Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Chemical vapor deposited single layer graphene as transparent electrodes for flexible photovoltaic devices Jiang, Zenan 2018

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata

Download

Media
24-ubc_2019_february_jiang_zenan.pdf [ 7.61MB ]
Metadata
JSON: 24-1.0376033.json
JSON-LD: 24-1.0376033-ld.json
RDF/XML (Pretty): 24-1.0376033-rdf.xml
RDF/JSON: 24-1.0376033-rdf.json
Turtle: 24-1.0376033-turtle.txt
N-Triples: 24-1.0376033-rdf-ntriples.txt
Original Record: 24-1.0376033-source.json
Full Text
24-1.0376033-fulltext.txt
Citation
24-1.0376033.ris

Full Text

CHEMICAL VAPOR DEPOSITED SINGLE LAYER GRAPHENE AS TRANSPARENT ELECTRODES FOR FLEXIBLE PHOTOVOLTAIC DEVICES by  Zenan Jiang  B.Sc., Tsinghua University, 2007 M. Sc., Simon Fraser University, 2010  A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF  DOCTOR OF PHILOSOPHY in THE FACULTY OF GRADUATE AND POSTDOCTORAL STUDIES (Electrical and Computer Engineering)  THE UNIVERSITY OF BRITISH COLUMBIA (Vancouver)   December 2018  © Zenan Jiang, 2018 ii  The following individuals certify that they have read, and recommend to the Faculty of Graduate and Postdoctoral Studies for acceptance, the dissertation entitled:  Chemical vapor deposited single layer graphene as transparent electrodes for flexible photovoltaic devices  submitted by Zenan Jiang in partial fulfillment of the requirements for the degree of Doctor of Philosophy in Electrical and computer engineering   Examining Committee: Dr. Peyman Servati  Supervisor  Dr. John Madden  Supervisory Committee Member  Dr. Guangrui (Maggie) Xia  Supervisory Committee Member Dr. Mu Chiao  University Examiner Dr. Shuo Tang  University Examiner  iii  Abstract  Graphene has attracted intensive attention for various electronic applications in the past decades given its unique properties. The synthesis of graphene by chemical vapor deposition (CVD) on copper foil provides the opportunity to deliver large-area, high quality, and continuous graphene films. The metal foil can be removed with a wet etching process. The transferred CVD graphene films can be integrated into existing semiconductor device manufacturing platforms, or into low-cost roll-to-roll manufacturing of flexible electronics.  Since graphene is a two-dimensional material, the optical, mechanical, and electrical properties can easily be altered with surface modification. Copper etchants used in graphene transfer process can lead to films with different levels of doping and mechanical strength. The topology and temperature dependent electrical properties of transferred graphene using three different etchants were investigated. All of the graphene samples demonstrate a doping level above 1013 cm-3. The graphene films prepared with cupric sulfate solution presents the most uniform and continuous layer, with the least density of defects. Metallic and organic residues, defects and grain boundaries, as well as intercalated water molecules, attribute to the variation in conductivity and permittivity of the films.  By coating the films with charge selective materials, graphene sheets with improved sheet resistance and transparency of about 90% were fabricated. The hole-selective transparent conductors show about 50% reduction in resistivity. All the samples demonstrate high stability with repeated bending of over 800 cycles. Organic photovoltaic (OPV) devices using the hole-selective graphene transparent conductors as electrodes were iv  fabricated on plastic substrates. Less than 5% fluctuation in power conversion efficiency (PCE) was noticed when the devices were bent up to 130 degrees.  As an extension of this work, the photovoltaic characteristics of inverted OPV devices fabricated with AlxZn(1-x)O as an electron transport layer with Al fraction of up to 11% were reported. The light-soaking effect can be eliminated by using more than 4% of Al doping. All devices demonstrate PCE over 3.4% with air-stability of over 150 days. The light-soaking mechanism is investigated by employing a numerical simulation on the devices.   v  Lay Summary  Graphene is a two-dimensional material with unique optical and electrical properties. Large-area graphene sheets have great potential in flexible electronics. The aim of this research is to investigate the electrical, optical, and mechanical properties of monolayer polycrystalline graphene films on various substrates, and explore possible applications of graphene in flexible optoelectronic devices. Transparent conductors, which are necessary components in solar cells and light-emitting diodes, fabricated using graphene sheets are studied. The graphene conductors are assembled for flexible organic photovoltaic devices as the transparent electrodes. The electrical and mechanical performances of the devices are characterized. The influence of different preparation techniques, device structures, and substrate choices on the power conversion behavior and flexibility of organic photovoltaic devices are investigated.  vi  Preface  All the research projects presented in this thesis are conducted in the Flexible Electronics and Energy Lab (FEEL) in the Department of Electrical and Computer Engineering of the University of British Columbia. All the research is under the supervision of Dr. Peyman Servati. Chapter 2 focuses on the electrical properties of large-area graphene films grown by CVD and transferred with a wet etching process. The author is the leading investigator. The author designed the projects, carried out most of the experiments, and analyzed the data. Mr. Weijun Luo and Dr. Guangrui Xia have collaborated on the Raman 2D mapping experiments. Dr. Zihe Ren helped with the dielectric spectroscopy measurements.  Chapter 3 is based on the development of CVD-grown graphene coated with charge selective materials as transparent conductors. The author takes responsibility for project design, the performance of experiments, and data analysis.  Chapter 4 is the extension of the work in chapter 3, by applying the graphene transparent conductors in the fabrication of organic photovoltaic devices. The author has been leading the projects, on program design, experiment conduction, and data analysis. Dr. Rowshan Rahmanian and Dr. Saeid Soltanian helped perform the polymer deposition with spray-coating.  Chapter 5 is related to the fabrication and investigation of inverted organic photovoltaic devices using zinc oxide and aluminum-doped zinc oxide as the electron transport layer. The author has been leading the project. Dr. Bobak Gholamkhass assisted vii  with the lifetime measurements of the solar cells. Dr. Saeid Soltanian provided useful help on the synthesis of ZnO nanoparticles.   Publications and Presentations  Journal papers: - Zenan Jiang, S. Soltanian, Bobak Gholamkhass, Peyman Servati ‘Light-soaking Free Organic Photovoltaic Devices with Sol-gel Deposited ZnO and AZO Electron Transport Layers’, RSC Advances, 2018. 8: p. 36542-36548 - Zenan Jiang, Zihe Ren, Weijun Luo, Guangtui (Maggie) Xia, S. Soltanian, Peyman Servati ‘Temperature-dependent conductivity of CVD-grown single-layer graphene’, in preparation - Zenan Jiang, S. Soltanian, Peyman Servati ‘Flexible organic solar cells using charge-selective graphene transparent electrodes’, in preparation  Conference Presentations:  - Zenan Jiang, S. Soltanian, Bobak Gholamkhass, Peyman Servati. ‘Light-soaking Effect in Inverted Solar Cells with Sol-gel Derived AZO and ZnO Electron Exaction Layers’, the 21st annual Pacific Centre for Advanced Materials and Microstructures (PCAMM 2016) Meeting, Vancouver, BC, Canada, Dec 10, 2016 - Zenan Jiang, Saeid Soltanian, Peyman Servati. ‘Highly Flexible Organic Photovoltaic Devices Based on Monolayer Graphene–Metal Oxide Hybrids as viii  Transparent Electrode’, the 2015 MRS Spring Meeting and Exhibit, San Francisco, USA, Apr 06-10, 2015 - Zenan Jiang, Zihe Ren, Peyman Servati. ‘Aging Effects and Evolution of Temperature-Dependent Polarization of Single Layer Graphene’, the 2015 MRS Spring Meeting and Exhibit, San Francisco, USA, Apr 06-10, 2015 - Zenan Jiang, Saeid Soltanian, Bobak Gholamkhass, Peyman Servati. ‘Flexible Organic Photovoltaic Devices Using Single-Layer Graphene Combined with Metal Oxides as Transparent Electrode’, the 4th international symposium on Graphene devices, Bellevue, USA, Sep 21-25, 2014   ix  Table of Contents  Abstract ............................................................................................................................. iii Lay Summary .....................................................................................................................v Preface ............................................................................................................................... vi Table of Contents ............................................................................................................. ix List of Tables .................................................................................................................. xiii List of Figures ................................................................................................................. xiv List of Abbreviations ................................................................................................... xxiv Acknowledgements ...................................................................................................... xxvi Dedication .................................................................................................................... xxvii Chapter 1: Introduction ....................................................................................................1 1.1 Graphene: synthesis and properties .....................................................................1 1.2 Flexible electronics and transparent electrodes ...................................................3 1.3 Organic photovoltaic devices ...............................................................................7 1.4 Research objectives ............................................................................................13 1.5 Thesis Overview ................................................................................................13 Chapter 2: Temperature-dependent electrical properties of CVD-grown graphene 15 2.1 Introduction and motivation ...............................................................................15 2.2 Sample preparation: CVD growth and film transfer ..........................................16 2.2.1 Graphene growth ............................................................................................16 2.2.2 Graphene Transfer .........................................................................................20 2.3 Measurements and related background theory ..................................................22 x  2.3.1 Raman Spectroscopy ......................................................................................22 2.3.2 Scattering-type scanning near-field optical microscopy (s-SNOM) ..............23 2.3.3 Dielectric spectroscopy ..................................................................................24 2.4 Results and Discussion ......................................................................................27 2.4.1 Morphology study ..........................................................................................27 2.4.2 Raman spectroscopy ......................................................................................31 2.4.3 s-SNOM test...................................................................................................36 2.4.4 Dielectric spectra ...........................................................................................41 2.4.4.1 DC conductivity .....................................................................................41 2.4.4.2 AC conductivity .....................................................................................46 2.4.4.3 Permittivity ............................................................................................50 2.5 Conclusion .........................................................................................................55 Chapter 3: Graphene and graphene/charge-selective layer hybrid transparent conductors .........................................................................................................................56 3.1 Introduction and motivation ...............................................................................56 3.2 Sample preparation and measurements ..............................................................59 3.2.1 Sample preparation ........................................................................................59 3.2.2 Measurement techniques ................................................................................60 3.3 Graphene sheets on different substrates .............................................................62 3.3.1 Graphene layer transfer ..................................................................................65 3.3.2 Sheet resistance and transparency ..................................................................68 3.3.3 Flexibility .......................................................................................................69 3.4 Hole-selective graphene transparent conductors ...............................................71 xi  3.4.1 PEDOT:PSS coated graphene transparent conductors ..................................71 3.4.1.1 Fabrication .............................................................................................71 3.4.1.2 Results ....................................................................................................72 3.4.2 MoO3 coated graphene transparent conductors .............................................78 3.4.2.1 Fabrication .............................................................................................78 3.4.2.2 Results ....................................................................................................81 3.4.3 Summary ........................................................................................................85 3.5 Electron-selective graphene transparent conductors ..........................................87 3.5.1 Fabrication .....................................................................................................87 3.5.1.1 ZnO colloidal nanoparticles as electron-selective materials ..................87 3.5.1.2 ZnO films prepared with the sol-gel method as electron-selective materials 89 3.5.2 Results ............................................................................................................90 3.6 Conclusion .........................................................................................................92 Chapter 4: Flexible organic solar cells with hole-selective graphene transparent electrodes ..........................................................................................................................93 4.1 Introduction and motivation ...............................................................................93 4.2 Sample preparation ............................................................................................94 4.2.1 Device structure and fabrication ....................................................................94 4.2.2 Measurements ................................................................................................97 4.3 Experimental results...........................................................................................97 4.3.1 OPV devices with hole-selective graphene TCE on glass substrates ............97 xii  4.3.2 Spin-coated OPV devices with hole-selective graphene TCE on PEN substrates ..................................................................................................................100 4.3.3 Spray-coated OPV devices with hole-selective graphene TCE on PEN substrates ..................................................................................................................104 4.4 Conclusion .......................................................................................................107 Chapter 5: Light-soaking Free Organic Photovoltaic Devices with Sol-gel Deposited ZnO and AZO Electron Transport Layers .................................................................108 5.1 Introduction and motivation .............................................................................108 5.2 Experiments .....................................................................................................110 5.3 Results and Discussion ....................................................................................112 5.3.1 Morphology Study .......................................................................................112 5.3.2 Photovoltaic properties ................................................................................114 5.3.3 Light-soaking effect .....................................................................................119 5.4 Conclusions ......................................................................................................124 Chapter 6: Conclusion and future work ......................................................................126 6.1 Contribution .....................................................................................................126 6.2 Future work ......................................................................................................128 Bibliography ...................................................................................................................130  xiii  List of Tables  Table 1.1 Comparison of transparent conductors: Indium-tin oxide (ITO), carbon nanotube (CNT), dielectric/metal/dielectric (DMD), metal nanowire (NW), and single/multi-layer graphene (from Ref [37]). .......................................................................6 Table 2.1 Average positions, FWHM, and integrated intensity of Raman peak of graphene samples. ..............................................................................................................33 Table 3.1 Technical properties of glass and flexible substrates [105, 106]. ......................63 Table 3.2 Sheet resistance and transmittance at 550 nm of SLG on various substrates. The transmittance shown here is the average value of more than 50 samples for each type of substrate. The transmittance is calculated with respect to the power of incident light. .....68 Table 4.1 Essential parameters of the OPV devices using hole-selective graphene TCE on glass substrates. ..................................................................................................................99 Table 4.2  Essential parameters of the OPV devices with spin-cast active layers using hole-selective graphene TCE on PEN..............................................................................102 Table 4.3  Essential parameters of the OPV devices with spray coated active layers using hole-selective graphene TCE on PEN..............................................................................105 Table 5.1 Extracted parameters, short-current density (Jsc), open-circuit voltage (Voc), fill factor (FF) and power conversion efficiency (PCE) for the OPV devices. .....................115  xiv  List of Figures  Figure 1.1 Graphene as transparent conductor. (a) Transmittance as a function of wavelength and (b) thickness dependence of the sheet resistance for different transparent conductors. Two limiting lines (and the shaded area) for graphene are calculated from theoretical predictions. (c) Transmittance versus sheet resistance for different transparent conductors. Shaded area for graphene films calculated from theoretical predictions [40]. (d) Transmittance versus sheet resistance for graphene films grouped according to production strategies. A theoretical line is also plotted for comparison. (From Ref. [40]) .7 Figure 1.2 Schematic of a typical BHJ OPV device: (a) band diagram and charge flow; (b) device structure for a BHJ device with a normal structure; (c) band diagram and (d) device structure of an inverted BHJ device. ......................................................................10 Figure 1.3 Schematic of the I-V curves of an OPV device. ...............................................11 Figure 2.1 a) Photograph and b) schematic of the tube furnace used for graphene growth. ...............................................................................................................................18 Figure 2.2 a) Schematic of graphene growth process; (b) photo of the freshly grown graphene taken from the furnace........................................................................................19 Figure 2.3 a) Schematic of the graphene transfer process: graphene-coated Cu foils were spin-caoted with PMMA and floated on etchant solutions for Cu removal; the samples were rinsed with DI water twice before sccoping up with the substrates; PMMA was removed in acetone bath after the samples were dried in a vacuum oven overnight; b) photograph of a graphene sheet transferred to the SiO2/Si substrate. ................................21 xv  Figure 2.4 Schematic of working principle of the scattering-type scanning near-field optical microscopy (s-SNOM): surface plasmons that are induced by the tip and the subsequent interference due to the emission of scattered and reflected plasmonic waves..................................................................................................................................24 Figure 2.5 a) Schematic of the circuit used in the dielectric spectroscopy; b) illustration of the graphene samples used for dielectric spectroscopy measurements. ............................25 Figure 2.6 Optical and AFM images of graphene sheets SiO2/Si substrates using different copper etchants: (a, d) optical and (g) AFM images of G-Cu films; (b, e) optical and (h) AFM images of G-Fe films; (c, f) optical and (i) AFM images of G-PSF films. Scale bar of all the AFM images is shown in the lower right corner of the figure............................28 Figure 2.7 SEM images of the graphene samples using different copper etchants: (a, d) G-Cu films; (b, e) G-Fe films; (c, f) G-PSF films. Legends are listed on the right side of the figure. (SEM parameters: EHT 10.00 kV, working distance 19.5 mm, magnification 10.5 kX.) ....................................................................................................................................30 Figure 2.8 Raman spectra of a) G and b) 2D peaks of graphene transferred by CuSO4 (G-Cu), FeCl3 (G-Fe), and (NH4)2(SO4)2 (G-PSF). .................................................................31 Figure 2.9 The histogram of a) c) the position distribution, and b) d) FWHM distribution of graphene transferred by CuSO4 (G-Cu), FeCl3 (G-Fe), and (NH4)2(SO4)2 (G-PSF). ....32 Figure 2.10 2D mapping of graphene samples transferred with different copper etchants. a), b), and c) show the intensity ratio of 2D and G peaks; d), e), and f) show the intensity ratio of D and G peaks. Scale bars are shown on the right side of the figure. ...................35 xvi  Figure 2.11  Topography (a, d, g) and infrared near-field amplitude (b, e, h) and phase (c, f, i) images taken simultaneously at the wavenumber of 1047 cm-1 (~ 9.6 µm in wavelength). .......................................................................................................................37 Figure 2.12  Topography, surface plasmon interferometry images taken simultaneously at the wavenumber of 890 cm-1 (~11.3 µm in wavelength) of (d-i) G-Cu films and (j-o) G-PSF films. The images of the mechanically exfoliated graphene (a-c) were measured as a reference. ............................................................................................................................40 Figure 2.13 Temperature-dependent DC conductivity and relative permittivity of graphene samples transferred by a) CuSO4 (G-Cu), b) FeCl3 (G-Fe), and c) (NH4)2(SO4)2 (G-PSF). .............................................................................................................................42 Figure 2.14 Frequency dependence of the real part of the complex electrical conductivity of graphene transferred by CuSO4 at a series of temperatures: a) -50 ºC to 95 ºC; b) 95 ºC to 150 ºC; c) 150 ºC to 190 ºC; d) 190 ºC to 300 ºC. .........................................................47 Figure 2.15 Frequency dependence of the real part of the complex electrical conductivity of graphene transferred by FeCl3 at a series of temperatures: a) -50 ºC to 95 ºC; b) 95 ºC to 150 ºC; c) 150 ºC to 190 ºC; d) 190 ºC to 300 ºC. .........................................................49 Figure 2.16 Frequency dependence of the real part of the complex electrical conductivity of graphene transferred by (NH4)2(SO4)2 at a series of temperatures: a) -50 ºC to 95 ºC; b) 95 ºC to 150 ºC; c) 150 ºC to 190 ºC; d) 190 ºC to 300 ºC. ...............................................50 Figure 2.17 Frequency dependence of the real part of the complex permittivity of graphene transferred by CuSO4 at a series of temperatures: a) -50 ºC to 95 ºC; b) 95 ºC to 150 ºC; c) 150 ºC to 190 ºC; d) 190 ºC to 300 ºC. .............................................................51 xvii  Figure 2.18 Frequency dependence of the real part of the complex permittivity of graphene transferred by FeCl3 at a series of temperatures: a) -50 ºC to 95 ºC; b) 95 ºC to 150 ºC; c) 150 ºC to 190 ºC; d) 190 ºC to 300 ºC. .............................................................52 Figure 2.19 Frequency dependence of the real part of the complex permittivity of graphene transferred by (NH4)2(SO4)2 at a series of temperatures: a) -50 ºC to 95 ºC; b) 95 ºC to 150 ºC; c) 150 ºC to 190 ºC; d) 190 ºC to 300 ºC. ...............................................53 Figure 2.20 Temperature dependency relaxation frequency of graphene sheets prepared with different copper etchants. ...........................................................................................54 Figure 3.1 Schematic of samples during I-V measurements .............................................60 Figure 3.2 Schematic of the bending test apparatus: the two ends of a sample are held by one fixed arm and one bending arm which is controlled by a stepper motor. The center of the sample is aligned to a 0.2 mm diameter thick wire and is bent over the wire during the test. (Adapted from Ref. [104])..........................................................................................61 Figure 3.3 Morphology and transmittance of glass and the flexible substrates: AFM images of a) PI, b) PET, and c) PEN substrates over an area of 50×50 µm; d) optical transmittance of substrates at wavelength 400-800 nm. ....................................................64 Figure 3.4 Optical and AFM images of the graphene sheets on different substrates. Scale bar of the AFM images is shown on the right side of l). ...................................................67 