
Taniguchi, T. Kawai and T. Aida, Science, 2006, 314, 1761.
198. G. Mazzeo, J. L. Reverchon, J. Y. Duboz and A. Dussaigne, IEEE Sens. J., 2006,
6, 957.
199. I. Rodionov, J. M. Bidault, I. Crotty, P. Fonte, F. Galy, V. Peskov and O. Zanette,
in 2005 IEEE Nuclear Science Symposium Conference Record, Vols 1-5, ed. B.
Yu, 2005, 3045.
200. J. Hass, W. A. de Heer and E. H. Conrad, J. Phys.-Condens. Mat., 2008, 20,
323202.
201. J. Hass, R. Feng, J. E. Millan-Otoya, X. Li, M. Sprinkle, P. N. First, W. A. de
Heer, E. H. Conrad and C. Berger, Phys. Rev. B, 2007, 75, 214109.
202. F. Varchon, P. Mallet, J. Y. Veuillen and L. Magaud, Phys. Rev. B, 2008, 77,
165415.
203. T. Ohta, A. Bostwick, J. L. McChesney, T. Seyller, K. Horn and E. Rotenberg,
Phys. Rev. Lett., 2007, 98, 206802.
204. S. Latil, V. Meunier and L. Henrard, Phys. Rev. B, 2007, 76, 201402(R).
205. M. Sprinkle, D. Siegel, Y. Hu, J. Hicks, A. Tejeda, A. Taleb-Ibrahimi, P. Le Fevre,
F. Bertran, S. Vizzini, H. Enriquez, S. Chiang, P. Soukiassian, C. Berger, W. A. de
Heer, A. Lanzara and E. H. Conrad, Phys. Rev. Lett., 2009, 103, 226803.
206. D. L. Miller, K. D. Kubista, G. M. Rutter, M. Ruan, W. A. de Heer, P. N. First and
J. A. Stroscio, Science, 2009, 324, 924.
154
207. Y. Kopelevich, J. H. S. Torres, R. R. da Silva, F. Mrowka, H. Kempa and P.
Esquinazi, Phys. Rev. Lett., 2003, 90, 156402.
208. Y. F. Chen, M. H. Bae, C. Chialvo, T. Dirks, A. Bezryadin and N. Mason, Journal
of physics. Condensed matter : an Institute of Physics journal, 2010, 22, 205301.
209. J. S. Hu and T. F. Rosenbaum, Nat. Mater., 2008, 7, 697.
210. J. H. Ngai, Y. Segal, D. Su, Y. Zhu, F. J. Walker, S. Ismail-Beigi, K. Le Hur and
C. H. Ahn., Phys. Rev. B, 2010, 81, 241307.
211. T. Ando, A. B. Fowler and F. Stern, Rev. Mod. Phys., 1982, 54, 437.
212. Z. M. Liao, H. C. Wu, S. Kumar, G. S. Duesberg, Y. B. Zhou, G. L. W. Cross, I.
V. Shvets and D. P. Yu, Adv. Mater., 2012, 24, 1862.
213. E. T. i. M. I. J. L. Olsen, New York, 1962).
214. W. R. Branford, A. Husmann, S. A. Solin, S. K. Clowes, T. Zhang, Y. V.
Bugoslavsky and L. F. Cohen, Appl. Phys. Lett., 2005, 86, 202116.
215. A. Husmann, J. B. Betts, G. S. Boebinger, A. Migliori, T. F. Rosenbaum and M. L.
Saboungi, Nature, 2002, 417, 421.
216. A. Patane, W. H. M. Feu, O. Makarovsky, O. Drachenko, L. Eaves, A. Krier, Q.
D. Zhuang, M. Helm, M. Goiran and G. Hill, Phys. Rev. B, 2009, 80, 115207.
217. H. Xu, J. Lu, Y. Xia, J. Yin and Z. Liu, Solid State Commun., 2012, 152, 1150.
218. P. P. Freitas, R. Ferreira, S. Cardoso and F. Cardoso, J. Phys.-Condens. Mat.,
2007, 19, 165221.