Figure 3.5 Transmittance of graphene sheets on different substrates. The transmittance is calculated with respect to the power of incident light. ......................................................69 Figure 3.6 Flexibility of graphene samples on different substrates: a) change of sheet resistance with bending angle up to 130 degrees; b) change of resistance during xviii  continuous bending up to 800 bending cycles. The original values of the three plots in b) are around 0. The curves are plotted with a vertical offset for clarity. ..............................70 Figure 3.7 Optical, SEM, and AFM images of 25 nm thick PEDOT:PSS coated graphene sheets on PET (a, b) and PEN (c, d, and e). (SEM parameters: EHT 10.00 kV, working distance 19.5 mm, magnification 10.5 kX.) .......................................................................72 Figure 3.8 a) Reduction of sheet resistance and b) transmittance at 550 nm of PEDOT:PSS coated graphene on PET samples. The transmittance of all the samples is calculated with respect to the light after substrate absorption. ..........................................73 Figure 3.9 Reduction of sheet resistance and transmittance at 550 nm of graphene on PEN with PEDOT:PSS coating with: a, d) different thickness; b, e) annealing time; c, f) annealing temperature. The transmittance of all the samples is calculated with respect to the light after substrate absorption. ....................................................................................75 Figure 3.10 Flexibility of PEDOT:PSS coated graphene samples on PEN. a-c) the increment of sheet resistance when the PEDOT:PSS coated graphene samples were bent up to 130 degree; d-e) change of sheet resistance of PEDOT:PSS coated graphene samples during continuous bending. The original values of the curves in d) and e) are around 0. The curves are plotted with a vertical offset for clarity. ....................................77 Figure 3.11 Flexibility of PEDOT:PSS coated graphene samples on PET. a) The increment of sheet resistance when the PEDOT:PSS coated graphene samples were bent up to 130 degree; b) change of sheet resistance of PEDOT:PSS coated graphene samples during continuous bending. The original values of the curves in b) are around 0. The curves are plotted with a vertical offset for clarity. ...........................................................78 xix  Figure 3.12 Optical, SEM, and AFM images of 20 nm MoO3 coated graphene sheets on PET (a, b) and PEN (c, d, and e). SEM parameters: EHT 10.00 kV, working distance 19.5 mm, magnification 10.5 kX. ..............................................................................................80 Figure 3.13 Reduction of sheet resistance and transmittance at 550 nm of graphene on PEN (a, b) and PET (c, d) with MoO3 film. .......................................................................82 Figure 3.14 Flexibility of MoO3 coated graphene samples on PEN (a, b) and PET (c, d). a, c) the increment of sheet resistance when the MoO3 coated graphene samples were bent up to 130 degree; b, d) change of sheet resistance of MoO3 coated graphene samples during continuous bending. The original values of the curves in b) and d) are around 0. The curves are plotted with a vertical offset for clarity. ....................................................83 Figure 3.15 AFM images of MoO3 coated graphene samples on PET after bending. .......84 Figure 3.16 Schematic of graphene coated with charge selective materials during bending. ..............................................................................................................................85 Figure 3.17 a) Transmittance vs. sheet resistance plot of SLG, PEDOT:PSS/SLG, and MoO3/SLG stacks on PEN from this work (transmittance are calculated relative to PEN), compared with typical transparent conductors. ITO data is adapted from Ref. [22], the data of large-area multi-layer graphene is adapted from Ref. [11], and the data of conductive PEDOT:PSS with additives is adapted from Ref. [115]. b) Transmittance of SLG, PEDOT:PSS/SLG, and MoO3/SLG stacks on PEN from this work (transmittance are calculated relative to PEN) compared to ITO (~ 60 nm in thickness) over 400-800 nm wavelengths........................................................................................................................86 xx  Figure 3.18 Optical and SEM images of ZnO colloidal nanoparticles coated graphene sheets on PI. (SEM parameters: EHT 10.00 kV, working distance 19.5 mm, magnification 10.5 kX.) .....................................................................................................88 Figure 3.19 Optical and SEM images of ZnO sol-gel film coated graphene sheets on PI using a, d) 2-ME, b, e) IPA, c, f) MeOH as solvents. (SEM parameters: EHT 10.00 kV, working distance 19.5 mm, magnification 10.5 kX.) ........................................................89 Figure 3.20  Sheet resistance and transmittance of ZnO thin film coated graphene on PI (transmittances are calculated relative to PI), compared with typical transparent conductors. ITO data is adapted from Ref. [22], the data of large-area multi-layer graphene is adapted from Ref. [11], and the data of conductive PEDOT:PSS with additives is adapted from Ref. [115]. .................................................................................91 Figure 3.21 Flexibility of ZnO coated graphene samples on PI: a) change of sheet resistance of ZnO coated graphene samples during continuous bending. The original values of the curves are around 0. The curves are plotted with a vertical offset for clarity; b) the increment of sheet resistance when the MoO3 coated graphene samples were bent up to 130 degree. ................................................................................................................91 Figure 4.1 (a) Band alignment of layers in the OPV devices; (b) schematic OPV devices structures; the overlapped area between the top and bottom contact is the effective device area. ....................................................................................................................................96 Figure 4.2 a) Front and back sides of OPV devices using hole-selective graphene as TCE; the green box marks the graphene covered area. b) Reflective and c) transmitted optical images of OPV devices using hole-selective graphene as TCE on glass substrates. .........98 xxi  Figure 4.3 a) Current density-voltage characteristics and b) external quantum efficiency of OPV devices using hole-selective graphene TCE on glass substrates. .......................100 Figure 4.4 a) Spin-coated P3HT:PCBM films on hole-selective graphene TCE on PEN substrates. b) Front and c) back sides of OPV devices using hole-selective graphene as TCE; the green box marks the graphene covered area. d) Reflective and e) transmitted optical images of OPV devices using hole-selective graphene as TCE on PEN. ............101 Figure 4.5 a) Current density-voltage characteristics and b) external quantum efficiency of OPV devices with spin-cast active layers using hole-selective graphene TCE on PEN. .................................................................................................................................103 Figure 4.6 Plots of essential parameters vs. bending angles of the OPV devices with spin-cast active layers using hole-selective graphene TCE on PEN........................................104 Figure 4.7 a) Sprayed P3HT:PCBM films on hole-selective graphene TCE on PEN substrates before and after annealing. b) Reflective and c) transmitted optical images of OPV devices using hole-selective graphene as TCE on PEN. .........................................105 Figure 4.8 a) Current density-voltage characteristics and b) external quantum efficiency of OPV devices with spray coated active layers using hole-selective graphene TCE on PEN. .................................................................................................................................106 Figure 4.9 Plots of essential parameters vs. bending angles of the OPV devices with spray coated active layers using hole-selective graphene TCE on PEN. ..................................107 Figure 5.1 a) The energy level alignment and b) device structure in an inverted OPV using P3HT:PCBM. .........................................................................................................111 Figure 5.2 SEM images of 25 nm thick of a) ZnO, b) 1%-AZO, c-f) Al-doped ZnO with different Al fraction nanoparticles coated on a glass substrate. All the samples were xxii  annealed at 250 ºC. a) and b) have higher magnification (EHT 10.00 kV, working distance 9.4 mm, with in-lens camera), and c-f) have the same magnification (EHT 10.00 kV, working distance 9.6 mm with in-lens camera). .......................................................113 Figure 5.3 a) EDX spectra of 25 nm thick of AZO films with various Al fraction. The spectra were normalized to the Oxygen peak. The Al peak is zoomed in and shown in the insect. b) Integrated area of the Al peak vs. Al molar ratio in the precursor inks. ..........114 Figure 5.4 Photovoltaic performance of the inverted devices using ZnO and AZO as electron transport layers: a) current density-voltage (J-V) characterizations at the day of fabrication; b) external-quantum efficiency (EQE) normalized to the highest absorbing wavelength (520 nm) of P3HT; c) J-V and EQE of sample A4 tested at the day of fabrication and 150 days layer; d) EQE of the sample ZnO within 50 days of fabrication normalized to the highest EQE value...............................................................................116 Figure 5.5 Photovoltaic parameters of OPV devices with 1% AZO nanoparticle film as ETL. Device structures of all samples were the same. The AZO films were heated at a various temperature in the ambient atmosphere after deposition. ...................................117 Figure 5.6 Light-soaking effect in the inverted OPV devices with ZnO/AZO electron transport layers: a) periodically J-V characterizations of sample A1 measured under continuous light illumination; normalized b) fill factor and c) EQE of each sample with illumination time. .............................................................................................................120 Figure 5.7 a) Transmittance versus wavelength plot and b) (αh)2 versus photon energy plot of AlxZn(1-x)O nanoparticles, x ranged from 0 to 11 %, deposited on glass substrates. c) optical bandgap of ZnO nanoparticles doped with various level of Al; the inset schematic shows the shifting of Fermi level when Al is added into ZnO nanoparticles. 122 xxiii  Figure 5.8 The simulated band diagram of the inverted OPV devices using a) ZnO (black lines) and b) AZO (red lines) with 4% Al electron transport layers. ...............................123  xxiv  List of Abbreviations  2-ME   2-methoxyethanol AFM   Atomic force microscope  ALD    Atomic Layer Deposition  AZO   Aluminum doped ZnO BHJ   Bulk heterojunction CB   Chlorobenzene CNT   Carbon nanotube CVD   Chemical vapor deposition DMD   Dielectric/metal/dielectric EDX   Energy-dispersive X-ray spectroscopy EQE   External quantum efficiency  ETL    Electron transport layer FF   Fill factor FWHM  Full-width at half maximum GO   Graphene Oxide HOMO  Highest occupied molecular orbital HTL    Hole transport layer IPA   Isopropanol IQE   Internal quantum efficiency  IR   Infrared ITO   Indium-tin oxide xxv  LPCVD  Low-pressure chemical vapor deposition LUMO  Lowest unoccupied molecular orbital MeOH  Methanol MFC   Mass flow controllers NP   Nanoparticle NW   Nanowire OPV   Organic photovoltaic device  P3HT   Poly(3-hexylthiophene-2,5-diyl) PCBM  [6,6]-Phenyl C61 butyric acid methyl ester PCE    Power conversion efficiency PEDOT:PSS   Poly(3,4-ethylenedioxythiophene)-polystyrene sulfonate PEN   Polyethylene-naphthalate PET   Polyethylene terephthalate PI   polyimide  PMMA  Poly(methylmethacrylate) PSF   Ammonia persulfate RMS   Root-mean-square s-SNOM  Scattering-type scanning near-field optical microscopy SEM   Scanning electron microscopic SLG   Single layer graphene TCE   Transparent conducting electrodes UV   Ultra-violet xxvi  Acknowledgements  I would like to express my sincere gratitude to my supervisor, Dr. Peyman Servati, for his insightful supervision, on both scientific and life attitude.  I would like to acknowledge the research associate, Dr. Saeid Soltanian, and postdoctoral fellow, Dr. Bobak Gholamkhass for their immense help and guidance. I would like to offer my deepest appreciation to Dr. Guangrui Xia and Dr. John Madden. As the members of the supervision committee, they have provided valuable feedback and support over the research. I would like to acknowledge my dearest colleagues in the FEEL lab, for the help and friendship.  I would like to thank Dr. Patricia Mooney, Dr. Simon Watkins, Dr. Yatung Cherng, Ms. Grace Li, Mr. Weijun Luo, for their useful discussions and assistance.  I would like to thank my parents for making me believe that science is the one true thing to believe in and pursuit.  Finally, I would like to thank my husband, Dr. Zihe Ren, for being a perfect companion on scientific discussions, creative thinking, housework sharing, and life.  xxvii  Dedication       Never grow older    1  Chapter 1: Introduction   1.1 Graphene: synthesis and properties In 2004, single layer graphene (SLG) was obtained by Dr. Geim and his group for the first time through scotch tape exfoliation [1], which then caused unprecedented attention to graphene and the related materials for a variety of applications. This discovery was followed by intensive studies on graphene in the past decade. Several record-high characteristics have been reported for graphene, and the fact that these unique properties all combine in a single material makes graphene an important and promising two-dimensional material to be explored for diversified applications.  The unique properties of graphene are the most important reason for the intensive studies of researchers and fast progress. Several reported characteristics of graphene have set up the new records. The electron mobility in suspended graphene is demonstrated to be as high as 2.5×105 cm2V-1s-1 at room temperature [2], and ~ 6×106 cm2V-1s-1 at 4K [3], due to the ballistic transport of charge carriers within the two-dimensional sheet. Graphene is also highly transparent, with >90% transparency under ultra-violet and visible (UV-Vis) irradiations. Due to the sp2 hybridization, the covalent bonds between nearby carbon atoms in pristine graphene are very strong, leading to Young’s modulus of 1 TPa and intrinsic strength of 130 GPa [4]. Graphene has a high thermal conductivity of up to 5300 W m-1 K-1 [5]. Moreover, it has low permeability to gases and can pass high electric current densities.  As the minimum of its conduction band and the maximum of the valence bands overlap, single layer graphene is a 0 eV band gap material behaving as a metal. By doping 2  the graphene films, the Fermi-level can be shifted up and down. In bilayer graphene, a bandgap can be introduced when a transverse electrical field is applied [6-8], turning the material into a semiconductor.   Several methods have been employed to obtain large scale and reproducible graphene sheets. Chemical vapor deposition (CVD) of graphene onto transition metals provides a balance between consistency of material quality and cost-effectiveness. The initial purpose of exposing metals to carbon precursors was to study the catalytic and thermionic activities of carbon at metal surfaces [9]. The interest switched to the actual growth of graphene since 2004. Low-pressure CVD on iridium single crystal was found to produce graphene [10]. However, this is not a promising technique for graphene production due to the difficulties of transferring graphene from the chemically inert Ir to other substrates as well as the high cost of the metal itself [11]. On the other hand, growth on metals like Ni and Co, which is compatible with silicon industry, received attention. However, because of the high carbon solubility in these metals, multi-layer graphene is generally formed, and the areas covered with single-layer graphene are highly non-uniform.  The first reported CVD growth of large-scale uniform single layer graphene was in 2009 [12], where polycrystalline copper foil was used as the metal catalyst. Methane, the carbon source, was decomposed thermo-catalytically on the surface of the copper foil. The growth process is self-limiting and stops when the Cu surface is fully covered by graphene due to the low solubility of carbon in copper. The as-grown graphene film consists of mostly single-layer graphene, with typically less than 5% coverage of multilayers [12].  By increasing the size of copper foil, graphene grown by CVD was demonstrated to scale up, with large grain sizes in the range of 20 to 500 μm and electron mobility values 3  of up to 25000 cm2 V-1 s-1 at room temperature after transferring to SiO2 substrates [13]. Graphene sheets were produced using roll-to-roll process, with width as high as ~50 cm and electron mobility higher than 7000 cm2V-1s-1 [14]. By applying the low-pressure CVD synthesis together with the roll-to-roll transfer techniques, researchers can directly transfer single-layer graphene films to flexible substrates to fabricate transparent conductive films. Graphene films that are 100 m long and have the sheet resistance as low as 150 Ohm/square after doping have been attained using photo-curable epoxy resins [15].  The great and interesting properties of graphene lead to many practical applications. The transparency, conductivity and tensile strength can be exploited in flexible electronics; the controllable bandgap and high electron mobility can be applied in transistors; high specific strength lets graphene be useful as additives in polymers to enforce mechanical properties; the large surface/volume ratio of graphene also makes it a promising candidate for sensing applications; and the impermeability to gases also provides solution for protective coatings.   1.2 Flexible electronics and transparent electrodes Flexible electronics refer to the electronic devices and circuits built up on ultra-thin and flexible substrates. With their bendable, foldable or even stretchable nature, consumer devices can become lighter, smarter, and better-integrated. Flexible electronics can potentially bring significant impact on wearable technologies, internet-of-things, medical devices, etc. By building up devices and electronic systems on bendable or even stretchable platforms, we will be able to produce devices and integrate large systems by using novel 4  manufacturing methods, e.g., printing, roll-to-roll deposition, lamination, etc., which are inconceivable for the current rigid platform. “Soft” materials are employed for flexible systems, whereby organic materials with high elasticity are considered and applied.  With a bright future lying ahead, there are still challenges left before the commercial uptake of these technologies. Flexible electronic devices need to be mechanically and electrically robust; performance of the devices has to match with the current rigid ones; materials with certain transparency and low resistivity are required; and they also need to be environment-friendly if applied on wearable electronics.  One of the most obvious demands to realize flexible electronics and optoelectronics is the transparent electrodes. The current transparent conductors are mostly semiconductors of binary or ternary metal oxides, such as doped Indium oxides, Zinc oxides, and the dominant Indium tin oxides (ITO), which provide the best combination of transparency and conductivity [16-18]. Typical ITO contains about 90% In2O3 and about 10% SnO2, in which Sn atoms act as n-type dopants. The electrical and optical properties of ITO are determined by impurity levels [16]. Commercially available ITO has above 80% transparency and sheet resistance as low as 10 Ohm/sqr on glass substrates. The shortcomings of ITO are obvious as well: high synthesis cost due to the cost of Indium, difficulties in processing and patterning [16, 19], and limitations in flexible applications due to its brittleness [20].  Metal nanowires and grids, carbon nanotubes, and organic polymer PEDOT:PSS are investigated as potential replacements for ITO. These transparent conductors have advantages of low-cost solution processing and good mechanical flexibility [21-30]. 5  However, metal nanowires have rough surfaces and poor thermal and chemical stability [24, 25]. Carbon nanotubes also suffer from poor surface morphology, as well as the low conductivity [27, 28]. Dielectric/metal/dielectric (DMD) multi-layer electrodes have also been noted as alternatives for ITO [31-36]. With acceptable electrical and optical properties and good flexibility, DMD conductors have very narrow emission width and strong angular dependence of the incident and emission light, which makes them challenging for light-emitting diodes and photovoltaic devices [37]. Organic poly(3,4-ethylenedioxythiophene) polystyrene sulfonate (PEDOT:PSS) can be conductive and transparent as well. However, its poor stability under ultra-violet irradiation is a major limit [38, 39]. Detailed comparison and parameters are shown in Table 1.1 and Figure 1.1.  As an ultra-thin and strong film with high transparency, graphene is a promising candidate for flexible conductors. The transparent conductive film can be made by simply transferring synthesized graphene onto flexible substrates. Single layer graphene has >90% transmittance for the UV-Visible wavelengths due to the ultra-thin film (Figure 1.1a). Nevertheless, the actual resistivity of graphene is highly dependent on the synthesis techniques (Figure 1.1d). Among all, graphene grown by CVD and transferred to the SiO2 wafer by the roll-to-roll method shows the highest conductivity (blue rhombus dots in Figure 1.1, chemically doped graphene sheet, up to four layers). Overall, graphene shows prominent optical and electrical potential as transparent conductors, compared to other ITO replacements.  6  Table 1.1 Comparison of transparent conductors: Indium-tin oxide (ITO), carbon nanotube (CNT), dielectric/metal/dielectric (DMD), metal nanowire (NW), and single/multi-layer graphene (from Ref [37]).  TE Sheet Resistance (Ohm/sqr) Trans- parency Surface Flexibility Cost Stability Other issues ITO <50 > 80% Smooth Poor High High High deposition T Graphene >100 > 90% Smooth Good Low High Complicated processing CNT 500-1000 70-80% Rough Good Low High Non-uniformity DMD 15-100 > 80% Smooth Good Low High Microcavity,  Emission width Metal NW 10-100 > 80% Rough Good High Poor Light scattering PEDOT: PSS 50-300 80-95% Smooth Good Low Poor Adhesion 7   Figure 1.1 Graphene as transparent conductor. (a) Transmittance as a function of wavelength and (b) thickness dependence of the sheet resistance for different transparent conductors. Two limiting lines (and the shaded area) for graphene are calculated from theoretical predictions. (c) Transmittance versus sheet resistance for different transparent conductors. Shaded area for graphene films calculated from theoretical predictions [40]. (d) Transmittance versus sheet resistance for graphene films grouped according to production strategies. A theoretical line is also plotted for comparison. (From Ref. [40])  1.3 Organic photovoltaic devices Solar energy is one of the most important clean and renewable energy resources in the world. It has the largest available volume as compared to the other known sources 8  (Solar: 3850000EJ, Wind: 2250EJ, Biomass: 3000EJ) [41, 42]. Nowadays, solar energy is collected mostly in two ways: photo-thermal and photovoltaic effect. The first method has been used for a long time in human history. Sunlight is reflected and focused on the glass tube in which water flows through and being vaporized. The vapor is then transported to a turbine engine in the power plant to generate electricity. The long-term cost of solar-thermal energy in some area can almost match with the traditional fossil fuel energy. However, due to the complication of the construction of optical reflectors, large area solar-thermal energy power plant can only be built in limited areas.  The photovoltaic effect is the generation of electric current in a material under radiation. Electrons in the valence band absorb the incident photons with energy higher than the bandgap of the active layer material and being stimulated to the conduction band. The generated electron-hole pairs, i.e., excitons, are then separated at the interface of a junction (p-n junction or metal-semiconductor junction) and transferred to corresponding electrodes, resulting in current flow. The light energy is converted into electrical energy. Organic photovoltaic (OPV) devices have attracted many researchers in recent years. Due to the high solubility of materials in common organic solvents, it is possible to manufacture flexible solar cells with low-cost roll-to-roll processes. However, the biggest drawback for OPV devices currently is the low internal quantum efficiency (IQE) and power conversion efficiency (PCE), mainly due to the large bandgap of organic materials and the low charge separation rate. The highest reported PCE is still below 11% [43-45], which is far below the commercially available Si-based PV devices (~20%) [43]. Also, instabilities of active light absorbing materials and the interfacial layers, caused by 9  oxidation, reduction, UV exposure, etc., can result in degradation of device performance, undermining the long-term cost-effectiveness [46].  A typical heterojunction OPV device consists of an electron donor layer and electron acceptor layer between the two electrodes, one of which needs to be transparent/semi-transparent to allow the light to come in. The photo-generated excitons, which are the bounded electron-hole pairs, move to the donor-acceptor interfaces and are dissociated by the local electric field caused by the difference of electron affinity and ionization energies of the donor and acceptor regions. The charges are then collected by the corresponding electrodes, forming the output electrical current. A schematic device and its band structure are shown in Figure 1.2.  The charge carriers in the organic semiconductors usually have much lower mobility and a shorter lifetime than those in the inorganic semiconductors. The excitons can only travel through a short distance before recombination, which is one of the reasons for the low internal quantum efficiency (IQE) in organic solar cells. Therefore, to shorten the carrier transport distance and to elevate the possibility of exciton dissociation, a bulk heterojunction (BHJ) structure, whereby the electron donor and acceptor materials are mixed to form a nanoscale blend, and are deposited together as a single layer, is employed in the OPV devices [47], as shown in Figure 1.2 (b). In this system, the transport length for charge carriers is limited to a nanometer scale [48], since the electron donor and acceptor materials tightly interpenetrate into each other. This leads to a significant rise in IQE due to better charge dissociation and massive increment of interfacial areas.  Two charge selection layers are often added between the electrodes and the active organic semiconductors. The layers utilize carefully chosen materials that have the right 10  electron affinity and ionization energies, so only the selected carriers can pass through. They are named as the hole transport layer (HTL) and the electron transport layer (ETL), respectively for holes and electrons. These layers help eliminate charge recombination at the interface of electrodes and the active layer and improve the overall device performance.   Figure 1.2 Schematic of a typical BHJ OPV device: (a) band diagram and charge flow; (b) device structure for a BHJ device with a normal structure; (c) band diagram and (d) device structure of an inverted BHJ device.  As most of the light absorbing polymers are electron donor materials, to place this layer closer to the transparent electrode can lead to higher absorption. Therefore, OPV 11  devices with the hole transport layer next to the transparent electrode are called as “normal” devices, while the ones with the electron transport layer next to the transparent electrode are the “inverted” devices. The schematics of the band diagram and structure of the normal and inverted devices are shown in Figure 1.2.   Figure 1.3 Schematic of the I-V curves of an OPV device.  Several essential parameters are employed to characterize the performance of an OPV device. As is shown in Figure 1.3, the red line is a typical I-V curve of an ideal OPV device under irradiation. The short-circuit current ISC is the current through the solar cell when the voltage across the solar cell is zero. It is the largest current which may be obtained from the solar cell. The open-circuit voltage VOC is the voltage occurred at zero current, which corresponds to the forward bias required to balance the bias associated with the light-generated current. The ideal working condition of the device is set to where the maximum 12  output power Pmax occurs. Therefore, the PCE is defined as the ratio of the maximum output power to the input power of irradiation Pin.  𝐏𝐂𝐄 =𝐏𝐦𝐚𝐱𝐏𝐢𝐧 × 𝟏𝟎𝟎%       Equation 1.1 The fill factor (FF) is the ratio of the optimal output power to the theoretical maximum output power: 𝑭𝑭 =  𝑷𝒎𝒂𝒙𝑰𝑺𝑪×𝑽𝑶𝑪× 𝟏𝟎𝟎%      Equation 1.2 And the external quantum efficiency (EQE), which is used to characterize the conversion rate of photons to electrons, is calculated as   𝐄𝐐𝐄(𝛌) =  𝐭𝐡𝐞 𝐧𝐮𝐦𝐛𝐞𝐫 𝐨𝐟 𝐨𝐮𝐭𝐩𝐮𝐭 𝐞𝐥𝐞𝐜𝐭𝐫𝐨𝐧𝐬𝐭𝐡𝐞 𝐧𝐮𝐦𝐛𝐞𝐫 𝐨𝐟 𝐢𝐧𝐜𝐢𝐝𝐞𝐧𝐭 𝐩𝐡𝐨𝐭𝐨𝐧𝐬 × 𝟏𝟎𝟎%  Equation 1.3 The major option of transparent electrodes utilized in OPV devices is conductive metal oxides: ITO and Fluorine-tin oxide, which are discussed in the previous session. Graphene is incorporated into the OPV devices as a transparent electrode that can provide high transmittance over the whole UV-Vis spectra, lower manufacturing cost, and flexibility.  Current studies on graphene as transparent electrodes are mainly focused on improving the film preparation techniques, modifying the electrical and optical properties of graphene film, etc. Two major concerns are noted. First, most of the reported graphene included devices demonstrate lower PCE than the ones using the same structure and materials but the ITO electrode [49-54]. This is primarily because of the low conductivity of graphene films. Intensive investigations have been and are ongoing for better preparation 13  method and post-growth treatment to obtain graphene films with better quality. Also, the graphene films are in direct contact with the substrates, the interfacial layers, and/or the active layers in the OPV devices. Hence, the surface structure and properties of graphene will be tuned by the adjacent layers. 1.4 Research objectives This research aims to investigate the electrical, optical, and mechanical properties of CVD-growth monolayer graphene films transferred onto various substrates. Hybrid transparent conductor fabricated using graphene sheets combined with charge selective materials is studied. The graphene and the hybrid films are assembled for flexible OPV devices as transparent electrodes. The electrical and mechanical performances of the devices are characterized. Discussions on the influence of different preparation techniques, coupling layers, and substrate choices on the power conversion behavior and flexibility of OPV devices are conducted.  1.5 Thesis Overview Six chapters are included in the dissertation. The present chapter introduced the background of the CVD growth and featured properties of graphene and its potential application as transparent electrodes in organic energy harvesting devices. The working principles and structure of the organic photovoltaic devices are introduced as well.  Chapter 2 presents the electrical properties of CVD-grown single-layer graphene transferred with different copper etchants. Various characterization techniques have been 14  applied to investigate the influence of different etchants on its optical and electrical properties. Chapter 3 demonstrates the performance of graphene flexible transparent conductors. Charge selective materials to modify the conductivity and transparency of graphene. Both hole- and electron-selective transparent conductors are fabricated and studied.  Chapter 4 is an extension of the previous chapter by applying the hole-selective transparent conductors as electrodes in flexible OPV devices. The solar energy conversion performance of the devices, as well as the flexibility, are discussed.  In chapter 5, inverted organic solar cells using zinc oxide and aluminum-doped zinc oxide as electron transport layer are fabricated. The widely existed issue of “light-soaking effect” in OPV devices containing a metal oxide layer is investigated. The mechanism of the effect is analyzed with experimental and simulation results.     15  Chapter 2: Temperature-dependent electrical properties of CVD-grown graphene  2.1 Introduction and motivation Graphene has great potential to provide solutions for the next generation of micro- and nano-electronics, and photonic devices [40, 55]. Intensive research has been done for the application of graphene in high-frequency transistors and two-dimensional devices [56-59]. The synthesis technique of graphene by CVD on copper foil provides the opportunity of delivering large-area, high quality, and continuous graphene films. The metal foil can be removed with a wet etching process. The transferred graphene films can be integrated into the existing semiconductor device manufacture platforms. However, certain criteria need to be met when incorporating graphene into devices. Poly(methylmethacrylate) (PMMA) or thermal-released tapes are commonly used as supporting materials during wet etching [14, 60]. The residue polymer impurities can lead to a significant variation of the morphology [61, 62]. Metallic contamination coming from the etching formula as well as from the metal foil can alter the electrical and mechanical properties of the materials [63]. These trace impurities can potentially lead to serious degradation of device performance and lifetime [64]. The most commonly used etchants of copper foil are sodium or ammonia persulfate (PSF) and ferric chloride [65-67]. Cupric sulfate in acidic solution is also used since it does not bring in the extra type of metal ions. Previous work has demonstrated that the selection of copper etchant has a strong influence on the quality of graphene. Graphene transferred with PSF has fracture load one order of 16  magnitude higher than the ones transferred with ferric chloride [68]. Ions in the etchants may attach to the surface of graphene and cause doping effect on the graphene films [69]. The contaminations on graphene surfaces can introduce interface defects to the adjacent materials in the devices, which may cause further problems. Therefore, it is essential to study the effect of etchant on graphene films, especially on the electrical properties. Most of the reported studies are focused on the close to 0 K temperature where graphene shows the unique semi-metal behavior. To the best of the author’s knowledge, no temperature dependency electrical properties of graphene have been done in the temperature range which the devices are usually running at.  In this chapter, graphene sheets grown with CVD and transferred with three different copper etchants were studied. Raman spectroscopy, near-field scanning microscopy, as well as dielectric spectroscopy, were performed to investigate the electrical and topography of the films. All the graphene samples exhibit a certain level of p-type doping. The density of defects and grain boundaries were studied. Conductivity and permittivity of the graphene films were measured at a series of temperatures and frequency. The charge transport mechanisms of electrons in different graphene films were discussed.  2.2 Sample preparation: CVD growth and film transfer 2.2.1 Graphene growth Single layer graphene was synthesized by low-pressure chemical vapor deposition (LPCVD) using high purity copper foil (25 μm in thickness, Alfa-Aesar) as the metal catalyst.  17  The photo and schematic of the CVD tube furnace are shown in Figure 2.1. The CVD system consists of a furnace, a quartz tube, a rotary pump, two mass flow controllers (MFC), one flow meter, one pressure transducer, one pressure controller, two gas bottles, and multiple valves for flow control. The methane (Praxair 3.7 ultrahigh purity 99.97%) and hydrogen (Praxair 5.0 ultra-high purity 100.00%) gas are connected to the MFC. The flow rate is monitored and controlled by using a CCR-400 MFC flow meter. The rotary pump (Varian DS302 1HP) used to pump down the CVD chamber is controlled by an MKS 253B exhaust throttle valve positioned between the quartz tube and the pump. The pressure is controlled and monitored by MKS Type 651C pressure controller. A 3-Zone Carbolite TZF furnace is used to achieve a uniform growth temperature of 1000oC over a length of 20 cm in the center.  18   Figure 2.1 a) Photograph and b) schematic of the tube furnace used for graphene growth.  The copper foil pieces (~1×1 cm) were placed in a quartz boat and loaded into the CVD furnace. The chamber was evacuated to a base pressure of less than 10 mTorr (base pressure). The system was heated to a growth temperature of 1000 °C under hydrogen gas (90 mTorr + base pressure). The copper was annealed for 30 min in hydrogen flow for the removal of native oxides and other contaminants as well as for the growth of copper grain. Methane gas was subsequently introduced into the tube (460 mTorr + base pressure) as the carbon source. The flow of methane was maintained for 30 min in order to achieve full decomposition of methane and the deposition of graphene on the copper surface. The 19  temperature, partial pressure of methane and hydrogen were set to achieve supersaturation condition, which led to nucleation of graphene islands on the catalyst. As the reaction proceeded, more carbon atoms covalently attached to the edge of the graphene islands. Methane molecules lost contact with copper catalyst once one graphene island meets another, ending the reaction locally. This process leads to the growth and connection of neighboring domains with the formation of a continuous layer eventually. After the synthesis was completed, the chamber was slowly cooled down to room temperature under the hydrogen atmosphere (90 mTorr + base pressure). The schematic diagram of the graphene growth process is shown in Figure 2.2.    Figure 2.2 a) Schematic of graphene growth process; (b) photo of the freshly grown graphene taken from the furnace.   20  2.2.2 Graphene Transfer The as-grown graphene films were transferred to the SiO2 on Si substrates using a PMMA film as the sacrificial physical support layer. The substrates are undoped (001) Si wafer coated with 420 nm thick thermal oxide films. PMMA in chloroform solution (Microchem 950 PMMA C4) was spin cast on the graphene films with a spinning speed of 3500 rpm for 45 seconds, resulting in a 500 nm thick of film. The coated films were cured at 180 ºC for 1 minute in order to remove the solvent. After curing the stack was floated on the etchant solution, with the PMMA covered side of the graphene facing up. Graphene films covering both the top and the bottom of the copper foil were exfoliated during the etching process. Since the graphene on the bottom of the copper foil was not covered by the PMMA layer, it broke into small pieces and deposited at the bottom of the etching solution. After the entire copper foil is etched, the PMMA coated graphene was transferred using a plastic spoon into DI water bath for cleaning. After rinsing twice in DI water bath, the graphene films were scooped up with substrates and placed in a vacuum oven overnight to dry. Before the removal of PMMA, the graphene samples were heated at 180 ºC to for 30 minutes on a hot plate in air. In the end, PMMA was dissolved in a 50 ºC acetone bath for 2 hours.  The graphene samples were blown dry with Nitrogen after acetone and IPA rinsing. A schematic of the graphene growth and transfer process is shown in Figure 2.3.  21   Figure 2.3 a) Schematic of the graphene transfer process: graphene-coated Cu foils were spin-caoted with PMMA and floated on etchant solutions for Cu removal; the samples were rinsed with DI water twice before sccoping up with the substrates; PMMA was removed in acetone bath after the samples were dried in a vacuum oven overnight; b) photograph of a graphene sheet transferred to the SiO2/Si substrate.  The heat treatment helps to reduce the density of cracks in the transferred graphene. The copper foil usually has a high surface roughness, which will be further increased during annealing and graphene growth, due to surface reconstruction [70]. The as-grown graphene film follows the morphology of the copper foil. Small gaps will appear between graphene and substrate when the graphene films are transferred to the substrates. Cracks and wrinkles are therefore introduced to graphene when PMMA is removed. By applying the heat 22  treatment, the PMMA coated graphene films become more flexible. The gaps between the graphene and the substrate are reduced, and the density of mechanical defects is lowered. During the transfer process of graphene onto different substrates, it has been noticed that the mechanical and electrical quality of graphene sheets changes when using different copper etching solutions. Three choices of etchant are ferric chloride (FeCl3) solution, ammonia persulfate ((NH4)2(SO4)2) solution, and copper sulfate (CuSO4) in hydrochloric acid solution, in which Fe3+, (SO4)2 2-, and Cu2+ ions behave as oxidants, respectively, according to these reactions: 2𝐹𝑒3+ + 𝐶𝑢 →  𝐶𝑢2+ + 2𝐹𝑒2+      𝐶𝑢2+ + 𝐶𝑢 → 2𝐶𝑢+        (𝑆𝑂4)22− + 𝐶𝑢 →  𝐶𝑢2+ + 2𝑆𝑂42−      The as-grown graphene/copper stacks are subjected to wet etching using etchants with three different recipes: (1) FeCl3 aqueous solution: FeCl3 (0.2 M) in DI water, (2) CuSO4 in aqueous hydrochloric acid solution: CuSO4.5H2O (0.4 M) and HCl (6.1 M) in DI water, and (3) (NH4)2(SO4)2 aqueous solution (PSF): (NH4)2(SO4)2 (0.4 M) in DI water. It takes about 30 min to remove the 25 μm-thick Cu foils for all etchants.   2.3 Measurements and related background theory 2.3.1 Raman Spectroscopy Raman spectroscopy is a versatile tool to examine the properties of graphene. The Raman measurements were conducted with Horiba LabRam HR800 Raman system in the backscattering configuration with the excitation wavelength of 441.6 nm. The Raman spectra were collected through Olympus 50X (NA = 0.55) objective lens and recorded with 23  a grating of 2400 lines mm−1, resulting in a spectral resolution of 0.27 cm−1. The acquisition time and accumulation times were optimized with 6 s and 2 times (×2), respectively, to get a signal-to-noise ratio of more than 100 and avoid damages to the graphene samples due to a long time laser exposure. The 2D Raman mapping images were acquired using the following configurations: 20×20 μm were scanned for each sample; the step sizes were both 1 μm for X and Y direction; the laser spot size was ∼1 μm2.   2.3.2 Scattering-type scanning near-field optical microscopy (s-SNOM) Scanning near-field optical microscopy (SNOM) is an optical microscopy technique for investigating nanostructures, which causes minimal damages to the sample and exceeds the spatial resolution limit of the incident light wavelength.  The working principle of the scattering-type scanning near-field optical microscopy (s-SNOM) is illustrated in Figure 2.4. The metallic AFM tip is brought very close to the surface of the graphene samples. Incident infrared (IR) light is focused on the AFM tip, introducing a strong localized field around the tip apex, similar to the “lightening-rod effect” [71]. The scattered light caused by the nanoscale tip-sample coupling is detected in the form of elastic scattering, inelastic scattering, surface vibration, etc. In the graphene on SiO2/Si samples, the localized electric field can generate surface plasmons around the tip. The plasmonic wave propagates inside the graphene film. It is scattered back when hitting on defects or grain boundaries. The interference of the incident and reflective lights contains information about the defects distributions.     24   Figure 2.4 Schematic of working principle of the scattering-type scanning near-field optical microscopy (s-SNOM): surface plasmons that are induced by the tip and the subsequent interference due to the emission of scattered and reflected plasmonic waves.  The Anasys Nano IR-2 system was used in this work. The diameter of the tip used in the measurements was ~25nm, which allows the same level of spatial resolution in the imaging. CO2 laser covering the wavelength range of 9.5-11.3 µm was selected as the IR source. The topography of the samples was taken simultaneously with the plasmon interferometry.    2.3.3 Dielectric spectroscopy Dielectric spectroscopy measures various electrical properties of materials as a function of frequencies and temperatures [72]. In this work, dielectric spectroscopy measurements were conducted with a Novocontrol high-resolution broadband dielectric spectrometer. The schematic drawing of the circuit is shown in Figure 2.5 a).  25  The illustration of the samples prepared for the measurements was shown in Figure 2.5 b). Square shaped graphene films were transferred to SiO2 or Si substrates. Silver paste was applied on the two parallel sides of the film to form electrodes for a parallel plate capacitor. The input voltage was set to 0.01 V with the frequency varying between 103 Hz and 107 Hz. In this circuit, sample and reference were connected in series with two voltmeters measuring their voltage signals, respectively, and one current meter measuring the charging/discharging current signals.    Figure 2.5 a) Schematic of the circuit used in the dielectric spectroscopy; b) illustration of the graphene samples used for dielectric spectroscopy measurements.  Due to the serial connection between the sample and reference, the charge on the reference capacitor (Qr) must be equal to the charge on the sample capacitor (Qs). The complex impedance of the sample (Z*) can be calculated with the following equation: 𝒁∗(𝝎) =  𝒁′ + 𝒊𝒁" =  𝑽𝒔𝑰𝒔      Equation 2.1 where Vs and Is are the complex voltage and current measured on the sample. 26  As the external voltage applied to the circuit is time-dependent, the permittivity and conductivity we obtain are therefore complex numbers:  𝜺∗(𝝎) =  𝜺′ − 𝒊𝜺" =−𝒊𝝎𝒁∗(𝝎)𝟏𝑪𝟎    Equation 2.2 and   𝝈∗(𝝎) =  𝝈′ − 𝒊𝝈" =𝟏𝒁∗(𝝎)𝒅𝑨     Equation 2.3 where 𝜀′ and 𝜀" are the real and imaginary part of the complex permittivity, respectively; 𝜎′ and 𝜎" are the real and imaginary part of the complex conductivity, respectively; 𝐶0 =𝜀0𝐴𝑑 is the empty cell capacitance, ε0 is the vacuum permittivity, ω is the angular frequency, d is the plate separation, and A is the area of plate.  The area of the plate  𝐴 = 𝑡ℎ𝑖𝑐𝑘𝑛𝑒𝑠𝑠 𝑜𝑓 𝑔𝑟𝑎𝑝ℎ𝑒𝑛𝑒 × 𝑤𝑖𝑑𝑡ℎ 𝑜𝑓 𝑒𝑙𝑒𝑐𝑡𝑟𝑜𝑑𝑒 = 𝑡 × 𝑊 where W is the width of electrode, and t is the thickness of graphene. As graphene is a two-dimensional material, the thickness of graphene is negligible. Hence, the capacitance can be written as 𝑪𝟎 = 𝜺𝟎𝑨𝒅=  𝜺𝟎𝒕×𝑾𝒅= 𝜺𝟎−𝟐𝑫  𝑾𝒅    Equation 2.4 where 𝜀0−2𝐷 =  𝜀0 × 𝑡 = 8.85 × 10−14 𝐹 𝑐𝑚⁄ × 0.5 𝑛𝑚 = 4.43 × 10−21 𝐹 , assuming the thickness of graphene is 0.5 nm.   Subsequently,  𝝈∗(𝝎) =  𝝈′ − 𝒊𝝈" =𝟏𝒁∗(𝝎)𝒅𝑨=  𝟏𝒁∗(𝝎)𝒅𝑾    Equation 2.5 and  𝜺∗(𝝎) =  𝜺′ − 𝒊𝜺" =−𝒊𝝎𝒁∗(𝝎)𝟏𝜺𝟎−𝟐𝑫 𝒅𝑾     Equation 2.6 The real part of the complex permittivity still has the unit of 1, while the unit of the real part of the conductivity becomes siemens (S) or 1/Ω. 27   2.4 Results and Discussion 2.4.1 Morphology study Figure 2.6 and 2.7 demonstrate the images of the graphene samples transferred by CuSO4 (G-Cu), FeCl3 (G-Fe), and (NH4)2(SO4)2 (G-PSF) under an optical microscope, atomic force microscope (AFM), and scanning electron microscope (SEM).  As shown in Figure 2.6 a), d), and g), the G-Cu film is continuous and smooth in general. AFM measurements indicate a root-mean-square (RMS) roughness at 1.4±0.2 nm over the 5×5 µm measured area. The film consists of single-layer (green arrows in Figure 2.7) and bi-layer graphene (orange arrows in Figure 2.7), shown as the light grey and dark grey color in Figure 2.7 a) and d), respectively. Wrinkles and cracks (yellow arrows in Figure 2.7), as well as small amount of PMMA residues (light-blue circles in Figure 2.7), can also be observed in the optical and SEM images. It is believed that the microscopic line structures appeared in the optical images come from the mismatch between graphene and substrates during transfer. The nm-scale cracks and wrinkles are mainly due to the difference of thermal expansion between graphene and copper [73]. Unlike most materials, graphene has a negative thermal expansion coefficient. As a result, graphene and copper foil experienced opposite strain during heating and cooling, creating the cracks and wrinkles. The wrinkles usually have the height of less than 5 nm. It can be seen that G-Cu has a more uniform surface when compared to G-Fe and G-PSF layers.   28   Figure 2.6 Optical and AFM images of graphene sheets SiO2/Si substrates using different copper etchants: (a, d) optical and (g) AFM images of G-Cu films; (b, e) optical and (h) AFM images of G-Fe films; (c, f) optical and (i) AFM images of G-PSF films. Scale bar of all the AFM images is shown in the lower right corner of the figure.   G-Fe contains more microscopic defects due to ruptures and stacking of multiple layers (Figure 2.6 b and e). Consequently, more PMMA residue was left on the film due to the structural defects [9]. The RMS roughness of the G-Fe sample increased to 5.4±0.5 nm. The size of the PMMA residue was also measured with a height of 15 nm. The SEM images 29  (Figure 2.7 b and e) show water flow marks as well as nanometer-size dark particles (blue arrows in Figure 2.7) under the graphene film. The watermarks are originated from the water trapped between the PMMA/graphene stacks and the substrates [62]. The dark particles are believed to be residual metallic contamination introduced during the transfer process. Lupina et al. have reported that the copper surface concentration ranging from 1013 to 1015 atoms/cm2 has been observed in CVD-grown large area graphene samples regardless of either FeCl3 or PSF used for wet etching [63]. It is likely that some Cu atoms are isolated from the Cu foil due to the rough surface of Cu foil during high-temperature growth and the formation of graphene wrinkles. This part of the residual Cu may be “enclosed” into the graphene layer, and therefore inaccessible for Cu etchants [12, 74]. It was also noticed that in the optical microscopic images, more microscopic fragments and holes were found due to ruptures and stacking of multi-layers in the G-Fe films than in the other two types of samples. This phenomenon has also been reported before that graphene processed with Fe3+ etchant demonstrates one order of magnitude lower in fracture load than the ones processed with PSF [68]. The reason for the discrepancy in film quality for G-Cu and G-Fe is still not clear. However, previous research [75] has shown that chemical modification with divalent ions, such as Mg2+, can enhance the mechanical strength of the Graphene Oxide (GO) film. Although the detected amount of Cu2+ or Fe3+ ions is not significant in energy dispersive X-ray spectroscopy (EDX) measurement, it is suspected that in the etching solution, the divalent Cu2+ ions with similar ionic radius of Mg2+ (0.57 Å for both Mg2+ and Cu2+ ions with VI-coordinates) may give rise to a similar effect.   30   Figure 2.7 SEM images of the graphene samples using different copper etchants: (a, d) G-Cu films; (b, e) G-Fe films; (c, f) G-PSF films. Legends are listed on the right side of the figure. (SEM parameters: EHT 10.00 kV, working distance 19.5 mm, magnification 10.5 kX.)  G-PSF film is also continuous but has more microscopic features, compared to G-Cu film, as is shown in Figure 2.6 c) and f). The SEM images (Figure 2.6 c and f) further confirm this observation, showing that the G-PSF film contains more fine wrinkles. The persulfate anions can slowly react with water molecules in PH neutral solutions 2(𝑆𝑂4)22− +  2𝐻2𝑂 → 4𝐻𝑆𝑂4− +  𝑂2 releasing oxygen gas. It is observed that the generated oxygen gas forms small bubbles which are attached to the PMMA/graphene stacks in the bottom. These bubbles slow down the etching process and create deformation locally in the graphene films during wet etching and settlement of graphene on the substrates. Therefore, more wrinkles are generated in the G-PSF samples. The RMS roughness of G-PSF sample over the measured area is 1.4±0.2 nm, similar to G-Cu samples. 31   2.4.2 Raman spectroscopy   Figure 2.8 Raman spectra of a) G and b) 2D peaks of graphene transferred by CuSO4 (G-Cu), FeCl3 (G-Fe), and (NH4)2(SO4)2 (G-PSF).  The Raman spectra of graphene transferred with different etchants are shown in Figure 2.8. The peak positions of exfoliated single-layer graphene using 442 nm laser are 32  about 1583 cm-1 and 2725 cm-1, for G and 2D band respectively [76]. The G peaks of graphene samples prepared with all three etchants align well with the reported values and with each other, while the 2D peaks of all samples demonstrate a certain amount of blue shift from the original 2725 cm-1. The D peak (at 1380 cm-1) which is activated by defects occurs in all the samples, but with low intensity. The 𝐷 + 𝐷" peak (at ~ 2450 cm-1) which is also related to defected samples appears with low intensities as well. This indicates the existence of sub-micron defects in the graphene samples.    Figure 2.9 The histogram of a) c) the position distribution, and b) d) FWHM distribution of graphene transferred by CuSO4 (G-Cu), FeCl3 (G-Fe), and (NH4)2(SO4)2 (G-PSF).   