219. X. Zhang, Q. Z. Xue and D. D. Zhu, Phys. Lett. A, 2004, 320, 471.
220. Y. F. Chen, M. H. Bae, C. Chialvo, T. Dirks, A. Bezryadin and N. Mason, J.
Phys.-Condens. Mat., 2010, 22, 205301.
221. B. Li, C. Y. Kao, J. W. Yoo, V. N. Prigodin and A. J. Epstein, Adv. Mater., 2011,
23, 3382.
222. N. P. Raju, V. N. Prigodin, K. I. Pokhodnya, J. S. Miller and A. J. Epstein,
Synthetic Met., 2010, 160, 307.
223. H. G. Johnson, S. P. Bennett, R. Barua, L. H. Lewis and D. Heiman, Phys. Rev. B,
2010, 82, 085202.
224. T. J. McArdle, J. O. Chu, Y. Zhu, Z. H. Liu, M. Krishnan, C. M. Breslin, C.
Dimitrakopoulos, R. Wisnieff and A. Grill, Appl. Phys. Lett., 2011, 98, 132108.
225. T. A. Len, L. Y. Matzui, I. V. Ovsienko, Y. I. Prylutskyy, V. V. Andrievskii, I. B.
Berkutov, G. E. Grechnev and Y. A. Kolesnichenko, Low Temp. Phys., 2011, 37,
819.
155
226. Y. Oshima, T. Takenobu, K. Yanagi, Y. Miyata, H. Kataura, K. Hata, Y. Iwasa
and H. Nojiri, Phys. Rev. Lett., 2010, 104, 168803.
227. R. Mahendiran, A. K. Raychaudhuri, A. Chainani and D. D. Sarma, Rev. Sci.
Instrum., 1995, 66, 3071.
228. J. W. Bai, R. Cheng, F. X. Xiu, L. Liao, M. S. Wang, A. Shailos, K. L. Wang, Y.
Huang and X. F. Duan, Nat. Nanotechnol., 2010, 5, 655.
229. K. Eto, Z. Ren, A. A. Taskin, K. Segawa and Y. Ando, Phys .Rev. B, 2010, 81,
195309.
230. A. A. Abrikosov, Phys. Rev. B, 1998, 58, 2788.
231. P. N. Argyres and E. N. Adams, Phys. Rev., 1956, 104, 900.
232. R. T. Bate, R. K. Willardson and A. C. Beer, J. Phys. Chem. Solids, 1959, 9, 119.
233. T. J. Diesel and W. F. Love, Phys. Rev., 1961, 124, 666.
234. J. W. Mcclure and W. J. Spry, Phys. Rev., 1968, 165, 809.
235. H. Tang, D. Liang, R. L. J. Qiu and X. P. A. Gao, ACS Nano, 2011, 5, 7510.
236. M. M. Parish and P. B. Littlewood, Nature, 2003, 426, 162.
237. V. I. Falko, J. Phys.-Condens. Mat., 1990, 2, 3797.
238. S. V. Morozov, K. S. Novoselov, M. I. Katsnelson, F. Schedin, L. A.
Ponomarenko, D. Jiang and A. K. Geim, Phys. Rev. Lett., 2006, 97, 016801.
239. W. R. Anderson, D. R. Lombardi, R. G. Wheeler and T. P. Ma, IEEE Electr.
Device. L., 1993, 14, 351.
240. G. Bergmann, Int. J. Mod. Phys. B, 2010, 24, 2015.
241. T. E. R. o. M. a. A. C. U. P. P L Rossiter, Cambridge, 1987).
242. G. Bergmann, Phys. Rev. B, 1983, 28, 2914.
243. R. S. Markiewicz and L. A. Harris, Phys. Rev. Lett., 1981, 46, 1149.
244. Z. Ovadyahu and Y. Imry, Phys. Rev. B, 1981, 24, 7439.
245. P. A. Lee and T. V. Ramakrishnan, Rev. Mod. Phys., 1985, 57, 287.