33  Table 2.1 Average positions, FWHM, and integrated intensity of Raman peak of graphene samples. Parameters Unit G-Cu G-Fe G-PSF G-peak Position  [cm-1] 1585.5±2.2 1584.8±1.3 1583.4±1.0 FWHM  [cm-1] 19.0±1.5 21.9±1.3 17.0±1.2 Amplitude arb. unit 274.9 770.4 594.6 Integrated  intensity ×104 1.6 5.3 3.2 2D-peak Position  [cm-1] 2731.5±3.7 2731.6±2.5 2731.2±4.3 FWHM  [cm-1] 48.2±5.2 48.2±4.0 45.8±4.4 Amplitude arb. unit 503.4 942.4 605.6 Integrated  intensity ×104 7.6 14 8.7  2D-mapping of Raman spectra were performed on the graphene samples over a 20×20 µm area. Raman spectrum was recorded at every point in the measured area. The statistic results collected for the 400 points are presented in Figure 2.9 and Table 2.1. From Figure 2.9 a) and c), it can be observed that the peak position distribution of G-Fe is narrower than the other two films. The average positions of G peaks are found at 1585.5 cm-1, 1584.8 cm-1, and 1583.4 cm-1 for G-Cu, G-Fe, and G-PSF samples respectively. The average positions of the 2D peak are found at 2731.5 cm-1, 2731.6 cm-1, and 2731.2 cm-1, 34  for G-Cu, G-Fe, and G-PSF samples respectively. The blue shift of both G and 2D peaks suggests p-type doping of graphene [77]. Hole concentration in the range of 1013 cm-3 is expected corresponding to 5 cm-1 blue shift in 2D peak, which is consistent with the surface copper concentration level discussed previously.  Both G and 2D peaks demonstrate a certain level of broadening in nearly all the samples as is shown in Figure 2.9 b) and d). The full-width at half maximum (FWHM) of G peak for G-Cu, G-Fe, and G-PSF are 19, 22, and 17 cm-1, respectively. In comparison, the typical FWHM of doped graphene with G peak at 1585 cm-1 is about 14 cm-1.  G-PSF sample shows the lowest broadening level. However, it is believed that the shift of Fermi level in doped graphene results in the sharpening of G peak [77]. One of the possible reason for such phenomenon is the stacking of multi-layer graphene. Further investigation is needed to explain the peak broadening effect in our samples.  Figure 2.10 displays the mapping of intensity ratio between 2D (I2D) and G (IG) peak (a-c), and the ratio between D (ID) and G peak (d-f). A significant change in the shape and intensity of the 2D peak occur when the number of graphene layer increases. Typically, 𝐼2𝐷 𝐼𝐺⁄  ratio equals 2 or 1 indicate a monolayer graphene or bi-layer graphene respectively. Lower 𝐼2𝐷 𝐼𝐺⁄  ratio suggests for 3 layers or more [78]. The broadening of the 2D peaks also suggests the multi-layer stacking [79]. As is shown in Figure 2.10 a-c), the G-Cu sample has the most uniform distribution of 𝐼2𝐷 𝐼𝐺⁄  ratio at about 2, indicating the existence of single layer in most of the measured area. The G-Fe and G-PSF samples show similar results, with 𝐼2𝐷 𝐼𝐺⁄  ratio from 0.5 to 1.5 within the mapping areas. This result is consistent with the microscopic images that G-Cu has the smoothest surface, while more overlapping 35  grains and wrinkles, as well as the microscopic fragments appeared in the G-Fe and G-PSF samples.    Figure 2.10 2D mapping of graphene samples transferred with different copper etchants. a), b), and c) show the intensity ratio of 2D and G peaks; d), e), and f) show the intensity ratio of D and G peaks. Scale bars are shown on the right side of the figure.   Quantifying defects in graphene is critical for microelectronic applications. With a low density of defects and disorder, the D peak appears and 𝐼𝐷 𝐼𝐺⁄  ratio increases. The Tuinstra-Koenig relation [80, 81] can be employed to quantitively study the density of defects in carbon related materials. The model is used to describe edge defects [82] which is exactly the situation in our samples since most of the defects originate from grain boundaries, ruptures, and wrinkles. In the low defect-density regime,  36  𝑰𝑫𝑰𝑮=  𝑪′(𝝀)𝑳𝑫       Equation 2.7 where LD is the average inter-defect distance,  𝐶′(𝜆) =  (2.4 × 10−10 𝑛𝑚−3) ∙ 𝜆4, and λ is the wavelength of Raman laser [83]. Substituting λ= 442 nm and 𝐼𝐷 𝐼𝐺⁄ = 0.2 (for G-Cu film) into Equation 2.7, the average distance between line defects LD can be calculated to be ~45 nm. This indicates that the average size of graphene grains is higher or equal to 45 nm. For the other two samples with a higher average 𝐼𝐷 𝐼𝐺⁄  ratio approximately at 0.3, the density of defects can be higher with average distance between defects at about 30 nm.  2.4.3 s-SNOM test The s-SNOM images of the graphene films prepared with different etchants were initially recorded with the IR laser of 1047 cm-1 (9.6 µm in wavelength). The near-field amplitude spectrum of SiO2 has a resonant center at 1128 cm-1. By applying incident light close to the substrate resonant wavelength, the image of the graphene samples can be obtained from the backscattered light. Figure 2.11 displays the AFM topography, near-field amplitude and phase spectra of G-Cu, G-Fe, and G-PSF over an area of 5×5 µm.  All the samples demonstrate inhomogeneous optical and electronic properties. The near-field amplitude contrast is sensitive to the number of layers. As is shown in Figure 2.11, the different colors in the amplitude mapping images are correlated to the different reflectivity of the graphene layers. The G-Fe sample demonstrates the most non-uniform film as well as multi-layer stacking along different directions. The G-Cu and G-PSF samples have similar mapping images showing thicker and thinner regimes on the 37  continuous films with high contrast. The G-Fe and G-PSF images both contain areas with strong contrast. This phenomenon is likely because of local doping of metal ions.   Figure 2.11  Topography (a, d, g) and infrared near-field amplitude (b, e, h) and phase (c, f, i) images taken simultaneously at the wavenumber of 1047 cm-1 (~ 9.6 µm in wavelength).   To investigate grain boundaries and point defects in the graphene films, the plasmon interferometry images were taken with the incident light of 890 cm-1 where graphene and the SiO2 substrates have a low level of phonon coupling. Only the graphene samples of G-38  Cu and G-PSF with better quality and continuity were studied. The mechanically exfoliated graphene sheet was also included as the reference with large grains and low density of defects. The scattering amplitude and phase images are shown in Figure 2.12. For the exfoliated graphene, the scattering signal shows double fringes along the defect (grain boundaries, grain overlapping, or wrinkles). The bright circles observed in Figure 2.12 b) is the bi-layer region. The double fringes are formed by the standing wave produced by the emitted and reflected plasmons. The distance between the double fringes DDF can be calculated as [84]:  𝑫𝑫𝑭 ≈  −𝜹𝟐𝝅𝝀𝒑     Equation 2.8 where λp is the wavelength of plasmon and δ is the phase shift. This model is valid for grain boundaries and other line defects with geometry smaller than λp. The wrinkles and cracks appeared in the graphene films usually have sizes of a few nanometers. Therefore, Equation 2.8 can be applied to these defects too.  Similar double fringes were observed in G-Cu and G-PSF samples. It was clearly shown that the CVD-grown graphene has much smaller grains with a diameter less than 1 µm. The double fringe in both G-PSF and G-Cu samples has less separation than that in the exfoliated graphene. This is mainly because of the discontinuity of plasmon propagation caused by a high density of defects. The distance between fringes is measured as 50 nm and 80 nm in G-Cu and G-PSF, respectively. The plasmon wavelength is proportional to the Fermi energy of graphene: 𝝀𝒑 =  𝟐𝒆𝟐𝑬𝑭𝝀𝑰𝑹𝟐𝒉𝟐𝒄𝟐𝜺𝟎(𝟏+𝜺𝒓)     Equation 2.9 39  where e is the electron charge, 𝜆𝐼𝑅 is the wavelength of input infrared light, h is the Planck constant, c is the vacuum light speed, and 𝜀𝑟 is the real part of the complex permittivity of the substrate.  Therefore, the Fermi level of the graphene samples becomes lower when the fringe separation increases. G-Cu samples have lower Fermi energy than the G-PSF samples, suggesting a lower dopant level in the G-Cu samples.   40   Figure 2.12  Topography, surface plasmon interferometry images taken simultaneously at the wavenumber of 890 cm-1 (~11.3 µm in wavelength) of (d-i) G-Cu films and (j-o) G-PSF films. The images of the mechanically exfoliated graphene (a-c) were measured as a reference.  41  2.4.4 Dielectric spectra Dielectric spectroscopy measurements were performed on the graphene films prepared with different copper etchants. AC conductivity and permittivity of graphene films were investigated with the frequency range from 103 to 107 Hz at a various temperature from -50 ºC to 300 ºC. The input RMS voltage is set to be 0.01 V.   2.4.4.1 DC conductivity The DC conductivity of the samples was extracted from the frequency dependent AC conductivity at the low-frequency plateau region. Figure 2.13 shows the temperature dependent DC resistivity and dielectric constant of the graphene samples.  Graphene has zero bandgap at 0 K. When the temperature goes up, more electrons can be excited to higher levels, and the conductivity increases. The resistivity of graphene, which is the inverse of conductivity, increases with the rise of temperature. It indicates that graphene behaves more like metal rather than a semiconductor in the measured temperature range.  The G-Cu sample has the lowest conductivity over the entire temperature range, comparing to G-PSF and G-Fe samples. As discussed in the previous section, the Cu ions doping level in G-Fe and G-PSF samples are quite similar, while G-Fe has extra Fe3+ doping. Moreover, the Fermi level of G-PSF is higher than that of G-Cu, suggesting higher doping concentration in G-PSF than in G-Cu. Therefore, G-Cu has the lowest carrier concentration, and consequently the lowest conductivity.   42   Figure 2.13 Temperature-dependent DC conductivity and relative permittivity of graphene samples transferred by a) CuSO4 (G-Cu), b) FeCl3 (G-Fe), and c) (NH4)2(SO4)2 (G-PSF).  The conductivity of a metal or semiconductor follows the relation of: 𝝈 = 𝒏𝒆𝝁       Equation 2.10 where e is the charge of the electron, n is the density of conduction electrons, and µ is the mobility. The mobility describes the ability of charge carriers moving in the materials and can be written as 𝛍 =𝐞𝛕𝐦𝐞∗       Equation 2.11 43  where τ is the mean free time which is the average time between scattering events, and 𝑚𝑒∗  is the effective mass of the charge carriers. If multiple scattering mechanisms work on the conduction electrons at the same time, all the processes contribute to the overall resistivity of the material by reducing the mean scattering time.  𝟏𝝉=  ∑𝟏𝝉𝒊𝒊        Equation 2.12 Therefore, the resistivity can be written as  𝝆 =𝟏𝒏𝒆𝟐∑𝒎𝒆∗𝝉𝒊𝒊        Equation 2.13 The CVD graphene films are polycrystalline materials. The conductivity is affected by two major mechanisms: the inter-grain charge transport and the intra-grain transport. Theoretically, electrons in the perfect single crystalline graphene sheets without defects have ballistic propagation [85]: conduction electrons experience no scattering from the lattice. In the CVD-grown samples, the conduction electrons can be scattered by the interface between graphene and substrates, by point or line defects in the grain, and by the grain boundaries. The mean free time of conduction electrons in the grain is typically higher than that of the intra-grain transported ones since the scattering cross-section of the intra-grain defects is larger [86]. Therefore, the overall resistivity of the polycrystalline graphene samples can be attributed to different scattering mechanisms.  The resistivity of a metallic material ρ(T) usually follows a linear relation with temperature as: 𝝆(𝑻) =  𝝆𝟎(𝟏 + 𝜶(𝑻 − 𝑻𝟎))    Equation 2.14 44  where ρ0 is the resistivity at T0, and α is the temperature coefficient of resistivity. So, in the CVD-grown samples, the resistivity shows multiple linear regions with increasing temperatures. There are several characteristic temperatures regimes in the plots of DC resistivity vs. temperature, marked as I, II, III and IV in Figure 2.13.  First, a slight increase in conductivity is observed for G-Fe and G-PSF films from -50 to 10 ºC (Regime I of both samples). As T continues to increase, the conductivity for G-Fe and G-PSF starts to decrease. However, the resistivity of G-Cu sample barely changed in that regime (I). To understand this phenomenon, previous studies by Huang et al. [87] on GO with dielectric spectroscopy suggests that this is mainly due to the “defrosting” of water molecules and its intercalation effect. Similarly, G-PSF and G-Fe samples both show lower 𝐼2𝐷 𝐼𝐺⁄  ratio, corresponded to multi-layer stacking of graphene sheets. Therefore, it is likely for G-Fe and G-PSF films to have water molecules intercalated between the substrates and graphene films, within the multi-layer stacking, and in the wrinkles during the wet transfer process. These intercalated molecules are activated while the temperature goes above 0 ºC, contributing extra carriers to the graphene films and therefore elevating the conductivity. On the other hand, the G-Cu sample with mostly monolayer and less wrinkles have fewer intercalated water molecules. Hence, no clear conductivity increase has been noticed at this temperature range.   For G-Cu sample, the resistivity starts to increase from about 40 ºC to about 240 ºC and then saturates at higher temperatures. In this temperature range, the resistivity first follows a power law (in zone II) and then increases more linearly. For the G-PSF films, the resistivity follows a two-step linear increment from 40 ºC to 300 ºC whereas the resistivity 45  of the G-Fe sample exhibits a three-step increase in the same temperature range. The reasons behind the different increasing patterns could be the effect of several mechanisms working at the same time. Since the G-PSF and G-Fe samples have the strong feature of multi-layer graphene stack, additional scattering from the multiple layers with different stacking orientation is likely to be one of the factors to consider [79]. The enhanced amount of overlapping grains and wrinkles in G-PSF and G-Fe can lead to several linear increment regimes. Further investigations on the cause of different temperature dependency need to be performed.  The real permittivity at the low frequency 1 kHz of the graphene samples is also plotted in Figure 2.13. The dielectric constant of all three samples shows the same trend: a slow increasing from -50 to 138 ºC, followed by a fast increasing with higher temperatures. G-Fe and G-PSF both show a peak permittivity at about 240 ºC (242 ºC for G-Fe and 230 ºC for G-PSF), while the permittivity of G-Cu sample slowly plateaued from the same temperature. The initial increase of permittivity is due to the vibration of the lattice, leading to higher scattering of charge carriers. The fast increasing and later on decreasing of dielectric constant is suspected to be related to PMMA residue. The PMMA has a melting temperature at about 140 ºC, and boiling temperature at about 200 ºC [88]. When the temperature increases, the PMMA residue melt slowly and interact with the graphene film creating more diploes at the interface. Later on, at higher temperatures, PMMA is gradually removed, causing a decrease of dipoles. As discussed in the previous sections, G-Cu films may have less residue since the surface is smoother than the other two samples. Therefore, less PMMA is left on the films, causing less change of permittivity at high temperatures.  46  2.4.4.2 AC conductivity Frequency-dependent ac conductivity 𝜎′  follows the expression of the power law [89]: 𝝈′(𝝎) =  𝝈𝑫𝑪 + 𝑨𝝎𝒔     Equation 2.15 where 𝜎𝐷𝐶 is the dc conductivity, A is the pre-exponential factor depending on temperature, and s is the frequency exponent generally in the range of 0-1. There is a critical frequency 𝜔𝑐 = 2𝜋𝑓𝑐 , beyond which the power law is valid. Therefore, Equation 2.15 can also be written as: 𝝈′(𝝎) 𝝈𝑫𝑪⁄ =  𝟏 + (𝝎 𝝎𝒄⁄ )𝑵    Equation 2.16 where N is the frequency exponent. Figure 2.14-2.16 shows the AC conductivity of the G-Cu, G-Fe, and G-PSF samples at different temperatures. The AC conductivity plotted are normalized to the DC conductivity at the according temperature.  All the samples demonstrate similar behaviors on the AC conductivity. As the frequency increases, the conductivity demonstrates a two-stage increment both following the power law at lower temperatures, with the first critical frequency 𝑓𝑐1 at 104 -105 Hz, and the second critical frequency 𝑓𝑐2 at about 107 Hz. The plateaued conductivity value between the first and second stages (105 -106 Hz) decreases with increasing temperatures. When the temperature reaches about 150 ºC (~136 ºC for G-Cu, ~158 ºC for G-Fe and G-PSF), the conductivity at 105 -106 Hz stops from decreasing. With temperatures higher than that, the conductivity at the same frequency range increases with increasing temperatures. The associated critical frequency for the first stage also decreases with increasing temperatures before it increases.   47   Figure 2.14 Frequency dependence of the real part of the complex electrical conductivity of graphene transferred by CuSO4 at a series of temperatures: a) -50 ºC to 95 ºC; b) 95 ºC to 150 ºC; c) 150 ºC to 190 ºC; d) 190 ºC to 300 ºC.  The origin of AC conductivity comes from the drifting of charge carriers with the external electric field. When an AC field is applied, the charge carriers can be drifted back and forth when the electric field flips directions. The drift velocity of charge carriers is again proportional to the mobility. As discussed in the previous section, the mobility of electrons or holes in our graphene samples is controlled by two dominant mechanisms: the scattering from inter-grain and intra-grain transport. The inter-grain transport charges have 48  lower effective mass and longer mean free time, thus higher mobility. Therefore, when the frequency increases, the charge carriers move within each grain respond to the external field faster. Moreover, the charges are slowed down when reaching the grain boundaries. Therefore, the AC conductivity exhibits a two-stage increment. When temperature increase, the thermal vibration of the lattice becomes stronger. The scattering cross-section becomes larger, causing a decrease in the conductivity at higher frequency ranges.  When the temperature reaches the critical temperature at 150 ºC and higher, it is suspected that the graphene lattice experiences a relaxation, leading to less scattering and higher conductivity. Graphene has a negative thermal expansion coefficient [90], which means that the graphene films shrink with increasing temperatures. The density of the existing line defects, overlapping of grains and wrinkles is likely to be reduced as the graphene films shrink and become flatter. The scattering cross-section is therefore reduced, leading to lower scattering rate and an increased conductivity. G-Fe and G-PSF samples demonstrate a higher critical temperature than G-Cu sample due to the higher density of such defects. More thermal energy is required for the grains in G-Fe and G-PSF samples to reach the same level of relaxation. Therefore, the AC conductivity of these two samples shows higher critical temperatures.    49   Figure 2.15 Frequency dependence of the real part of the complex electrical conductivity of graphene transferred by FeCl3 at a series of temperatures: a) -50 ºC to 95 ºC; b) 95 ºC to 150 ºC; c) 150 ºC to 190 ºC; d) 190 ºC to 300 ºC.    50   Figure 2.16 Frequency dependence of the real part of the complex electrical conductivity of graphene transferred by (NH4)2(SO4)2 at a series of temperatures: a) -50 ºC to 95 ºC; b) 95 ºC to 150 ºC; c) 150 ºC to 190 ºC; d) 190 ºC to 300 ºC.  2.4.4.3 Permittivity The relative permittivity of the graphene samples was measured in the frequency range of 103 -107 Hz, where the space charge and dipolar polarization dominates the permittivity behavior.  The frequency dependency of the real part of the complex permittivity 𝜀𝑟′  follows the relation of:  𝜺𝒓′ (𝝎) = 𝟏 + 𝑵𝜶𝜺𝟎(𝟏+𝝎𝟐𝝉𝟐)     Equation 2.17 51  where N is the density of atoms per unit volume, α is a constant called the polarizability, 𝜀0 is the vacuum permittivity, and τ is the relaxation time dependent on temperature. The according frequency 𝑓0 = 1 𝜏⁄  marks the frequency when the dipoles fail to respond to the external field.    Figure 2.17 Frequency dependence of the real part of the complex permittivity of graphene transferred by CuSO4 at a series of temperatures: a) -50 ºC to 95 ºC; b) 95 ºC to 150 ºC; c) 150 ºC to 190 ºC; d) 190 ºC to 300 ºC.  52   Figure 2.18 Frequency dependence of the real part of the complex permittivity of graphene transferred by FeCl3 at a series of temperatures: a) -50 ºC to 95 ºC; b) 95 ºC to 150 ºC; c) 150 ºC to 190 ºC; d) 190 ºC to 300 ºC.  Fig. 2.17-19 shows the real part of the complex permittivity of the graphene samples. Again, all three samples demonstrate similar frequency dependency at a series of temperatures. At the lower temperature range, the permittivity at low-frequency increases, and the relaxation frequency decreases at the same time. This is again due to the thermal vibration of the lattice which prevents the space charges and the existing dipoles to align with the external electric field. 53    Figure 2.19 Frequency dependence of the real part of the complex permittivity of graphene transferred by (NH4)2(SO4)2 at a series of temperatures: a) -50 ºC to 95 ºC; b) 95 ºC to 150 ºC; c) 150 ºC to 190 ºC; d) 190 ºC to 300 ºC.  As the temperature increases, the permittivity at low frequency keeps increasing, while the relaxation frequency starts to increase when the temperature goes up. When the temperature reaches more than 240 ºC, the lower frequency permittivity of G-Fe and G-PSF starts to decrease. The trend of permittivity fluctuation at low frequency is interpreted in the previous section.  54  By fitting the permittivity vs. frequency curves with equation 2.17, the relaxation frequency at different temperatures is extrapolated and plotted in Figure 2.20. It is noticed that when the temperature goes up from -50 ºC to 80 ºC, the relaxation frequency first decreases and maintains the same value until the temperature reaches about 200 ºC. The relaxation frequency shows a rapid increase with higher temperatures. It is believed that the reason behind the rapid frequency increase is related to the relaxation of grains. As discussed in the previous section, the density of the line defects and wrinkles on the graphene films is likely to be reduced when the temperature goes higher. Therefore, the space charges and dipoles have higher flexibility to adapt to the external electric field, leading to higher relaxation frequency.    Figure 2.20 Temperature dependency relaxation frequency of graphene sheets prepared with different copper etchants.   55  2.5 Conclusion In summary, the graphene films grown with CVD on copper foil and transferred to SiO2/Si substrates with wet etching process was investigated. Three different etchants were used during the transfer. The G-Cu samples demonstrate a uniform and continuous single layer; the G-PSF samples are continuous but with multiple layers; the G-Fe samples have more ruptures and overlapping fragments. All three samples demonstrate a certain level of p-type doping. The grain boundaries and wrinkles on the graphene films were studied. The estimated length between defects is about 45 nm for G-Cu sample and lower for the other two samples. The conductivity and permittivity of the graphene films were studied with dielectric spectroscopy. All the samples display a similar trend of AC conductivity and permittivity at the temperature range of -50 ºC to 300 ºC. G-Cu sample shows the lowest DC conductivity due to its fewer layers and lower doping level. Several factors contribute to the difference of the electrical properties, including the PMMA residue, the density of line defects and grain boundaries, and the water molecule intercalation. The underlying mechanism leading to the different trending of DC conductivity vs. temperature of the three types of graphene films is still inconclusive. Further study in the charge transport mechanisms in polycrystalline graphene is needed.    56  Chapter 3: Graphene and graphene/charge-selective layer hybrid transparent conductors  3.1 Introduction and motivation Transparent conductors are important in the new generation industries of solar cells, touch panels, and displays. Indium tin oxide (ITO) is currently the dominant option for most of the commercially available electronic products, providing a benchmark combination of transparency (> 90%) and low sheet resistance (< 100 Ω/square) at the same time. However, due to the rapidly growing demands of the market as well as the high-temperature vacuum fabrication process, the cost of ITO has been rising significantly. Also, the brittleness of ITO limits its application in the next generation of flexible electronic products.  As a two-dimensional material, graphene is naturally a good candidate of the electrode in optoelectronic devices due to its high conductivity and transparency in ultra-violet (UV) and visible wavelengths. In order to replace ITO with graphene, a key issue is to reduce the sheet resistance of graphene. The reported lowest sheet resistance of graphene is about 30 Ohm/square for multiple layer graphene, and most of the literature demonstrates that the sheet resistance values of single-layer graphene (SLG) are still several hundred or thousand Ohms, which is too high for graphene to be used as a transparent electrode. Due to its ultra-low thickness and high surface-to-volume ratio, graphene is very sensitive to all kinds of surface modifications. The properties of graphene can be easily tuned by interfacial doping or coupling from the adjacent layers or substrates.  57  Surface transfer doping is one of the methods to manipulate the conductivity of graphene.  The doping is achieved by carrier exchange between graphene and the materials which are absorbed or deposited on the surface of graphene [91-93]. Charge transfer is determined by the relative position of the Fermi level of graphene and the HOMO/ LUMO levels of the adjacent materials. The doping leads to a shift in Fermi level, and increases the carrier concentration and therefore the conductivity of SLG layers.  Similarly, the electrical properties of graphene can be largely affected by the substrates. It has been reported that the suspended graphene has high electron mobility over 200,000 cm2V-1S-1 [94]. However, the mobility of exfoliated graphene films placed on the SiO2 wafer can be reduced by one to two orders of magnitude [95-97]. Electrons in actual devices experience strong impurity scattering mainly due to the interactions of graphene with the underlying substrate, which limits the electron mean free path. Surface traps, interfacial phonons, and fabrication residues on top of or underneath the graphene sheets may contribute to the loss of mobility too. Many studies of using graphene as transparent conductors have been reported in the past decades. Multilayer graphene grown on nickel foil usually has higher conductivity than single-layer graphene grown on copper. The sheet resistance of multilayer graphene samples can be reduced to the range of 200-300 Ω/square [98, 99] with the cost of low transmittance. Stacking of several single layer graphene can improve the conductivity too [100], but the complicated transfer process would bring more defects and contaminations.   Chemical doping is another commonly used methods to enhance the conductivity of graphene. Typical dopants include AuCl3, HNO3, HCl, SOCl2, TCNQ, FeCl3, etc. It has been reported that AuCl3, HNO3, and FeCl3 can reduce the sheet resistance of graphene up 58  to 80%, to less than 200 Ω/square [101-103]. By applying multiple layers stacking and chemical doping together, Bae et al. presented 4-layer graphene with sheet resistance as low as 30 Ω/square [14]. However, most of the dopants are unstable in the air so the doping effect would gradually fade away over a few weeks or even shorter [42, 43 67]. In addition, the dopants can be washed away by solvents, which brings difficulties for device fabrication as the deposition of other solution-processed materials on top of the doped graphene would eliminate the doping effect [101]. More exploration of graphene conductors with high and stable conductivity and transmittance is still needed.  Typical organic photovoltaic devices and organic light emitting diodes employ a structure that can be summarized as: the active layer which absorbs or emits light is sandwiched between the anode and cathode; one of them needs to have a certain level of transparency in order to allow photons passing through; one or several interfacial layers are inserted between the active materials and the electrodes, to help charge transfer and selection. Commonly used materials for interfacial layers include poly(3,4-ethylenedioxythiophene) polystyrene sulfonate (PEDOT:PSS), nanocrystalline MoO3, and ZnO. In order to incorporate graphene into optoelectronic devices, it is essential to understand how graphene is coupled with different substrates as well as the charge selective layers. Subsequent deposition of the commonly required charge selective materials for OPV device on the graphene sheets not only help fill up the ruptures that are generated during the transfer process but also affect the conductivity of graphene. If SLG/charge selective material hybrid conductors are applied as transparent electrodes in OPV devices, the carrier transport materials can work in both conduction and charge selection. We will 59  be able to eliminate unnecessary interfaces, i.e., unexpected interfacial reactions with graphene, simplify the manufacturing process of OPV devices, and reduce the total cost eventually.  3.2 Sample preparation and measurements 3.2.1 Sample preparation As demonstrated in the previous chapter, single layer graphene was grown with LPCVD using Cu foil (25 µm in thickness) as the catalyst. The as-grown graphene sheets were then transferred to substrates using PMMA as the sacrificial layer. To fabricate OPV devices with reliable and optimal performance, the transparent electrodes are preferred to have a lower density of defects. Therefore, CuSO4 in hydrochloride acid solution was selected as the etchant to remove the Cu foil.  Various types of materials have been employed as substrates in the study, including glass, polyimide (PI), Polyethylene terephthalate (PET), and Polyethylene-naphthalate (PEN), which are commonly used in organic device fabrication.  The PMMA-coated graphene on substrates samples was then left in a vacuum chamber for 24 hours to remove the leftover water completely. The samples were heated in the ambient atmosphere at 110 ºC to 180 ºC for 30 min. After that, the PMMA sacrificial layer was removed in an acetone bath at 50 ºC for 2 hours. The graphene coated substrates were rinsed with acetone, IPA, and DI water, and dried using a nitrogen blow gun.  The graphene samples were then coated with different charge selective materials to modify its surface and electrical properties. Detailed treating procedures will be introduced in the according sections.  60   3.2.2 Measurement techniques To characterize the performance of graphene as a transparent conductor, sheet resistance, transparency over visible and UV wavelengths, as well as the flexibility of each type of graphene samples were tested.    Figure 3.1 Schematic of samples during I-V measurements  Limited by the size of the graphene growth chamber, the graphene sheet used in this research were all in a square shape with the size of approximately 1 cm by 1 cm. Due to the complicated growth and transfer process, as shown in the previous chapter, it is very common to have cracks and wrinkles as well as PMMA residue on the graphene sheets. Therefore, the commonly used four-point probe or van der Pauw technique is not useful in characterizing the conductivity of the graphene samples. Similar methods were used in the previous chapter, by applying two long electrodes on both ends of the graphene sheet, as shown in Figure 3.1. Current-voltage (I-V) characteristics were measured with a computer-controlled Keithley 2400 Source Meter. The sheet resistance of graphene can be calculated as:  61  𝑹𝒔 = 𝑹𝑾𝑳=𝑽𝑰𝑾𝑳    Equation 3.1 where R is the resistance, V and I are the voltage and current obtained in the I-V test, and W and L are the width and length of the graphene sheet.  A xenon lamp (150 W, Newport Co.) equipped with a monochromator (Cornerstone 130, Newport Co.) was employed as the light source for transparency test. The optical power was measured using a power/energy meter (Newport Co.). Scanning electron microscopic (SEM) were recorded using Zeiss Sigma FE-SEM, while atomic force microscope (AFM) images were measured using Asylum Research MFP 3D and Bruker Dimension Icon system. Film thicknesses were measured with a Bruker Dektak XT profilometer.    Figure 3.2 Schematic of the bending test apparatus: the two ends of a sample are held by one fixed arm and one bending arm which is controlled by a stepper motor. The center of the sample is aligned to a 0.2 mm diameter thick wire and is bent over the wire during the test. (Adapted from Ref. [104])   62  The conductivity of the samples was investigated during a two-step test, to study the flexibility of the graphene conductors. A bending test apparatus was designed and manufactured as shown in Figure 3.2. The sample (shown in orange color) is held between a fixed holding arm and a rotating arm which is driven by a stepper motor. A molybdenum wire of diameter 0.2 mm is held by two fixed metal arms in-aligned with the stepper motor’s rotary center. A wire is connected to each electrode of the graphene sheet, to conduct resistance test while bending.  I-V characteristics of the samples were first measured while the samples were bent at a series of angles, from 0 to 130 degrees, to obtain the change of resistance versus radius of curvatures.  Later on, the samples went through a continuous bending process to investigate the stability of the transparent conductors during bending. The bending arm was set to rotate by 100 degrees, and back to its original position as one full bending cycle, the bending rate was set to 25 cycles per minute. The change of sheet resistance was recorded to characterize the stability of the transparent conductors.   3.3 Graphene sheets on different substrates Substrate selection is one of the key elements for flexible electronic devices. Typical requirements include good mechanical strength, low surface roughness, high transparency, high thermal stability, low moisture and oxygen permeability, etc. Polyimide (PI, Kapton® Polyimide Film), polyethylene terephthalate (PET, Teijin® Tetoron® film), and polyethylene naphthalate (PEN, Teonex®) are widely used polymer films in flexible electronic applications. Detailed technical properties of these polymer films, as well as glass slides, are listed in Table 3.1. The transmittance of each substrate at 550 nm and the 63  surface roughness were measured in our lab, while other parameters are exacted from Ref. [105, 106].  Table 3.1 Technical properties of glass and flexible substrates [105, 106].  All the flexible substrates employed in this research have the same thickness of 50 µm. The AFM images of flexible substrates, PI, PET, and PEN, over an area of 50×50 µm, are demonstrated in Figure 3.3 a), b), and c).  The root-mean-square roughness of the substrates are calculated to be 25±2, 8.4±0.8, and 7.5±0.8 nm over the entire measured area, respectively for PI, PET, and PEN. The surface of PI substrates is significantly rougher than the other two substrates.  Property Unit Glass PI PET PEN Thickness μm 500 50 50 50 Transmittance @550nm % 93 31 92 87 Surface roughness nm 7.2±0.8 25±2 8.4±0.8 7.5±0.8 Glass transition temperature °C - 360 110 155 Melting point °C >1000 - 258 269 Thermal expansion coefficient ppm/K 3 17 22 21 Water absorption % - 1.3 0.4 0.3 Tensile strength MPa N/A 280 230 220 Tensile elongation % N/A 80 120 120 64   Figure 3.3 Morphology and transmittance of glass and the flexible substrates: AFM images of a) PI, b) PET, and c) PEN substrates over an area of 50×50 µm; d) optical transmittance of substrates at wavelength 400-800 nm.  Up to 70% of the solar energy that reaches the ground of earth lies into the UV and visible light range [107]. To maximize the power conversion efficiency of OPV devices, the substrates need to have high transparency level at the same wavelength range. As shown in Figure 3.3 d), PET has almost the same transparency level as glass covering more than 90% over 400 to 800 nm wavelength range. PEN has slightly lower transparency level with 65  80% to 90% over the measured range. PI shows high absorption in the short wavelength range with only ~30% transparency at 550 nm. Therefore, PI is not good for OPV applications. However, since the substrate transparency of OLEDs is not as critical as OPV devices, PI can be used for OLEDs as long as the emission light does not overlap with the absorption spectrum of PI.  Heat treatment is a commonly used approach to improve the performance of the functional layers in OLED and OPV devices. Therefore, it is essential to choose the substrates with relatively high glass transition and melting temperatures to give enough flexibility for material curing and annealing. PET and PEN have glass transition temperature at 110 °C and 155 °C, respectively. Although PI has low transparency, especially in the green and blue light region, it has much higher glass transition point which allows heat treatment at high temperatures. Therefore, PI is also considered as a candidate for the substrate in our study.   3.3.1 Graphene layer transfer As described in the previous chapter, PMMA residue on the graphene sheets is difficult to remove and has a strong impact on the properties of graphene. The heat treatment in ambient atmosphere for 30 min of PMMA-coated graphene on substrates samples is used to soften the PMMA layer and to help dissolve PMMA in the following step. Typical PMMA has a melting temperature of 160 ºC. Therefore, the annealing temperature is set to 165 ºC for graphene sheets on PI, 110 ºC for ones on PET, and 140 ºC for PEN.  66  It is also noticed during the transfer process that graphene shows stronger adhesion to plastic substrates than to glass. This phenomenon is possible because plastics tend to have stronger hydrophobicity than glass or other metal oxides. Further fabrication process or treatments would lead to less impact/damage on graphene sheets placed on plastic substrates. Figure 3.4 shows the surface morphology of graphene samples on different substrates. Graphene sheets placed on plastic substrates show fewer wrinkles than the ones on glass slides. The samples with higher baking temperature before PMMA removal also demonstrate lower density of defects and cracks. The root-mean-square roughness for the 20×20 µm measured area is 13.8±1.8 nm for glass, 14.3±2.1 nm for PI, 13.5±1.9 nm for PET, and 11.2±1.5 nm for PEN. The surface roughness of all samples is in the same range with each other, despite the difference of roughness of the substrates. This is mainly due to the existence of PMMA and copper residue left on the graphene sheets, which averages out the original roughness on the substrates.   67   Figure 3.4 Optical and AFM images of the graphene sheets on different substrates. Scale bar of the AFM images is shown on the right side of l).  68  3.3.2 Sheet resistance and transparency The transmittance and sheet resistance measured on samples of graphene on different plastic substrates are listed in Table 3.2. As discussed in the previous section, the density of defects, as well as the leftover residue between graphene sheets and the substrates, can be reduced by using plastic substrates. Therefore, the sheet resistance of graphene sheets transferred to plastic substrates was lower than the ones placed on glass slides. Since the graphene sheets were transferred using the scooping technique piece by piece, the sheet resistance of each piece of graphene shows significant variation. It is difficult and also unreliable to calculate the average resistance of the graphene sheets. According to the measurements of more than 200 pieces of graphene sheets on all substrates, it is fair to claim that the sheet resistance of graphene transferred to plastic films can be as low as 700 Ohms, in comparison of the lower limit of ~1000 Ohms of the graphene transferred to a glass slide.   Table 3.2 Sheet resistance and transmittance at 550 nm of SLG on various substrates. The transmittance shown here is the average value of more than 50 samples for each type of substrate. The transmittance is calculated with respect to the power of incident light. Property Unit Glass PI PET PEN Sheet resistance of the best sample Ω/sqr 1033 690 752 714 Transmittance @550nm % 90 27 90 85  69  The transmittance of single-layer graphene on different substrates in the visible range is shown in Figure 3.5. Graphene shows a uniform absorption of about 3% of the incident light across the whole measured range. The average transmittance of graphene on plastic substrates at 550 nm is 27%, 90%, and 85%, respectively.    Figure 3.5 Transmittance of graphene sheets on different substrates. The transmittance is calculated with respect to the power of incident light.   3.3.3 Flexibility To investigate the flexibility of graphene on plastic samples, the conductivity measurements were performed during a two-step bending test. The results are shown in Figure 3.6. When the samples were bent up to 130 degrees, the sheet resistance increases with increasing bending angles, for all the samples. The graphene sheets on PI substrates experienced higher strain since PI has a higher stiffness than the other two plastic films. Therefore, the graphene on PI samples shows a 70  higher increment of sheet resistance than the other two type of samples. Figure 3 b) demonstrates the stability of graphene under long-time continuous bending. One bending cycle was defined as the sample is bent from 0 to 100 degree and back to original position. The rate was set to be 25 cycles per minute. All the three series of samples demonstrate high stability during the test. The sheet resistance oscillated at about 2% of its original value for SLG/PI, and at about 1.5% for SLG/PET and SLG/PEN samples. No significant relaxation has been noticed during the test, which indicates that the coupling between the graphene sheets and the substrates are consistent and stable.     Figure 3.6 Flexibility of graphene samples on different substrates: a) change of sheet resistance with bending angle up to 130 degrees; b) change of resistance during continuous bending up to 800 bending cycles. The original values of the three plots in b) are around 0. The curves are plotted with a vertical offset for clarity.   71  3.4 Hole-selective graphene transparent conductors Charge selection materials are essential for high-performance optoelectronic devices. In this section, two of the most commonly used hole-selective materials: semiconducting polymer PEDOT:PSS and inorganic MoO3 are assembled with graphene sheets and studied. Since both PEDOT:PSS and MoO3 require annealing temperature less than 150 ºC after coating,  only PET and PEN substrates with high transparency level were used in this research.  3.4.1 PEDOT:PSS coated graphene transparent conductors 3.4.1.1 Fabrication Filtered (0.45 µm) PEDOS:PSS aqueous solution (Heraeus Clevios P VP Al 4083) was mixed with IPA (3:1 vol ratio), followed by vigorous stirring at room temperature for 24 hours. The mixture was spun cast on the graphene film. The thickness of the PEDOT:PSS layer can be adjusted by altering the spinning rate and time. The coated films were then heated in ambient condition for 15 min at up to 140 ºC when PEN was used as substrates and at 110 ºC for PET.  Optical microscopic, SEM, and AFM images of the PEDOT:PSS coated SLG samples are shown in Figure 3.7. PEDOT:PSS of 25 nm thickness is selected as the example here for both substrates. It is seen that the PEDOT:PSS forms a continuous and uniform layer on the graphene sheet, without clusters or defects. The peaks and valleys occurred in the AFM image (Figure 3.7 e) are originated from the substrates and the PMMA residue. The RMS roughness of the PEDOT:PSS/SLG/PEN stack is 7.5±0.5 nm, in the 20×20 µm measured area, which is significantly reduced from the non-coated ones. This indicates that the solution processed PEDOT:PSS film fills up the unevenness of as-transferred graphene surface, providing a smoother surface which can be beneficial for future device fabrications. 72   Figure 3.7 Optical, SEM, and AFM images of 25 nm thick PEDOT:PSS coated graphene sheets on PET (a, b) and PEN (c, d, and e). (SEM parameters: EHT 10.00 kV, working distance 19.5 mm, magnification 10.5 kX.)  3.4.1.2 Results Figure 3.8 shows the sheet resistance and transmittance of the PEDOT:PSS/SLG transparent conductors on PET substrates. A series of PEDOT:PSS samples with thicknesses ranging from 20 73  to 43 nm were fabricated. All the PEDOT::PSS layers were annealed at 110 ºC for 15 min after spin-coating. The plotted results are averaged for ten samples fabricated with a given condition. It is noted that the sheet resistance of the samples can be reduced by more than 20% by incorporating 25 nm of PEDOT:PSS film. The sheet resistance of the stack further decreases with thicker PEDOT:PSS film. Meanwhile, the transmittance of the samples demonstrates a decrease of ~4% with 20 nm of PEDOT:PSS film. The transmittance of samples at 550 nm is plotted for comparison of optical properties. The transmittance of all the samples maintains at the same level or demonstrates a slow decrease when the thicker film was coated. The transmittance values are calculated using the exit light power after the substrate as the reference, to provide a fair comparison with samples made with other substrates.    Figure 3.8 a) Reduction of sheet resistance and b) transmittance at 550 nm of PEDOT:PSS coated graphene on PET samples. The transmittance of all the samples is calculated with respect to the light after substrate absorption.  74  Figure 3.9 shows the sheet resistance and transmittance of the PEDOT:PSS/SLG transparent conductors on PEN substrates. Similarly, a series of samples with different PEDOT:PSS thicknesses were fabricated. The typical annealing temperature of PEDOT:PSS in OPV device fabrication is 140 ºC. Since the glass transition temperature of PEN is higher than that, samples with 25 nm thick film annealed for the various duration and at various temperatures were also fabricated to study the influence of curing conditions. PEDOT:PSS layer with a fixed thickness of 25 nm while annealed at a series of temperatures between 95 to 140 ºC for 15 min. Samples that annealed at 140 ºC for a variable time from 0 to 15 min were fabricated and measured.  Similar to the PEDOT:PSS/SLG conductors on PET, the PEN samples also demonstrate up to 40% reduction in sheet resistance with 53 nm thick PEDOS:PSS layer. The resistance becomes lower when thicker PEDOT:PSS layer is incorporated. Nevertheless, as the PEDOT:PSS polymer has low conductivity of 10-3 S/cm [108] (corresponding to several MΩ for 20-50 nm thick layer), the decrease of resistance cannot be entirely attributed to the addition of a parallel conductive film. The electrical coupling between PEDOT:PSS and graphene should have contributions to the overall change in the resistance. HOMO level of PEDOT:PSS is at -5.2 eV, lower than the work function of graphene at 4.5 eV. It is possible that the holes in the PEDOT:PSS layer is drifted to graphene because of the valence band potential difference between the two layers. The carrier concentration in graphene is increased, resulting in lower resistivity.   75   Figure 3.9 Reduction of sheet resistance and transmittance at 550 nm of graphene on PEN with PEDOT:PSS coating with: a, d) different thickness; b, e) annealing time; c, f) annealing temperature. The transmittance of all the samples is calculated with respect to the light after substrate absorption.  The annealing process of PEDOT:PSS also has some effect on the sheet resistance. It can be noticed in Figure 3.9 b) and c) that the sheet resistance was further reduced when the polymer film was cured at a higher temperature and over a longer period. PEDOT:PSS is an aqueous suspension. The excessive water in the film can be removed by annealing, leading to better recrystallization of the polymer with increased conductivity.  The transmittance at 550 nm of the PEDOT:PSS/graphene stack on PEN shows 2-4% of decreasing with PEDOT:PSS up to 53 nm. The transmittance was also enhanced when the 76  PEDOT:PSS layer was cured for a longer time at a higher temperature. The density of defects is reduced when the PEDOT:PSS film is cured. Therefore, the incident photons are likely to experience less scattering and absorption, leading to the observed higher transmission.   Figure 3.10 and 3.11 show the results for a flexibility test of the PEDOT:PSS/graphene conductors on PEN and PET substrates. Figure 3.10 a), b), and c) illustrate the normalized change of resistance of PEDOT:PSS/graphene/PEN samples with respect to bending angles. The sheet resistance of all samples increases with higher bending angles. When the samples were bent over a 0.2mm thin wire, the sheet resistance increases to less than 4% up to 130 degrees. The increase in the resistance is lower at the beginning and becomes large in magnitude when thicker PEDOT:PSS layer is coated. It is also noticed that the changes in sheet resistance reduce when the PEDOT:PSS layer is cured at a higher temperature and for a longer period, which is consistent with the previous interpretation of polymer recrystallization. The normalized change of resistance of samples under continuous bending up to 1000 cycles was plotted in Figure 3.10 d) and e). All the samples with proper annealing demonstrate high stability over a long time bending, with up to 4.5% shift of resistivity. Almost all the samples demonstrate a certain level of relaxation at the initial stage of the bending experiment before entering the steady state. On the other hand, the samples which were annealed at lower temperatures show a higher change of resistance (10%) and prolonged relaxation. The sheet resistance at the resting position keeps shifting, especially for the samples with a lowest annealing temperature of 95 ºC.  Since the PEDOT:PSS/graphene stack on PET substrates were all cured with lower temperature, the samples show higher shifts of resistance when they are bent up to 130 degrees, compared to the samples on PEN with the same PEDOT:PSS thickness. All the samples display a 77  less than 4% change of resistance during the continuous bending process, indicating good flexibility.   Figure 3.10 Flexibility of PEDOT:PSS coated graphene samples on PEN. a-c) the increment of sheet resistance when the PEDOT:PSS coated graphene samples were bent up to 130 degree; d-e) change of sheet resistance of PEDOT:PSS coated graphene samples during continuous bending. The original values of the curves in d) and e) are around 0. The curves are plotted with a vertical offset for clarity.  78   Figure 3.11 Flexibility of PEDOT:PSS coated graphene samples on PET. a) The increment of sheet resistance when the PEDOT:PSS coated graphene samples were bent up to 130 degree; b) change of sheet resistance of PEDOT:PSS coated graphene samples during continuous bending. The original values of the curves in b) are around 0. The curves are plotted with a vertical offset for clarity.  3.4.2 MoO3 coated graphene transparent conductors MoO3 is a widely used metal oxide for hole selection. The work function and ionization affinity of MoO3 thin films can change with different environmental conditions [109, 110]. Typically, the work function of ~5.3 eV can be achieved when MoO3 is exposed to air for more than one hour. Since many of organic semiconductors have similar HOMO levels, MoO3 is, therefore, a good hole transport material for many organic devices.   3.4.2.1 Fabrication The MoO3 thin film was deposited on the graphene sheet using thermal evaporation. The deposition was conducted in a vacuum chamber at a vacuum level of 2×10-8 Torr. Subsequently, 79  the MoO3 thin film was annealed in an ambient atmosphere for 10 min to enhance the mechanical strength. A series of samples with a different MoO3 thickness from 2 nm to 40 nm was prepared. The curing temperature was set to 110 ºC for PET samples and 140 ºC for PEN samples.   Optical microscopic, SEM, and AFM images of the MoO3 coated SLG samples are shown in Figure 3.12. MoO3 of 20 nm thickness is selected as the example here for both substrates. The MoO3 layer is uniformly distributed on the graphene sheets, without any sign of nucleation islands. The RMS roughness of the 20×20 µm area is 11.4±1.9 nm, similar to the roughness before MoO3 coating.   80   Figure 3.12 Optical, SEM, and AFM images of 20 nm MoO3 coated graphene sheets on PET (a, b) and PEN (c, d, and e). SEM parameters: EHT 10.00 kV, working distance 19.5 mm, magnification 10.5 kX.   81  3.4.2.2 Results Figure 3.13 shows the sheet resistance and transmittance of the MoO3/SLG transparent conductors on PEN and PET substrates. The samples illustrate a more than 35% reduction in resistance with only 2 nm of MoO3. When the thickness of MoO3 increases, the resistance slowly decreases to about 55% of the non-coated value. When the MoO3 film is more than 10 nm thick, it is noted that the thickness of the MoO3 layer does not affect the relative change of resistance. A similar effect has also been reported by Meyer et al. [111]. It is believed that interface traps are generated between MoO3 and graphene, leading to charge transport between the two layers, and thus increases the conductivity of graphene.  The transmittance of the MoO3/graphene stack increases slightly with 2-4 nm MoO3 and decreases with thicker MoO3 layers. The transmission at 550 nm can be reduced from ~97% (single layer graphene) to 94% with 20 nm MoO3, and to 80% with 40 nm MoO3 layer. It has been observed that 2 nm of thermal deposited MoO3 form discrete islands rather than a continuous film on 100 nm thick gold film [112]. The increase in transparency with 2 nm MoO3 can be interpreted as the MoO3 islands acting as absorbing centers which concentrate the incoming photons and transport them into the graphene sheets.   82   Figure 3.13 Reduction of sheet resistance and transmittance at 550 nm of graphene on PEN (a, b) and PET (c, d) with MoO3 film.   Figure 3.14 shows the flexibility test results of the MoO3/graphene conductors on PEN and PET substrates. Almost all the samples demonstrate the similar or even lower shift of resistance than the non-coated samples, except for the 40 nm thick MoO3 sample. The MoO3/graphene conductors show less than 4% increase in resistance when bent up to 130 degrees on both PET and PEN substrates. Meanwhile, as shown in Figure 3.14 b) and d), all the samples show very high stability after a long time bending with less than 2% shifting of resistance and almost no relaxation.    83   Figure 3.14 Flexibility of MoO3 coated graphene samples on PEN (a, b) and PET (c, d). a, c) the increment of sheet resistance when the MoO3 coated graphene samples were bent up to 130 degree; b, d) change of sheet resistance of MoO3 coated graphene samples during continuous bending. The original values of the curves in b) and d) are around 0. The curves are plotted with a vertical offset for clarity.   84  However, unlike in the PEDOT:PSS/graphene samples, cracks were observed after a long time bending in the MoO3/graphene conductors, as shown in Figure 3.15. This is because of the brittle nature of metal oxides. Since Young’s modulus of MoO3 is one magnitude higher than PEDOT:PSS [113, 114], it is reasonable to believe that the MoO3 layer was damaged by bending. However, the existence of cracks has a limited effect on the sheet resistance, which further proves that the reduction of sheet resistance does not come from charge transport within the MoO3 layer.   Figure 3.15 AFM images of MoO3 coated graphene samples on PET after bending.   Figure 3.16 is a schematic diagram of the graphene sheet coated with hole/electron transport materials during bending. By incorporating a layer on top of graphene, the position of the neutral plane is also shifted towards graphene. The strain that graphene bears during bending are reduced with both soft PEDOT:PSS films or brittle MoO3 films. PEDOT:PSS does not break during bending, therefore, the strain generated in both PEDOT:PSS layer and the graphene layer affects the resistivity of the transparent conductors. While in the case of MoO3 film, the cracks released the strain in the MoO3 layer outside of graphene. Only the strain existed inside of the graphene layer influences the resistivity. This could be the reason why the MoO3/graphene samples experience less relaxation and lower resistance oscillation.  85   Figure 3.16 Schematic of graphene coated with charge selective materials during bending.  3.4.3 Summary Adding a layer of hole transport materials helps improve the conductivity of graphene. Both PEDOT:PSS/graphene and MoO3/graphene samples show a decrease in sheet resistance up 40% while maintaining similar transparency level. The transmittance and sheet resistance data of PEDOT:PSS/graphene and MoO3/graphene samples from this work were plotted in Figure 3.17 a) to compare with other conventional transparent conductors. The combination of sheet resistance and transmittance of our transparent conductor is comparable with ITO at the high resistance range. Figure 3.17 b) shows the transmittance curve of the hole-selective graphene conductors over the 400-800 nm range. The transmittance of ITO and single layer graphene are included as a reference. The PEDOT:PSS/graphene stack samples demonstrate higher transparency than ITO in most of the wavelengths. MoO3/graphene samples only show higher absorption than ITO in the short wavelength region, which can be beneficial for UV protection purposes. Usually, transparent conductors with less than 100 Ohm can be considered as the potential candidate for photovoltaic 86  applications. However, the hole-selective transparent conductors developed in this work may still be useful in touch panels or displays industry.      Figure 3.17 a) Transmittance vs. sheet resistance plot of SLG, PEDOT:PSS/SLG, and MoO3/SLG stacks on PEN from this work (transmittance are calculated relative to PEN), compared with typical transparent conductors. ITO data is adapted from Ref. [22], the data of large-area multi-layer graphene is adapted from Ref. [11], and the data of conductive PEDOT:PSS with additives is adapted from Ref. [115]. b) Transmittance of SLG, PEDOT:PSS/SLG, and MoO3/SLG stacks on PEN from this work (transmittance are calculated relative to PEN) compared to ITO (~ 60 nm in thickness) over 400-800 nm wavelengths. 87  In addition, both of the PEDOT:PSS/graphene and MoO3/graphene conductors show high flexibility, with less than 4% change in conductivity under 130-degree bending. The samples also demonstrate strong stability during long time continuous bending. Therefore, both of the hole-selective graphene conductors present potential to be applied as flexible transparent electrodes in flexible electronics.  3.5 Electron-selective graphene transparent conductors Materials with high work function are required for electron transport and injection. Ca thin film, ZnO, and TiOx are the most commonly used electron transport materials. Ca is an active metal and can be oxidized easily. Therefore, it is not a good option for developing durable devices. In this work, ZnO is employed for the fabrication of electron-selective conductors. ZnO nanoparticles can be prepared by two methods: the hydrolysis method and the sol-gel method. Both synthetic methods require high-temperature treatment to help improve the properties of ZnO. Therefore, only PI with high melting temperature was used in this part of the research.   3.5.1 Fabrication 3.5.1.1 ZnO colloidal nanoparticles as electron-selective materials The hydrolysis preparation was adapted from Ref. [116] by Pacholski et al. ZnO colloidal nanoparticles were synthesized by hydrolysis and condensation of zinc acetate dihydrate with KOH in methanol. By controlling the time of condensation, we managed to have the average diameter of the particles to be ~10-15 nm. The colloidal particles were rinsed with methanol (MeOH). Chloroform, chlorobenzene (CB), and methanol were added into the particles to reach a concentration of ~8 mg/ml.  88  After the nanoparticle synthesis, the suspension ink was spin cast on the graphene sheets on PI. The layer was heated on a hot plate in ambient condition after spin-coating at 250 ºC for solvent removal and grain growth.   Figure 3.18 Optical and SEM images of ZnO colloidal nanoparticles coated graphene sheets on PI. (SEM parameters: EHT 10.00 kV, working distance 19.5 mm, magnification 10.5 kX.)  Optical microscopic and SEM images of ZnO nanoparticles coated graphene samples on PI are shown in Figure 3.18. Suspension inks with all three solvents form a continuous and uniform layer of ZnO on graphene. However, the layers formed using chloroform and chlorobenzene inks contain a significantly higher concentration of cavities and cracks. This is possibly due to the high evaporation rate of chloroform and chlorobenzene, not providing enough time for ZnO particles to settle during the drying process. In comparison, ZnO suspension ink with methanol forms a smoother surface with tiny cavities and no cracks.   89  3.5.1.2 ZnO films prepared with the sol-gel method as electron-selective materials The sol-gel preparation of ZnO films was adapted from Ref. [117]. One Molar of zinc acetate dehydrates [Zn(CH3COO)2·2H2O] (Sigma-Aldrich) was dissolved in 2-methoxyethanol [C3H8O2] (2-ME), isopropanol (IPA), or methanol. One molar percentage of ethanolamine [C2H7NO] (Sigma-Aldrich) was added as a stabilizer. The solutions were mixed at 60 °C until all the solids were fully dissolved (~30 min) in ambient atmosphere.  The as-prepared ZnO precursor inks were spin cast on graphene sheets at 3000 rpm for 10 seconds followed by immediate thermal treatment at 250 °C for 5 minutes. The Zn ions react with oxygen in the air and form ZnO nanoparticles. Each coating and annealing cycle results in a ~25 nm thick ZnO layer.   Figure 3.19 Optical and SEM images of ZnO sol-gel film coated graphene sheets on PI using a, d) 2-ME, b, e) IPA, c, f) MeOH as solvents. (SEM parameters: EHT 10.00 kV, working distance 19.5 mm, magnification 10.5 kX.) 90  Graphene exhibits strong hydrophobicity during the spin-coating process of the precursor inks. Therefore, 2-ME and IPA, both with a higher melting point, cannot form a uniform film during annealing or even spin-coating. Methanol with lower melting point helps achieve fast drying and ZnO formation. As is shown in the images of Figure 3.19, 2-ME and IPA precursor inks form discrete islands on the PI substrates. Only the ink with MeOH forms a uniform layer of ZnO. However, the SEM images show that there still exists a high density of small cavities all over the film.    3.5.2 Results Sheet resistance and transmittance measurements were performed on the three samples with continuous colloidal nanoparticle layers and the sol-gel ZnO with MeOH layer. The results are shown in Figure 3.20 a) and b). Only the sol-gel ZnO layer exhibits about 20% reduction in resistance. Although the conduction band energy of ZnO is -4.3 eV, higher than the Fermi level of graphene (-4.5 eV), the intrinsic defects exist in ZnO shift its work function to the higher value. Therefore, the electrons in the ZnO cannot transport to graphene effectively due to the energy barrier in the interface, which causes the high resistance of ZnO/graphene stacks.  Due to the high band gap energy of ZnO, almost all the samples demonstrate higher transparency than ITO by more than 90% within 400-800 nm wavelengths. Graphene-coated ZnO nanoparticle layer with chloroform solvent shows lower transmittance, as the photons get scattered in the cavities of the film.  91   Figure 3.20  Sheet resistance and transmittance of ZnO thin film coated graphene on PI (transmittances are calculated relative to PI), compared with typical transparent conductors. ITO data is adapted from Ref. [22], the data of large-area multi-layer graphene is adapted from Ref. [11], and the data of conductive PEDOT:PSS with additives is adapted from Ref. [115].     Figure 3.21 Flexibility of ZnO coated graphene samples on PI: a) change of sheet resistance of ZnO coated graphene samples during continuous bending. The original values of the curves are around 0. The curves are plotted with a vertical offset for clarity; b) the increment of sheet resistance when the MoO3 coated graphene samples were bent up to 130 degree.  92  The samples coated with ZnO nanoparticles using methanol solvent and samples coated with the sol-gel ZnO in methanol were studied in the bending test. Both samples demonstrate about 2% change of resistance while bending to 130 degrees, which is comparable to the non-coated graphene sheets on PI. However, both ZnO coated samples experience a high level of instability in the continuous bending test. The resistance at resting position increases up to 15% during 700 cycles of bending. It is suspected that the ZnO layer created ruptures on graphene resulting in the increasing resistance. The electron-selective transparent conductors made of ZnO coated graphene do not show superior performance in terms of conductivity and flexibility. Therefore, it may not be a good option for future device fabrications.  3.6 Conclusion In summary, charge selective graphene transparent conductors were fabricated by integrating graphene with hole/electron-selective materials. The hole-selective conductors using PEDOT:PSS and MoO3 films both exhibit improved performances in the combination of conductivity, transparency, and flexibility. These conductors can potentially be used as electrodes in fabricating flexible optoelectronic devices. On the other hand, the integration of ZnO nanoparticles with graphene to fabricate electron-selective conductors was not as successful. The samples did not show significant improvement compared to the non-coated graphene. Other materials that have better alignment with the Fermi level of graphene will be required for future studies.      93  Chapter 4: Flexible organic solar cells with hole-selective graphene transparent electrodes  4.1 Introduction and motivation Owing to the growing demands on light-weighted mobile electronics with better durability, portable energy harvesting devices draw intensive attention in the past decades. Organic photovoltaic devices have attracted great interest from both academic and industrial researchers due to the low-cost fabrication process, scalability, and the potential application in flexible electronics [118, 119]. ITO has been the dominant choice of the transparent electrode. However, the continuous increasing prices of ITO due to the limited supply of indium and high demands in the market affect the total cost of the OPV devices greatly [120]. Furthermore, the brittleness of ITO limits its applications in flexible electronic devices. Alternative transparent electrodes, such as conductive polymers, carbon-based materials, metal nanostructures, and multi-layer indium-free oxides, have been studied and developed.  Graphene has been proposed as a promising candidate for ITO replacement due to its high transparency, mechanical and chemical stability. CVD-grown graphene which is compatible with roll-to-roll process provides the opportunity of potentially low-cost mass production of the flexible devices [14, 40]. Recent studies have shown significant improvements in solar cells using graphene electrodes [121-123]. OPV devices using multi-layer graphene with chemical treatments as transparent electrodes have demonstrated PCE of 2.6% with P3HT:PCBM active film [124]. PCE of 7% has been achieved on devices with PTB7:PCBM film and electrodes made of stacking 94  multiple single-layer graphene films [100]. However, single layer graphene with exceptional transparency suffers from its low conductivity for fabricating OPV devices with high PCE.  As is discussed in the previous chapter, single layer graphene coated with hole-selective materials presents improved conductivity while maintaining the same level of transparency. The work presented in this chapter is an extension of the previous one. In this chapter, OPV devices using such conductors as transparent conducting electrodes (TCE) were fabricated and investigated. The hole-selective graphene electrodes demonstrate acceptable performance compared to the ITO electrode. PCE of 2.1 % and 1.5 % have reached by devices fabricated on rigid and flexible substrates. The devices using modified graphene electrodes have shown strong mechanical robustness under bending up to 130 degrees, corresponding to a radius curvature of less than 2mm. The results indicate that the hole-selective graphene transparent conductors can be a promising replacement of ITO on flexible solar harvesting devices.   4.2 Sample preparation 4.2.1 Device structure and fabrication For OPV device fabrication, P3HT and PCBM (Solaris, 1:1 wt ratio dissolved in dichlorobenzene) are used as an active polymer layer. Figure 4.1 shows the band alignment of possibly-included layers in the devices. Patterned graphene sheets were transferred to glass slides or PEN films. PEDOT:PSS or MoO3 films were deposited subsequently to form the hybrid transparent conductors. A second layer of hole transport film may be added for better charge separation.  Single layer graphene was grown with LPCVD using Cu foil (25 µm in thickness) as the catalyst. The as-grown graphene sheets were then transferred to substrates using PMMA as a 95  sacrificial layer and CuSO4 in hydrochloride acid solution as Cu etchant. PMMA/graphene stacks were transferred to a glass slide or PEN films with scooping technique. PMMA was consequently removed in an acetone bath at 50 ºC for 2 hours. The graphene coated substrates were rinsed with acetone, IPA, and DI water before dried using a nitrogen blow gun.  Filtered (0.45 µm) PEDOS:PSS aqueous solution (Heraeus Clevios P VP Al 4083) was mixed with IPA (3:1 vol ratio), followed by vigorous stirring at room temperature for 24 hours. The mixture solution was spin cast on the graphene film. The coated films were heated on a hot plate at 140 ºC for 15 min in air. The MoO3 thin film was deposited on the graphene sheet using thermal evaporation. The deposition was conducted in a vacuum chamber at a vacuum level of 2×10-8 Torr. Subsequently, the MoO3 thin film was annealed at 140 ºC in the air for 10 min to enhance the mechanical strength.  For the spin-coated samples, the bulk heterojunction films were deposited from a 1:1 solution of PC61BM and P3HT (regioregularity 95%, Solaris Chem Inc.) in 1,2-dichlorobenzene (1,2-DCB; 99%, Sigma-Aldrich) with the concentration of 40 mg/ml. The blend ink was spin-cast on the hole transport layers with a sequential rate of 250 rpm (3 s), 1000 rpm (15 s) and 1500 rpm (3 s) resulting in a 200 nm thick layer. The films were placed in a covered Petri dish immediately after coating to control the solvent evaporation rate (drying time: 10-20 min). In the end, Ca (20 nm) and Ag (100 nm) films were thermally evaporated through a shadow mask as the electron transport layer and cathode at a vacuum level of 2×10-8 Torr. The device area 0.2 cm2 is defined by overlapping the top and bottom electrodes.  96   Figure 4.1 (a) Band alignment of layers in the OPV devices; (b) schematic OPV devices structures; the overlapped area between the top and bottom contact is the effective device area.  The sprayed coated heterojunction films were deposited using Exacta Coat automatic spray-coating system equipped with a Sonotek Accumist ultrasonic atomizing nozzle. PC61BM and P3HT (1:1 weight ratio) was dissolved in a blend of 1,2-DCB and mesitylene (7:3 volume ratio) with the concentration of 8 mg/ml. The film was sprayed for a single run at room temperature. Fabrication parameters were set to achieve about 200 nm thick polymer layers [125]. The as-sprayed films were rest for 1 min and blow dried using the same spray nozzle. The bulk heterojunction films were cured at 140 ºC for 5 min. Similar to the spin-coated samples, Ca (20 97  nm) and Ag (100 nm) films were thermally evaporated through a shadow mask as the electron transport layer and cathode.  4.2.2 Measurements Current-voltage (I-V) measurements of devices were conducted on a computer-controlled Keithley 2400 Source Meter while the samples were kept in a home-built vacuum holder after transferring out of the glove box. A xenon lamp (150 W, Newport Co.) equipped with an AM1.5G filter was employed as the light source. The optical power was set to 100 mW/cm2, calibrated using a broadband power meter (Newport Co.). The external quantum efficiency (EQE) was measured using monochromatic light from a monochromator (Cornerstone 130, Newport Co.) and the incident power was measured using a power/energy meter (Newport Co.). Scanning electron microscopic (SEM) and atomic force microscope (AFM) images were recorded using Zeiss Sigma FE-SEM and Asylum Research MFP 3D respectively. Film thicknesses were measured with a Bruker Dektak XT profilometer.   4.3 Experimental results 4.3.1 OPV devices with hole-selective graphene TCE on glass substrates OPV devices were initially fabricated on glass substrates in order to study the performance of the hole-selective graphene TCEs. Spin-coating was used to prepare samples for all the glass substrate. Figure 4.2 a) shows the front and back sides of the OPV devices. The green boxes mark the graphene covered areas. Both PEDOT:PSS/graphene and MoO3/graphene transparent conductors were used as the anode in the devices. Optical microscopic images of the photoactive layers were taken in both reflected and transmitted mode, as is shown in Figure 4.2 b) and c). The 98  surface of the polymer layer showed a high degree of roughness due to the slow drying process of the P3HT:PCBM layer.    Figure 4.2 a) Front and back sides of OPV devices using hole-selective graphene as TCE; the green box marks the graphene covered area. b) Reflective and c) transmitted optical images of OPV devices using hole-selective graphene as TCE on glass substrates.  It has been observed that MoO3 can be washed away or damaged during the spin-coating of the PEDOT:PSS layer or the P3HT:PCBM layer especially when the MoO3 layer was thin and without proper annealing. It has been demonstrated in the previous chapter that the thickness of MoO3 has minimum effect on the conductivity and flexibility, but can significantly reduce transparency if it is over 30 nm. Therefore, 20 nm thick MoO3 was deposited for all the samples using MoO3/graphene as transparent electrodes. The as-deposited MoO3 layer was annealed at 150 ºC for 10 min to enhance its mechanical strength further. The MoO3/graphene TCEs treated with such methods showed no sign of structural damages during the following device fabrication procedures. The thickness of PEDOT:PSS was set to 42 nm for all the devices using 99  PEDOT:PSS/graphene transparent conductors as an anode. The sheet resistance for both hole-selective conductors lies in the range of ~500 Ohm/sqr, with about 92% transparency at 550 nm.  The reference device using single-layer graphene on ITO was fabricated in order to achieve both high conductivity and similar surface conditions as the hole-selective graphene conductor electrodes. The detailed sample list and the essential PV parameters are shown in Table 4.1.   Table 4.1 Essential parameters of the OPV devices using hole-selective graphene TCE on glass substrates. Electrode (anode) 2nd HTL VOC [V] JSC [mA/cm2] FF [%] PCE  [%] ITO/SLG PEDOT:PSS (25nm) 0.57 10.4 50.0 3.0 SLG/PEDOT:PSS -- 0.52 6.5 27.9 0.94 SLG/PEDOT:PSS MoO3 (20 nm) 0.56 8.7 43.0 2.1 SLG/MoO3 -- 0.54 3.6 29.0 0.57 SLG/MoO3 PEDOT:PSS (25 nm) 0.55 8.1 36.7 1.7  Figure 4.3 illustrates the current density-voltage (J-V) characteristics and the external quantum efficiency of the solar cells. All the samples using graphene anodes demonstrate lower short-circuit current than the graphene/ITO anode sample. This phenomenon is due to the higher series resistance in the graphene TCE samples than it in ITO samples (~30 Ohm/sq of ITO). The 100  samples with both PEDOT:PSS/graphene and MoO3/graphene TCEs show low fill factors. When a second hole transport layer was added, the short-circuit current of devices using both TCEs increased to more than 8 mA/cm2 with the open-circuit voltage of 0.56 V. This is due to more effective charge separation and transport at the interface of the active layer and electrodes. The samples with PEDOT:PSS/graphene TCEs demonstrate higher fill factor of 43% and a higher PCE of 2.1%.    Figure 4.3 a) Current density-voltage characteristics and b) external quantum efficiency of OPV devices using hole-selective graphene TCE on glass substrates.  4.3.2 Spin-coated OPV devices with hole-selective graphene TCE on PEN substrates The OPV devices made on glass slides demonstrate acceptable performance. Therefore, similar devices were fabricated with PEN substrates to study the flexibility behavior. Spin-coating was used for fabrication of the bulk heterojunction film of all the devices.  Figure 4.4 a) shows the top view of the active layer after coating. The PEN substrates were held by vacuum during spinning. Because of the low stiffness of PEN, the vacuum forced the 101  substrates to deform, resulting in uneven distribution of photoactive layers. As graphene only covers the bottom half of the substrates, the area that was away from the active device area were placed on the vacuum holder while spinning so that the working area of the solar cells can achieve more uniform films.      Figure 4.4 a) Spin-coated P3HT:PCBM films on hole-selective graphene TCE on PEN substrates. b) Front and c) back sides of OPV devices using hole-selective graphene as TCE; the green box marks the graphene covered area. d) Reflective and e) transmitted optical images of OPV devices using hole-selective graphene as TCE on PEN.  The OPV devices are listed in Table 4.2, together with the essential PV parameters. J-V and EQE characteristics are shown in Figure 4.5. The solar cells fabricated on PEN showed similar behavior as the ones made on rigid substrates. The samples without the second hole transport layer 102  have lower short-circuit current and lower open-circuit voltage due to charge recombination. The samples with the same devices structure demonstrate lower PCE on PEN than on glass, which is likely the consequences of the P3HT:PCBM film being non-uniform. PCE of 1.5 % and 1.4 % was achieved with the solar cells using the PEDOT:PSS/graphene and MoO3/graphene TCEs, respectively.   Table 4.2  Essential parameters of the OPV devices with spin-cast active layers using hole-selective graphene TCE on PEN. Electrode (anode) 2nd HTL VOC [V] JSC [mA/cm2] FF [%] PCE [%] SLG/PEDOT:PSS -- 0.51 5.6 30.4 0.87 SLG/PEDOT:PSS MoO3  (20 nm) 0.56 6.7 39.5 1.5 SLG/MoO3 -- 0.49 4.7 27.7 0.63 SLG/MoO3 PEDOT:PSS (25 nm) 0.56 6.9 35.4 1.4  103   Figure 4.5 a) Current density-voltage characteristics and b) external quantum efficiency of OPV devices with spin-cast active layers using hole-selective graphene TCE on PEN.  In addition, the sample devices have demonstrated good flexibility. The same bending apparatus used in the previous chapter were employed (please see section 3.2.2 for details). The OPV devices were held between the two metal arms. The bending arm was rotated to bend the sample to a certain angle over a ~0.2 mm wire. I-V measurements were performed while the sample was bent. Figure 4.6 illustrates the PV parameters of samples while bending up to 130 degrees. The devices have demonstrated high robustness under mechanical deformation. Samples with both PEDOT:PSS/graphene and MoO3/graphene as TCE show less than 5% degradation during bending. It indicates that our devices have great potential for low-cost roll-to-roll processing as well as for wearable electronics.      104   Figure 4.6 Plots of essential parameters vs. bending angles of the OPV devices with spin-cast active layers using hole-selective graphene TCE on PEN.  4.3.3 Spray-coated OPV devices with hole-selective graphene TCE on PEN substrates Spray coating technique was employed to address the non-uniformity issue caused by spin-coating. Figure 4.7 presents the images of the sprayed active polymer film. Similar to the process of spin-coating, the substrates were held by vacuum during the spray with smaller holes and lower pressure. The as-prepared films have demonstrated better uniformity than the spin-coated films.    105   Figure 4.7 a) Sprayed P3HT:PCBM films on hole-selective graphene TCE on PEN substrates before and after annealing. b) Reflective and c) transmitted optical images of OPV devices using hole-selective graphene as TCE on PEN.  Table 4.3  Essential parameters of the OPV devices with spray coated active layers using hole-selective graphene TCE on PEN. Electrode (anode) 2nd HTL VOC [V] JSC [mA/cm2] FF [%] PCE  [%] SLG/ PEDOT:PSS MoO3 (20 nm) 0.56 6.3 42.2 1.4 SLG/MoO3 PEDOT:PSS (25 nm) 0.56 6.6 37.9 1.5  Two series of samples were fabricated, as listed in Table 4.3. J-V and EQE characteristics were presented in Figure 4.8. The spray-coated samples exhibit similar J-V performance as the spin-coated samples. The PCE of 1.4% and 1.5% was obtained for the solar cells using the PEDOT:PSS/graphene and MoO3/graphene TCEs, respectively. Since the P3HT:PCBM layer was 106  formed from a lower concentration solution with different solvents, the spraying parameters and drying techniques that lead to optimal PV performance were still under exploration [125]. This is likely the reason why the sprayed samples can achieve more uniform active films while maintaining the same level of PCE as the spin-coated ones.   Figure 4.8 a) Current density-voltage characteristics and b) external quantum efficiency of OPV devices with spray coated active layers using hole-selective graphene TCE on PEN.  Bending test was performed on the sprayed devices as well. The samples once again present strong mechanical stability. Both of the samples have PCE of 1.4% while bent up to 130 degrees as shown in Figure 4.9.   107   Figure 4.9 Plots of essential parameters vs. bending angles of the OPV devices with spray coated active layers using hole-selective graphene TCE on PEN.  4.4 Conclusion In conclusion, flexible organic solar cells using hole-selective material coated graphene transparent conducting electrodes were fabricated and investigated. The devices fabricated on glass substrates show comparable PV performance with the ITO-based samples. Spin-coating of the active polymer materials on flexible substrates could cause uneven distribution and therefore affect the PCE of the devices. Spray-coating can be used to address the issue. The working parameters of the spray process need to be further studied to fabricate solar cells with optimal PCE. The flexible OPV devices exhibit a PCE of 1.5% regardless of the bending conditions. Such properties indicate a bright future of the application of single layer hole-selective graphene transparent conductors in a variety of flexible electronic applications.     108  Chapter 5: Light-soaking Free Organic Photovoltaic Devices with Sol-gel Deposited ZnO and AZO Electron Transport Layers  5.1 Introduction and motivation Organic photovoltaic (OPV) devices have drawn great attention in the past decade due to its potential for economical and efficient energy-harvesting applications. With a certified champion power conversion efficiency (PCE) of over 11.2% [126], solar cells with organic semiconducting light-absorbing layers are approaching closer to real life applications. Devices using polymers as the active layer can be fabricated using low-cost solution processing methods in ambient conditions and amenable to large-scale manufacturing.  Normal OPV devices use transparent Indium-tin oxide (ITO) electrodes as the anode and organic poly(3,4-ethylenedioxythiophene)-polystyrene sulfonate (PEDOT:PSS) as the anode buffer layer. However, PEDOT:PSS is corrosive to ITO due to its acidic nature and may decompose under long-term UV exposure [127, 128]. Therefore, OPV devices with normal structure are not ideal for robust long-term solar energy harvesting. Moreover, high work function metals that are used as cathode interfacial layer are sensitive to oxygen and humidity in the surrounding atmosphere. Thorough encapsulation and packaging become necessary to stop these devices from fast degradation.  On the other hand, OPV devices with an inverted structure can avoid the incorporation of unstable materials and have long-term stability. Inverted polymer solar cells with PCE of over 9% [129] has been reported. Metal oxide materials, e.g., ZnO and TiOx, are commonly used as an electron transport layer (ETL) deposited on the transparent electrode for inverted OPV devices. 109  Among all the metal oxides, zinc oxide (ZnO) stands out for its high electron mobility, high visible transparency, low cost, and environmentally friendly nature. ZnO nanoparticles can be synthesized using low-cost solution processing methods that make it compatible with the large area printing or roll-to-roll manufacturing.  One typical phenomenon that appears in solar devices using metal oxide ETLs is the so-called “light-soaking” effect, which refers to the improvement of device PCE under solar illumination over time. The performance of PV devices gradually increases with the increasing exposure time, and eventually reaches a maximum. The light-soaking effect poses as a major issue when solar cells experience variation in environmental illumination. Such an effect directly leads to a fluctuation in the output power and requires additional stabilizing electronics. Several hypotheses have been made to explain such behavior, including the energy barrier between the ETL and active organic layer blocks electron transport [130], oxygen absorption on the metal oxide surface creates deep level defects [131], and excess carriers tunnel through interface levels between metal oxide layer and Phenyl-C61-butyric acid methyl ester (PCBM) layers [132], etc. Currently, the origin of the effect is still not clear. It has been reported that the light-soaking effect can be overcome by employing Atomic Layer Deposition (ALD) of aluminum doped ZnO (AZO) as ETL [130], or by UV exposure [133] of the solar cells for a certain period. Nonetheless, UV irradiation causes degradation of the active polymers with the soaking effect reappearing after storing the devices in the dark. In comparison, Al-doped ZnO can be readily synthesized using low-cost solution processes, which is more suitable for manufacturing of light-soaking free devices.  In this work, the photovoltaic characteristics of inverted OPV devices fabricated with AlxZn(1-x)O as ETL with Al fraction of up to 11% are reported. The light-soaking effect can be 110  eliminated by using more than 4% of Al doping. All the OPV devices demonstrate PCE over 3.4% with air-stability of over 150 days. The suppression of light soaking effect by Al-doping is demonstrated at various Al fractions. The light-soaking mechanism is investigated by employing a numerical simulation on the devices.     5.2 Experiments The structure and the alignment of the highest occupied molecular orbital (HOMO) and lowest unoccupied molecular orbital (LUMO) energy levels for the inverted OPV devices are schematically depicted in Figure 5.1 [134].  ZnO and AZO nanoparticles (NP) with various Al fractions were synthesized with the sol-gel method, adapted from ref. [117]. One Molar of zinc acetate dehydrates [Zn(CH3COO)2·2H2O] (Sigma-Aldrich) was dissolved in 2-methoxyethanol [C3H8O2] (Sigma-Aldrich). 1 molar percentage of ethanolamine [C2H7NO] (Sigma-Aldrich) was added as a stabilizer. In the following steps, a series of solutions were prepared by adding the various amount of aluminum nitrate nonahydrate [Al(NO3)3·9H2O] to adjust the Al3+ ion dopants to target concentrations. The mixtures were then stirred at 60 ºC until all the powders were fully dissolved (~30 min) in ambient atmosphere.  The as-prepared ZnO and AZO precursor inks were spin cast on ITO (20 Ohm/sqr) substrates at 3000 rpm for 10 seconds followed by thermal treatment up to 250 ºC - 320 ºC for 5 minutes depending on the concentration of Al. The coating and annealing processes were repeated three times, resulting in a ~75 nm thick ZnO/AZO layer without pinholes [134].  The bulk heterojunction films were deposited from a 1:1 solution of [6,6]-Phenyl C61 butyric acid methyl ester (PCBM) and Poly(3-hexylthiophene-2,5-diyl) (P3HT, regioregularity 95%, 111  Solaris Chem Inc.) in 1,2-dichlorobenzene (1,2-DCB; 99%, Sigma-Aldrich) with the concentration of 40 mg/ml. The blend ink was sequentially spin-cast on the nanoparticle layers at the rates of 250 rpm (3 s), 1000 rpm (15 s) and 1500 rpm (3 s), resulting in a 200 nm thick layer. The films were placed in a covered Petri dish immediately after coating in order to control the solvent evaporation rate (drying time: 10-20 min). Finally, MoO3 (8 nm) and Ag (120 nm) were thermally deposited (at a vacuum level of 2×10-8 Torr) through a shadow mask to complete the PV device. The overlapped area of the top and bottom electrodes define an active device area of 0.2 cm2.  Figure 5.1 a) The energy level alignment and b) device structure in an inverted OPV using P3HT:PCBM.  Current density-voltage (J-V) measurements of devices were conducted on a computer-controlled Keithley 2400 Source Meter in a vacuum. A xenon lamp (150 W, Newport Co.) equipped with an AM1.5G filter was used as the light source. The optical power was 100 mW/cm2, calibrated using a broadband power meter (Newport Co.). The external quantum efficiency (EQE) was measured using monochromatic light (Cornerstone 130 Monochrometer, Newport Co.) The 112  incident power was measured using a power/energy meter (Newport Co.). Scanning electron microscopic (SEM) and atomic force microscope (AFM) images were recorded using Zeiss Sigma FE-SEM and Asylum Research MFP 3D respectively. Film thickness was measured with a Bruker Dektak XT profilometer. Absorption spectra were measured using a Varian Cary 7000 spectrometer.   5.3 Results and Discussion 5.3.1 Morphology Study ZnO and AZO nanoparticles with various Al fractions were synthesized using sol-gel method for only one layer on ITO-coated glass. The morphology of the nanoparticles was investigated on these samples. AFM analysis indicates that the thickness of a single-layered ZnO and AZO NPs are in the range of 25±2 nm. As is shown in the SEM images in Figure 5.2 a), the ZnO NPs demonstrated an average diameter of about 10 nm. The particles are uniformly deposited all around the surface. However, due to the agglomeration effect during high-temperature treatment, pores and cracks were created all over the film. Therefore, in order to obtain functional OPV devices, multiple depositions are required to avoid direct contact between the active polymer layer and the bottom electrode. In this work, it was found that the Al dopant has a strong influence on the size of the synthesized NPs. With the incorporation of only 1% Al, the average size of the NPs reduced significantly to less than 5 nm (Figure 5.2 b). One possible explanation is that during the synthesis of ZnO and AZO NPs, heating the as-deposited precursor films results in Zn ions reacting with oxygen to form nanocrystalline ZnO particles. Meanwhile, the Al dopants behave as impurities that segregate the grain boundaries and prevent the agglomeration, coalescence, and growth of 113  larger grains [135, 136]. The size of the NPs did not show significant reduction when more Al dopants were added (Figure 5.2 c-k).   Figure 5.2 SEM images of 25 nm thick of a) ZnO, b) 1%-AZO, c-f) Al-doped ZnO with different Al fraction nanoparticles coated on a glass substrate. All the samples were annealed at 250 ºC. a) and b) have higher magnification (EHT 10.00 kV, working distance 9.4 mm, with in-lens camera), and c-f) have the same magnification (EHT 10.00 kV, working distance 9.6 mm with in-lens camera). 114  The Al composition was studied with EDX. The amount of Al dopants detected in the samples shows a linear relation with the doping concentration, as shown in Figure 5.3 a) and b). The intensity of the Al Kα peak detected in the films increases linearly with the molar ratio of Al ions added to the precursor ink, which indicates the proportional incorporation of Al in the film.    Figure 5.3 a) EDX spectra of 25 nm thick of AZO films with various Al fraction. The spectra were normalized to the Oxygen peak. The Al peak is zoomed in and shown in the insect. b) Integrated area of the Al peak vs. Al molar ratio in the precursor inks.   5.3.2 Photovoltaic properties Figure 5.4 (a) shows the comparison of the current density-voltage (J-V) characteristics of the OPV devices with ETL at different Al fraction. The extracted PV parameters are summarized in Table 5.1. The J-V curves were performed continuously while the OPV devices were exposed to AM 1.5 G illumination until the devices reached the maximum output. All the parameters listed in Table 1 are extracted from the saturated PV performance. The VOC for the cell with all intermediate layer is approximately 0.57 V; while the short-circuit current demonstrates a slight 115  increase and then decrease with increasing Al doping level. Similar trends are noted for the fill factor and the PCE, which are proportional to the short-circuit current. But overall, the J-V characteristics of all devices have similar PV output, with PCEs between 3.5 - 4.0%.    Table 5.1 Extracted parameters, short-current density (Jsc), open-circuit voltage (Voc), fill factor (FF) and power conversion efficiency (PCE) for the OPV devices. Sample Al Fraction in ETL   [%] JSC [mA/cm2] VOC  [V] FF  [%] PCE  [%] ZnO 0 12.0 0.57 51.3 3.5 A1 1.0 12.0 0.57 53.5 3.7 A2 2.0 12.1 0.56 53.3 3.6 A4 4.0 12.3 0.57 56.0 3.9 A6 5.5 12.0 0.57 53.5 3.6 A7 7.5 12.0 0.57 52.5 3.6 A9 9.0 11.9 0.57 51.6 3.5 A11 11 11.7 0.56 51.9 3.4 116   Figure 5.4 Photovoltaic performance of the inverted devices using ZnO and AZO as electron transport layers: a) current density-voltage (J-V) characterizations at the day of fabrication; b) external-quantum efficiency (EQE) normalized to the highest absorbing wavelength (520 nm) of P3HT; c) J-V and EQE of sample A4 tested at the day of fabrication and 150 days layer; d) EQE of the sample ZnO within 50 days of fabrication normalized to the highest EQE value.   The annealing temperature of the ETL layers is a key factor for the fabrication of solar cells with high performance. It has been observed that devices with ZnO ETL showed highest PCE when the ZnO was annealed at 250 ºC. To reach the same performance, the annealing temperature has to increase for devices with Al-doped NPs ETL. With 1%-2% of Al doping, the best 117  conductivity and carrier transfer properties were achieved by rising annealing temperature to 290 ºC. For Al fraction higher than 2%, annealing temperature at 320 ºC is required to achieve optimum electrical properties, as is shown in Figure 5.5. The result is consistent with our previous observation regarding the reduction in the size of nanocrystals due to Al doping. The formation of ZnO nanoparticles requires higher energy to incorporate Al impurities into the unit cells and to form a uniform film with high conductivity and mobility [137, 138].    Figure 5.5 Photovoltaic parameters of OPV devices with 1% AZO nanoparticle film as ETL. Device structures of all samples were the same. The AZO films were heated at a various temperature in the ambient atmosphere after deposition. 118  External quantum efficiency (EQE) is defined as the ratio of the number of incident photons and the number of output electrons.  It demonstrates the ability of light absorption and charge collection of the solar cells at the same time. The EQE measurements of the ZnO/AZO devices were performed from 400 to 800 nm. Figure 5.4 (b) shows the normalized EQE curves with respect to the main absorbing wavelength of P3HT at about 520 nm. It is evident that a relative enhancement of EQE at 550 nm and 600 nm occurred for all the Al-doped samples. The enhancement effect on EQE is further increased with a higher Al doping level. A similar phenomenon has been reported before [139]. As the particle sizes are reduced with Al doping, the surface morphology of AZO layers is likely to be improved in comparison to the ZnO film. Al dopants also introduce electrical polarization into the ZnO lattice, which is likely to influence the wettability. When the P3HT:PCBM ink was spin cast on the ETL, an improvement of ordering could occur during the crystallization of P3HT, and subsequently enhance the quantum efficiency [140].  In addition to the test on the as-fabricated device, long-term stability tests have also been carried out. All the OPV devices were exposed to air at room temperature for over five months. J-V measurements were performed periodically during this time. Figure 5.4 c) illustrates the J-V properties measured at the day of fabrication and 150 days later of sample A4 which has 4% Al-doped ZnO as ETL. After exposure to an ambient environment for five months, the Voc and Jsc of the device were nearly unchanged, while the maximum output power decreased by ~32% (PCE from 3.4% to 2.4%) due to the reduction of FF. Figure 5.4 d) illustrates the normalized PCE of a device with ZnO ETL within the first 50 days of air exposure as an example. In contrast to conventional OPV devices, the PCE of our devices first increased for 3-6 days, followed by a gradual decline and eventually plateaued at ~60% of the maximum value. The PCE of devices with 119  Al incorporation show the similar enhancing-declining trend as well. The mechanism behind the increasing of PCE in the beginning stage is still under debate, as air and humidity exposure usually causes damages to conventional OPV devices. However, a similar phenomenon has been reported for P3HT solar cells with TiOx ETL [141]. One possible explanation is that the P3HT molecules are doped with oxygen when exposing to air, and the hole concentration is increased [142]. The interfacial electrostatic force between the electron donor and acceptor layers is increased, resulting in higher efficiency of exciton dissociation and therefore a rising PCE.   5.3.3 Light-soaking effect OPV devices were placed under continuous AM 1.5G illumination of 100 mW/cm2, to characterize the light-soaking effect. J-V curves of the cells were measured periodically with a certain time interval. The J-V plot of sample A1 is shown in Figure 5.6 a). The J-V curve in the beginning stage of the illumination appears as a straight line without rectifying behavior, similar to that of a resistor. The dark current (dash line) of sample A1 has the same shape. Solar devices with such J-V characteristics usually associated with a high series resistance (Rs) and low charge collection rates. With an increasing illumination time, the fill factor and the short-circuit current both increases and the devices start to show J-V characteristics of a diode. During the soaking process, the generated electrons were unable to be transferred to the respective electrode but recombined within the active layer, causing loss to the output power. After enough soaking time under light irradiation, the device parameters reached saturated values. Solar devices with Al-doped ZnO ETL were also tested. The FF and EQE values over the illumination time of all samples are presented in Figure 5.6 b) and c). Each curve is normalized to the highest FF or EQE value of the respective sample. As seen, the soaking time and the magnitude are directly related to the Al 120  doping fraction. When the Al incorporation reaches more than 4%, the light-soaking effect is eliminated. No shifting of FF or EQE is noticed during the continuous irradiation. Some samples also show decreasing FF and EQE after illumination for a few minutes. This is typical behavior for OPV devices, which is caused by increased charge scattering and recombination when the surface temperature rises due to irradiation.     Figure 5.6 Light-soaking effect in the inverted OPV devices with ZnO/AZO electron transport layers: a) periodically J-V characterizations of sample A1 measured under continuous light illumination; normalized b) fill factor and c) EQE of each sample with illumination time.  121  In order to understand the light-soaking mechanism, the UV-Visible absorption measurements were performed on ZnO, and AZO films coated glass samples to investigate the shifting of the bandgap with Al doping. Each sample was coated with 75±3 nm thick ZnO/AZO particles. The samples were cured in air at 250 ºC for ZnO, 290 ºC for AZO with 1% Al, and 320 ºC for samples with Al concentration > 1%. The absorption spectra (Figure 5.7 a) reveals above 90% transmittance over 400 nm to 800 nm on all the ZnO and AZO thin layers. The optical bandgap is extrapolated from a linear fit to the plot of the squared absorption coefficients vs. photon energy. As shown in Figure 5.7 b) and c), the optical bandgap increases with higher Al molar ratio of the NPs. The bandgap shifting is consistent with the well-known Moss–Burstein effect [143, 144] in which the absorption edge of a semiconductor is pushed to higher energies when charge carriers populate states close to the conduction band. The carrier concentration is increased with more Al incorporation, pushing the Fermi level closer to or even above the original conduction band of ZnO films. Assuming the conduction and valence bands of AZO are parabolic, the bandgap shifting ΔEg and the carrier densities ne follow the relationship: ∆𝐄𝐠 =𝐡𝟐𝟖𝐦𝐞∗ (𝟑𝛑𝐧𝐞)𝟐𝟑⁄      Equation 5.1 where me* is the effective mass of electrons, and h is the Planck constant. The estimated carrier density in sample A4 is about 1×1019 cm-3 if me* is assumed to be 0.17 times of the free electron mass. This value is close to other report electron concentration (in the 1020 cm-3 range) in AZO thin films [136, 137, 145].   122   Figure 5.7 a) Transmittance versus wavelength plot and b) (αh)2 versus photon energy plot of AlxZn(1-x)O nanoparticles, x ranged from 0 to 11 %, deposited on glass substrates. c) optical bandgap of ZnO nanoparticles doped with various level of Al; the inset schematic shows the shifting of Fermi level when Al is added into ZnO nanoparticles.  A numerical simulation based on a modified hetero-PN-junction model was established to calculate the band bending of the inverted OPV devices at 0 V bias. The HOMO and LUMO levels of each layer were obtained from ref. [109, 134, 146]. The carrier concentration of the polymer layers is extracted from the saturated short-circuit current of sample ZnO and A4, while the free electron concentration of ZnO and A4 are obtained from ref. [145, 147]. The thickness of each 123  layer is assumed to be the same as shown in Figure 5.1 b), while P3HT and PCBM layers each take half of the BHJ thickness.      Figure 5.8 The simulated band diagram of the inverted OPV devices using a) ZnO (black lines) and b) AZO (red lines) with 4% Al electron transport layers.  Trost et al. [130] believe that by shifting the Fermi level closer to the conduction band, the pre-existing mismatch between the conduction band of PCBM and ZnO would be reversed for PCBM and AZO interfaces. Also, the opposite interface dipoles could lead to the transport of the 124  generated electrons to the AZO side. However, as the conduction band of ZnO and AZO remains at the same energy, shifting of the Fermi level will only bend the conduction band of both layers, as illustrated in Figure 5.8 a) and b). The interfacial dipoles will remain in the same direction, as the barrier is the same as it without Al doping.  As shown in Figure 5.8 a) and b), the interface between ITO and ZnO/AZO should be considered as a Schottky contact instead of an Ohmic contact since there is a 0.4 eV mismatch between the conduction band of ZnO/AZO and the work function of ITO. When the free carrier density is low in the ETL, the entire layer is depleted to create enough bending on the conduction band to match with the adjacent PCBM and ITO layers. The generated electrons are then swept back to the BHJ and recombined with the generated holes, causing a low charge collection rate. When Al dopant is introduced into the ZnO lattice, the free carrier density in the ETL layer is increased, and the depletion region on both sides of the AZO layer is reduced. Eventually, when the depletion region is narrow enough to allow the generated electrons tunneling through the barriers, the J-V curves of the devices saturate and the light-soaking effect is fully eliminated. UV exposure, which injects photons with higher energy than the bandgap of ZnO, would excite bonded electrons in the ZnO to the conduction band and increasing the carrier density. Henceforth, the OPV devices will show saturation after a certain exposure time. Once the UV light is removed, the carrier density in ZnO decreases and the layer is again depleted and blocking electron transport.   5.4 Conclusions  ZnO and other metal oxides are commonly used in the OPV device fabrication as an electron extraction layer. ZnO nanoparticle ETL can be synthesized with low-cost solution process in the ambient environment, with potential for large-scale manufacturing. OPV devices with ZnO ETL 125  and inverted structure have a high level of air stability and a prolonged lift-time even without encapsulation. The light-soaking effect is widely observed in the OPV devices using metal oxide ETLs, which causes a major concern in the application of the inverted OPV devices. In this paper, we fabricated inverted OPV devices with Aluminum doped ZnO electron transport layers. The samples demonstrated high solar harvesting properties with power conversion efficiency up to 3.9%. Air-stability tests of up to 150 days were performed on devices with different Al doping levels. The devices maintained higher than 60% of the initial PCE after 50 days of open-air exposure. The mechanism of the light-soaking effect was investigated with experiments and simulations. The effect can be fully eliminated when the Al fraction in the AZO ETL is higher than 4%. The simulated band diagram of the OPV devices indicates that the low carrier density in the ZnO layer is the main reason for the light-soaking effect. The carrier in the thin ZnO layer is fully depleted to match with the conduction band edge of the adjacent PCBM and ITO layers that lead to the generated electrons being blocked from transferring to the ITO electrode. Doping the ZnO layer as well as exposing the devices under UV irradiation will introduce additional free carriers into the ETL and reduce the width of the depletion region at both sides of the ETL. Electrons, therefore, gain the opportunity to tunnel through both barriers and transport to the electrode.       126  Chapter 6: Conclusion and future work  This work focuses on the CVD-grown graphene films transferred to different substrates with several wet etching solutions. The temperature dependent electrical properties of graphene films transferred to SiO2/Si substrates were studied with optical and electrical methods. The influence of different copper etchants on graphene was discussed. The single layer graphene films were transferred to plastic substrates. The conductivity and transparency were modified with charge selective (both hole- and electron-selective) thin films for flexible transparent conductor applications. Flexible organic photovoltaic devices were fabricated with such modified graphene electrodes. The solar harvesting properties and flexibility were presented. One of the interesting electron-selective materials, ZnO nanoparticles, were applied in inverted OPV devices. The mechanism of the widely-existing “light-soaking effect” in such devices was studied with experimental and simulation data.   6.1 Contribution  A better understanding of the influence of the wet etching solutions on the transferred large-area graphene was obtained. The G-Cu samples demonstrate a uniform and continuous single layer; the G-PSF samples are continuous but with multiple layers; the G-Fe samples have more ruptures and overlapping fragments. All three samples demonstrate a certain level of p-type doping. The estimated length between defects is about 45 nm for G-Cu sample and lower for the other two samples. All the samples display a similar trend of AC conductivity and permittivity at the temperature range of -50 ºC to 300 ºC. G-Cu sample shows the lowest DC conductivity due to its fewer layers and lower doping level. 127  Several factors were discussed which may contribute to the difference of the electrical properties: the PMMA residue, the density of line defects and grain boundaries, the water molecule intercalation, etc.   Charge selective graphene transparent conductors with improved conductivity and strong flexibility were achieved. The hole-selective conductors using PEDOT:PSS and MoO3 films both exhibit about 50% reduction in sheet resistance while maintaining above 90% transparency. The samples demonstrate high flexibility with less than 5% change in resistivity while bent up to 130 degrees over a 0.2 mm thick metal wire. Long-term stability of the conductors was tested with continuous bending with the same conditions. Less than 5% of the change in sheet resistance was found in the optimized samples with more than 800 bending cycles.  Flexible organic solar cells using hole-selective material coated graphene transparent conducting electrodes were fabricated. The devices fabricated on glass substrates show comparable PV performance with the ITO-based samples with PCE up to 2.1 %. Spin-coating of the active polymer materials on flexible substrates could cause uneven distribution and therefore affect the PCE of the devices. Spray-coating can be used to address the issue. The flexible OPV devices exhibit a PCE of 1.5% while bent up to 130 degrees over a 0.2 mm thick metal wire.   The light-soaking effects in inverted OPV devices with zinc oxide (ZnO) and Aluminum-doped ZnO electron transport layers were investigated, which is important for the development of low-cost and durable solar cells. The samples demonstrated high solar harvesting properties with power conversion efficiency up to 3.9%. Air-stability tests of up to 150 days were performed on devices with different Al doping levels. The devices 128  maintained higher than 60% of the initial PCE after 50 days of open-air exposure. The light-soaking mechanism was investigated with experiments and simulations. It was shown that the light-soaking effect could be completely eliminated when the Al fraction of the AZO is higher than 4%. The simulated band diagram of the OPV devices indicates that the low carrier density in the ZnO layer by virtue of depletion is the main reason for the light-soaking effect. Doping the ZnO layer as well as exposing the devices under UV irradiation will introduce additional free carriers into the ETL and reduce the width of the depletion region at both sides of the ETL.  6.2 Future work  Further study in the charge transport mechanisms in polycrystalline graphene is needed. The reason causing the different trend of DC conductivity vs. temperature of the graphene films prepared with different copper etchants is still inconclusive. Surface characterization techniques, such as ToF-SIMS, can be used to learn more about the residual chemicals on the graphene films. In situ Raman spectroscopy of the graphene samples during heating and cooling, and intended doping may expose more information on change of conductivity.  The integration of ZnO nanoparticles with graphene to fabricate electron-selective conductors was not as successful so far. The samples did not show significant improvement compared to the non-coated graphene. Other materials with their Fermi levels better aligned with the graphene will be required for future studies.   Spray-coating is a powerful technique for fabricating large-area organic solar cells. Besides, it shows high potential for the application in flexible electronic devices. The preparation of polymer blends, the spray-coating, and the drying parameters need to be 129  investigated to obtain an optimized film. By using the modified graphene transparent conductors as electrodes and the spray-coating technique, roll-to-roll manufacture of OPV or other solution processed devices on flexible substrates can be achieved. Deposition parameters of each layer, post-deposition treatments, and other related parameters need to be studied for low-cost, high-speed mass production.     130  Bibliography  1. Novoselov, K.S., et al., Electric field effect in atomically thin carbon films. Science, 2004. 306(5696): p. 666-9. 2. Mayorov, A.S., et al., Micrometer-Scale Ballistic Transport in Encapsulated Graphene at Room Temperature. Nano Letters, 2011. 11(6): p. 2396-2399. 3. Baringhaus, J., et al., Exceptional ballistic transport in epitaxial graphene nanoribbons. Nature, 2014. 506(7488): p. 349-354. 4. Lee, C., et al., Measurement of the elastic properties and intrinsic strength of monolayer graphene. Science, 2008. 321(5887): p. 385-388. 5. Balandin, A.A., Thermal properties of graphene and nanostructured carbon materials. Nature Materials, 2011. 10(8): p. 569-581. 6. McCann, E. and V.I. Fal'ko, Landau-level degeneracy and quantum hall effect in a graphite bilayer. Physical Review Letters, 2006. 96(8). 7. Zhang, Y.B., et al., Direct observation of a widely tunable bandgap in bilayer graphene. Nature, 2009. 459(7248): p. 820-823. 8. Oostinga, J.B., et al., Gate-induced insulating state in bilayer graphene devices. Nature Materials, 2008. 7(2): p. 151-157. 9. Ferrari, A.C., et al., Science and technology roadmap for graphene, related two-dimensional crystals, and hybrid systems. Nanoscale, 2015. 7(11): p. 4598-810. 10. Nemes-Incze, P., et al., Revealing the grain structure of graphene grown by chemical vapor deposition. Applied Physics Letters, 2011. 99(2). 131  11. Li, X.S., et al., Transfer of Large-Area Graphene Films for High-Performance Transparent Conductive Electrodes. Nano Letters, 2009. 9(12): p. 4359-4363. 12. Krebs, F.C., et al., A round robin study of flexible large-area roll-to-roll processed polymer solar cell modules. Solar Energy Materials And Solar Cells, 2009. 93(11): p. 1968-1977. 13. Li, X.S., et al., Graphene Films with Large Domain Size by a Two-Step Chemical Vapor Deposition Process. Nano Letters, 2010. 10(11): p. 4328-4334. 14. Bae, S., et al., Roll-to-roll production of 30-inch graphene films for transparent electrodes. Nature Nanotechnology, 2010. 5(8): p. 574-578. 15. Kobayashi, T., et al., Production of a 100-m-long high-quality graphene transparent conductive film by roll-to-roll chemical vapor deposition and transfer process. Applied Physics Letters, 2013. 102(2). 16. Hamberg, I. and C.G. Granqvist, Evaporated Sn-Doped In2o3 Films - Basic Optical-Properties and Applications to Energy-Efficient Windows. Journal of Applied Physics, 1986. 60(11): p. R123-R159. 17. L. Holland, G.S., Vacuum, 1953. 3: p. 375–391. 18. Minami, T., Transparent conducting oxide semiconductors for transparent electrodes. Semiconductor Science and Technology, 2005. 20(4): p. S35-S44. 19. Granqvist, C.G., Transparent conductors as solar energy materials: A panoramic review. Solar Energy Materials and Solar Cells, 2007. 91(17): p. 1529-1598. 20. Sheraw, C.D., et al., Organic thin-film transistor-driven polymer-dispersed liquid crystal displays on flexible polymeric substrates. Applied Physics Letters, 2002. 80(6): p. 1088-1090. 132  21. Kumar, A. and C.W. Zhou, The Race To Replace Tin-Doped Indium Oxide: Which Material Will Win? Acs Nano, 2010. 4(1): p. 11-14. 22. Lee, J.Y., et al., Solution-processed metal nanowire mesh transparent electrodes. Nano Letters, 2008. 8(2): p. 689-692. 23. Zhong, C.M., et al., Materials and Devices toward Fully Solution Processable Organic Light-Emitting Diodes. Chemistry of Materials, 2011. 23(3): p. 326-340. 24. Allen, G.L., et al., Small Particle Melting of Pure Metals. Thin Solid Films, 1986. 144(2): p. 297-308. 25. Zhang, M., et al., Size-dependent melting point depression of nanostructures: Nanocalorimetric measurements. Physical Review B, 2000. 62(15): p. 10548-10557. 26. Zeng, X.Y., et al., A New Transparent Conductor: Silver Nanowire Film Buried at the Surface of a Transparent Polymer. Advanced Materials, 2010. 22(40): p. 4484-4488. 27. Hellstrom, S.L., H.W. Lee, and Z.N. Bao, Polymer-Assisted Direct Deposition of Uniform Carbon Nanotube Bundle Networks for High Performance Transparent Electrodes. Acs Nano, 2009. 3(6): p. 1423-1430. 28. King, P.J., et al., Percolation Effects in Supercapacitors with Thin, Transparent Carbon Nanotube Electrodes. Acs Nano, 2012. 6(2): p. 1732-1741. 29. Pang, S.P., et al., Graphene as Transparent Electrode Material for Organic Electronics. Advanced Materials, 2011. 23(25): p. 2779-2795. 30. Hwang, J.O., et al., Workfunction-Tunable, N-Doped Reduced Graphene Transparent Electrodes for High-Performance Polymer Light-Emitting Diodes. Acs Nano, 2012. 6(1): p. 159-167. 133  31. Jin, Y., et al., Solving Efficiency-Stability Tradeoff in Top-Emitting Organic Light-Emitting Devices by Employing Periodically Corrugated Metallic Cathode. Advanced Materials, 2012. 24(9): p. 1187-1191. 32. Sergeant, N.P., et al., Design of Transparent Anodes for Resonant Cavity Enhanced Light Harvesting in Organic Solar Cells. Advanced Materials, 2012. 24(6): p. 728-+. 33. Wang, Z.B., et al., Unlocking the full potential of organic light-emitting diodes on flexible plastic. Nature Photonics, 2011. 5(12): p. 753-757. 34. Cho, H., et al., Highly flexible organic light-emitting diodes based on ZnS/Ag/WO3 multilayer transparent electrodes. Organic Electronics, 2009. 10(6): p. 1163-1169. 35. Jung, G.H., et al., BCP/Ag/MoO3 Transparent Cathodes for Organic Photovoltaics. Advanced Energy Materials, 2011. 1(6): p. 1023-1028. 36. Hong, K., et al., Optical Properties of WO3/Ag/WO3 Multi layer As Transparent Cathode in Top-Emitting Organic Light Emitting Diodes. Journal of Physical Chemistry C, 2011. 115(8): p. 3453-3459. 37. Kim, S. and J.L. Lee, Design of dielectric/metal/dielectric transparent electrodes for flexible electronics. Journal of Photonics for Energy, 2012. 2. 38. Vitoratos, E., et al., Thermal degradation mechanisms of PEDOT:PSS. Organic Electronics, 2009. 10(1): p. 61-66. 39. Elschner, A., The spectral sensitivity of PEDOT:PSS films. Solar Energy Materials and Solar Cells, 2011. 95(5): p. 1333-1338. 40. Bonaccorso, F., et al., Graphene photonics and optoelectronics. Nature Photonics, 2010. 4(9): p. 611-622. 134  41. Archer, C.L. and M.Z. Jacobson, Evaluation of global wind power. Journal of Geophysical Research-Atmospheres, 2005. 110(D12). 42. Renewable biological systems for alternative sustainable energy production. FAO Agricultural Services Bulletin. 1997, Osaka, Japan. 43. Green, M.A., et al., Solar cell efficiency tables (Version 45). Progress in Photovoltaics, 2015. 23(1): p. 1-9. 44. Hosoya, M., Oooka, H., Nakao, H., Mori, S., Gotanda, T., Shida, N., Saito, M., Nakano, Y., Todori, K., Module development for polymer solar cells, in Grand Renewable Energy Conference. 2014: Tokyo. p. 21-37. 45. Liang, Y.Y., et al., For the Bright Future-Bulk Heterojunction Polymer Solar Cells with Power Conversion Efficiency of 7.4%. Advanced Materials, 2010. 22(20): p. E135-+. 46. Gholamkhass, B. and S. Holdcroft, Enhancing the durability of polymer solar cells using gold nano-dots. Solar Energy Materials And Solar Cells, 2011. 95(11): p. 3106-3113. 47. Peet, J., A.J. Heeger, and G.C. Bazan, "Plastic" Solar Cells: Self-Assembly of Bulk Heterojunction Nanomaterials by Spontaneous Phase Separation. Accounts of Chemical Research, 2009. 42(11): p. 1700-1708. 48. Shaw, P.E., A. Ruseckas, and I.D.W. Samuel, Exciton diffusion measurements in poly(3-hexylthiophene). Advanced Materials, 2008. 20(18): p. 3516-+. 49. Asmatulu, R., et al., Effects of UV degradation on surface hydrophobicity, crack, and thickness of MWCNT-based nanocomposite coatings. Progress in Organic Coatings, 2011. 72(3): p. 553-561. 135  50. Li, C. and G.Q. Shi, Synthesis and electrochemical applications of the composites of conducting polymers and chemically converted graphene. Electrochimica Acta, 2011. 56(28): p. 10737-10743. 51. Yong, V. and J.M. Tour, Theoretical Efficiency of Nanostructured Graphene-Based Photovoltaics. Small, 2010. 6(2): p. 313-318. 52. Wu, J.B., et al., Organic solar cells with solution-processed graphene transparent electrodes. Applied Physics Letters, 2008. 92(26). 53. Wang, X., et al., Transparent carbon films as electrodes in organic solar cells. Angewandte Chemie-International Edition, 2008. 47(16): p. 2990-2992. 54. Tung, V.C., et al., Low-Temperature Solution Processing of Graphene-Carbon Nanotube Hybrid Materials for High-Performance Transparent Conductors. Nano Letters, 2009. 9(5): p. 1949-1955. 55. Fiori, G., et al., Electronics based on two-dimensional materials. Nature Nanotechnology, 2014. 9: p. 768. 56. Koppens, F.H.L., et al., Photodetectors based on graphene, other two-dimensional materials and hybrid systems. Nature Nanotechnology, 2014. 9: p. 780. 57. Schwierz, F., Graphene transistors. Nature Nanotechnology, 2010. 5: p. 487. 58. Mehr, W., et al., Vertical Graphene Base Transistor. IEEE Electron Device Letters, 2012. 33(5): p. 691-693. 59. Vaziri, S., et al., A Graphene-Based Hot Electron Transistor. Nano Letters, 2013. 13(4): p. 1435-1439. 60. Lin, Y.-C., et al., Graphene Annealing: How Clean Can It Be? Nano Letters, 2012. 12(1): p. 414-419. 136  61. Ferrari, A.C., et al., Science and technology roadmap for graphene, related two-dimensional crystals, and hybrid systems. Nanoscale, 2015. 7(11): p. 4598-4810. 62. Kim, Y., et al., Influence of the transfer and chemical treatment of monolayer graphene grown for flexible transparent electrodes. Carbon, 2015. 81: p. 458-464. 63. Lupina, G., et al., Residual Metallic Contamination of Transferred Chemical Vapor Deposited Graphene. ACS Nano, 2015. 9(5): p. 4776-4785. 64. Vermeire, B., L. Lee, and H.G. Parks, The effect of copper contamination on field overlap edges and perimeter junction leakage current. IEEE Transactions on Semiconductor Manufacturing, 1998. 11(2): p. 232-238. 65. Li, X., et al., Large-Area Synthesis of High-Quality and Uniform Graphene Films on Copper Foils. Science, 2009. 324(5932): p. 1312-1314. 66. Suk, J.W., et al., Enhancement of the Electrical Properties of Graphene Grown by Chemical Vapor Deposition via Controlling the Effects of Polymer Residue. Nano Letters, 2013. 13(4): p. 1462-1467. 67. Alemán, B., et al., Transfer-Free Batch Fabrication of Large-Area Suspended Graphene Membranes. ACS Nano, 2010. 4(8): p. 4762-4768. 68. Lee, G.-H., et al., High-Strength Chemical-Vapor–Deposited Graphene and Grain Boundaries. Science, 2013. 340(6136): p. 1073-1076. 69. Zhan, D., et al., FeCl3-Based Few-Layer Graphene Intercalation Compounds: Single Linear Dispersion Electronic Band Structure and Strong Charge Transfer Doping. Advanced Functional Materials, 2010. 20(20): p. 3504-3509. 70. Suk, J.W., et al., Transfer of CVD-Grown Monolayer Graphene onto Arbitrary Substrates. ACS Nano, 2011. 5(9): p. 6916-6924. 137  71. Atkin, J.M., et al., Nano-optical imaging and spectroscopy of order, phases, and domains in complex solids. Advances in Physics, 2012. 61(6): p. 745-842. 72. Jonscher, A.K., Dielectric relaxation in solids. 1983, London: Chelsea Dielectrics Press. 73. Obraztsov, A.N., et al., Chemical vapor deposition of thin graphite films of nanometer thickness. Carbon, 2007. 45(10): p. 2017-2021. 74. Mattevi, C., H. Kim, and M. Chhowalla, A review of chemical vapour deposition of graphene on copper. Journal of Materials Chemistry, 2011. 21(10): p. 3324-3334. 75. Park, S., et al., Graphene oxide papers modified by divalent ions - Enhancing mechanical properties via chemical cross-linking. Acs Nano, 2008. 2(3): p. 572-578. 76. Wu, J.-B., et al., Raman spectroscopy of graphene-based materials and its applications in related devices. Chemical Society Reviews, 2018. 47(5): p. 1822-1873. 77. Das, A., et al., Monitoring dopants by Raman scattering in an electrochemically top-gated graphene transistor. Nature Nanotechnology, 2008. 3: p. 210. 78. Malard, L.M., et al., Raman spectroscopy in graphene. Physics Reports, 2009. 473(5): p. 51-87. 79. Ferrari, A.C. and D.M. Basko, Raman spectroscopy as a versatile tool for studying the properties of graphene. Nature Nanotechnology, 2013. 8: p. 235. 80. Tuinstra, F. and J.L. Koenig, Raman Spectrum of Graphite. The Journal of Chemical Physics, 1970. 53(3): p. 1126-1130. 81. Matthews, M.J., et al., Origin of dispersive effects of the Raman D band in carbon materials. Physical Review B, 1999. 59(10): p. R6585-R6588. 82. Lucchese, M.M., et al., Quantifying ion-induced defects and Raman relaxation length in graphene. Carbon, 2010. 48(5): p. 1592-1597. 138  83. Cançado, L.G., et al., General equation for the determination of the crystallite size La of nanographite by Raman spectroscopy. Applied Physics Letters, 2006. 88(16): p. 163106. 84. Fei, Z., et al., Electronic and plasmonic phenomena at graphene grain boundaries. Nature Nanotechnology, 2013. 8: p. 821. 85. Banszerus, L., et al., Ballistic Transport Exceeding 28 μm in CVD Grown Graphene. Nano Letters, 2016. 16(2): p. 1387-1391. 86. Avila, J., et al., Exploring electronic structure of one-atom thick polycrystalline graphene films: A nano angle resolved photoemission study. Scientific Reports, 2013. 3: p. 2439. 87. Huang, X.Y., et al., Temperature-dependent electrical property transition of graphene oxide paper. Nanotechnology, 2012. 23(45). 88. Choi, W., et al., Influence of removing PMMA residues on surface of CVD graphene using a contact-mode atomic force microscope. RSC Advances, 2017. 7(12): p. 6943-6949. 89. Jonscher, A.K., The ‘universal’ dielectric response. Nature, 1977. 267: p. 673. 90. Mann, S., R. Kumar, and V.K. Jindal, Negative thermal expansion of pure and doped graphene. RSC Advances, 2017. 7(36): p. 22378-22387. 91. Gierz, I., et al., Atomic Hole Doping of Graphene. Nano Letters, 2008. 8(12): p. 4603-4607. 92. Schedin, F., et al., Detection of individual gas molecules adsorbed on graphene. Nature Materials, 2007. 6(9): p. 652-655. 93. Ristein, J., Surface transfer doping of semiconductors. Science, 2006. 313(5790): p. 1057-1058. 94. Bolotin, K.I., et al., Ultrahigh electron mobility in suspended graphene. Solid State Communications, 2008. 146(9-10): p. 351-355. 139  95. Efetov, D.K. and P. Kim, Controlling Electron-Phonon Interactions in Graphene at Ultrahigh Carrier Densities. Physical Review Letters, 2010. 105(25). 96. Morozov, S.V., et al., Giant intrinsic carrier mobilities in graphene and its bilayer. Physical Review Letters, 2008. 100(1). 97. Chen, J.H., et al., Intrinsic and extrinsic performance limits of graphene devices on SiO2. Nature Nanotechnology, 2008. 3(4): p. 206-209. 98. Gong, Y., et al., Layer-Controlled and Wafer-Scale Synthesis of Uniform and High-Quality Graphene Films on a Polycrystalline Nickel Catalyst. Advanced Functional Materials, 2012. 22(15): p. 3153-3159. 99. Jiao, T., et al., Flexible solar cells based on graphene-ultrathin silicon Schottky junction. RSC Advances, 2015. 5(89): p. 73202-73206. 100. Park, H., et al., Flexible Graphene Electrode-Based Organic Photovoltaics with Record-High Efficiency. Nano Letters, 2014. 14(9): p. 5148-5154. 101. Yi, S., et al., Iron (III) Chloride doping of CVD graphene. Nanotechnology, 2014. 25(39): p. 395701. 102. Kim, C.O., et al., High photoresponsivity in an all-graphene p–n vertical junction photodetector. Nature Communications, 2014. 5: p. 3249. 103. Oh, S., B.-J. Kim, and J. Kim, Layer-by-layer AuCl3 doping of stacked graphene films. physica status solidi (RRL) – Rapid Research Letters, 2014. 8(5): p. 441-444. 104. Wang, Y., Highly Flexible Top-Emitting Phosphorescent Organic Light Emitting Diodes (OLEDs). 2014, MSc Thesis, University of British Columbia. 105. http://www.dupont.com/content/dam/assets/products-and-services/membranes-films/assets/DEC-Kapton-summary-of-properties.pdf. 140  106. http://www.teijindupontfilms.jp/english/product/hi_film.html. 107. https://www.astm.org/Standards/G173.htm. 108. http://www.heraeus.com/media/group/doc_group/products_1/conductive_polymers_1/p/ CLEVIOS_P_VP_AI_4083.pdf. 109. Zhong, J.Q., et al., Ionization potential dependent air exposure effect on the MoO3/organic interface energy level alignment. Organic Electronics, 2012. 13(12): p. 2793-2800. 110. Irfan, I., et al., Work function recovery of air exposed molybdenum oxide thin films. Applied Physics Letters, 2012. 101(9): p. 4. 111. Meyer, J., et al., Metal Oxide Induced Charge Transfer Doping and Band Alignment of Graphene Electrodes for Efficient Organic Light Emitting Diodes. Scientific Reports, 2014. 4. 112. Jiang, Z., B. Gholamkhass, and P. Servati, Effects of Interlayer Properties on the Performance of Tandem Organic Solar Cells with Low and High Band-gap Polymers. 2017, University of British Columbia: Unpublished Work. 113. Mai, L.Q., et al., Molybdenum oxide nanowires: synthesis & properties. Materials Today, 2011. 14(7-8): p. 346-353. 114. Baek, S., R.A. Green, and L.A. Poole-Warren, Effects of dopants on the biomechanical properties of conducting polymer films on platinum electrodes. Journal Of Biomedical Materials Research Part A, 2014. 102(8): p. 2743-2754. 115. Vosgueritchian, M., D.J. Lipomi, and Z.A. Bao, Highly Conductive and Transparent PEDOT:PSS Films with a Fluorosurfactant for Stretchable and Flexible Transparent Electrodes. Advanced Functional Materials, 2012. 22(2): p. 421-428. 141  116. Pacholski, C., A. Kornowski, and H. Weller, Self-assembly of ZnO: from nanodots to nanorods. Angew Chem Int Ed Engl, 2002. 41(7): p. 1188-91. 117. Chou, C.H., et al., A Metal-Oxide Interconnection Layer for Polymer Tandem Solar Cells with an Inverted Architecture. Advanced Materials, 2011. 23(10): p. 1282-1286. 118. Yu, G., et al., Polymer Photovoltaic Cells: Enhanced Efficiencies via a Network of Internal Donor-Acceptor Heterojunctions. Science, 1995. 270(5243): p. 1789-1791. 119. Brabec, C.J., et al., Polymer–Fullerene Bulk-Heterojunction Solar Cells. Advanced Materials, 2010. 22(34): p. 3839-3856. 120. Kumar, A. and C. Zhou, The Race To Replace Tin-Doped Indium Oxide: Which Material Will Win? ACS Nano, 2010. 4(1): p. 11-14. 121. Na, S.-I., et al., Efficient organic solar cells with solution-processed carbon nanosheets as transparent electrodes. Applied Physics Letters, 2013. 102(4): p. 043304. 122. Kymakis, E., et al., Flexible Organic Photovoltaic Cells with In Situ Nonthermal Photoreduction of Spin-Coated Graphene Oxide Electrodes. Advanced Functional Materials, 2013. 23(21): p. 2742-2749. 123. Liu, Z., J. Li, and F. Yan, Package-Free Flexible Organic Solar Cells with Graphene top Electrodes. Advanced Materials, 2013. 25(31): p. 4296-4301. 124. Sangchul, L., et al., Flexible organic solar cells composed of P3HT:PCBM using chemically doped graphene electrodes. Nanotechnology, 2012. 23(34): p. 344013. 125. Rahmanian, R., Charge-selective transparent conductors for solution processed organic solar cells, in Electrical and computer engineering. 2017, PhD Thesis, University of British Columbia. 142  126. Green, M.A., et al., Solar cell efficiency tables (version 51). Progress in Photovoltaics, 2018. 26(1): p. 3-12. 127. de Jong, M.P., L.J. van IJzendoorn, and M.J.A. de Voigt, Stability of the interface between indium-tin-oxide and poly(3,4-ethylenedioxythiophene)/poly(styrenesulfonate) in polymer light-emitting diodes. Applied Physics Letters, 2000. 77(14): p. 2255-2257. 128. Wong, K.W., et al., Blocking reactions between indium-tin oxide and poly (3,4-ethylene dioxythiophene): poly(styrene sulphonate) with a self-assembly monolayer. Applied Physics Letters, 2002. 80(15): p. 2788-2790. 129. He, Z.C., et al., Enhanced power-conversion efficiency in polymer solar cells using an inverted device structure. Nature Photonics, 2012. 6(9): p. 591-595. 130. Trost, S., et al., Overcoming the "Light-Soaking" Issue in Inverted Organic Solar Cells by the Use of Al:ZnO Electron Extraction Layers. Advanced Energy Materials, 2013. 3(11): p. 1437-1444. 131. Nam, C.Y., Ambient Air Processing Causes Light Soaking Effects in Inverted Organic Solar Cells Employing Conjugated Polyelectrolyte Electron Transfer Layer. Journal of Physical Chemistry C, 2014. 118(47): p. 27219-27225. 132. Manor, A., et al., Enhancing functionality of ZnO hole blocking layer in organic photovoltaics. Solar Energy Materials and Solar Cells, 2012. 98: p. 491-493. 133. Schmidt, H., et al., Transient characteristics of inverted polymer solar cells using titaniumoxide interlayers. Applied Physics Letters, 2010. 96(24). 134. Gholamkhass, B., N.M. Kiasari, and P. Servati, An efficient inverted organic solar cell with improved ZnO and gold contact layers. Organic Electronics, 2012. 13(6): p. 945-953. 143  135. Mayo, M.J. and D.C. Hague, Porosity-Grain Growth Relationships in the Sintering of Nanocrystalline Ceramics. Nanostructured Materials, 1993. 3(1-6): p. 43-52. 136. Hasuike, N., et al., Room Temperature Growth of Al-Doped ZnO Thin Films by Reactive DC Sputtering Technique with Metallic Target. Japanese Journal of Applied Physics, 2013. 52(1). 137. Hong, J., et al., The effect of growth temperature on physical properties of heavily doped ZnO:Al films. Physica Status Solidi a-Applications and Materials Science, 2009. 206(4): p. 697-703. 138. Dasgupta, N.P., et al., Atomic Layer Deposition of Al-doped ZnO Films: Effect of Grain Orientation on Conductivity. Chemistry of Materials, 2010. 22(16): p. 4769-4775. 139. Tsai, S.H., et al., Toward high efficiency of inverted organic solar cells: Concurrent improvement in optical and electrical properties of electron transport layers. Applied Physics Letters, 2013. 102(25). 140. Lloyd, M.T., et al., Impact of interfacial polymer morphology on photoexcitation dynamics and device performance in P3HT/ZnO heterojunctions. Journal of Materials Chemistry, 2009. 19(26): p. 4609-4614. 141. Kim, C.S., et al., Transient photovoltaic behavior of air-stable, inverted organic solar cells with solution-processed electron transport layer. Applied Physics Letters, 2009. 94(11). 142. Abdou, M.S.A., et al., Interaction of oxygen with conjugated polymers: Charge transfer complex formation with poly(3-alkylthiophenes). Journal of the American Chemical Society, 1997. 119(19): p. 4518-4524. 143. Jia, J.J., et al., Experimental observation on the Fermi level shift in polycrystalline Al-doped ZnO films. Journal of Applied Physics, 2012. 112(1). 144  144. Sternheimer, R.M., Kinematic Criterion for Meson Production in Fundamental Particle Collisions. Physical Review, 1954. 93(3): p. 642-643. 145. Kim, J.S., et al., Optical analysis of doped ZnO thin films using nonparabolic conduction-band parameters. Journal of Applied Physics, 2012. 111(12). 146. Davis, R.J., et al., Determination of energy level alignment at interfaces of hybrid and organic solar cells under ambient environment. Journal of Materials Chemistry, 2011. 21(6): p. 1721-1729. 147. Janotti, A. and C.G. Van de Walle, Fundamentals of zinc oxide as a semiconductor. Reports on Progress in Physics, 2009. 72(12).    

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
https://iiif.library.ubc.ca/presentation/dsp.24.1-0376033/manifest

Comment

Related Items