{"http:\/\/dx.doi.org\/10.14288\/1.0438661":{"http:\/\/vivoweb.org\/ontology\/core#departmentOrSchool":[{"value":"Applied Science, Faculty of","type":"literal","lang":"en"},{"value":"Electrical and Computer Engineering, Department of","type":"literal","lang":"en"}],"http:\/\/www.europeana.eu\/schemas\/edm\/dataProvider":[{"value":"DSpace","type":"literal","lang":"en"}],"https:\/\/open.library.ubc.ca\/terms#degreeCampus":[{"value":"UBCV","type":"literal","lang":"en"}],"http:\/\/purl.org\/dc\/terms\/creator":[{"value":"Pashaei, Parham","type":"literal","lang":"en"}],"http:\/\/purl.org\/dc\/terms\/issued":[{"value":"2024-01-12T22:39:15Z","type":"literal","lang":"en"},{"value":"2023","type":"literal","lang":"en"}],"http:\/\/vivoweb.org\/ontology\/core#relatedDegree":[{"value":"Doctor of Philosophy - PhD","type":"literal","lang":"en"}],"https:\/\/open.library.ubc.ca\/terms#degreeGrantor":[{"value":"University of British Columbia","type":"literal","lang":"en"}],"http:\/\/purl.org\/dc\/terms\/description":[{"value":"Two-dimensional materials are attractive choices for photovoltaic applications due to their unique material properties. Their extremely high absorption-to-thickness ratio supersedes conventional semiconducting crystals and their transparency and flexibility open new possibilities in photovoltaic applications. More importantly, van der Waals stacking allows the building of multi-stack solar cells. \r\nSeveral reports have demonstrated the potential of these materials for photovoltaic applications. However, the total efficiency of monolayer photovoltaic devices remains low. In this report, properties of multi-layer photovoltaic devices based on a van der Waals heterostructure diode are investigated. Our annealing method in ultra-high vacuum shows up to ~4x increase in current and our low-vacuum measurements show improvement in current and noise. Low-temperature measurement at 77K is conducted to understand the underlying working mechanism and for further explanation of photovoltaic behavior. These results are compared with reports of single-layer-based heterostructure photovoltaic diodes with similar conditions showing higher current density. To achieve these results, we optimize a process to produce exfoliated two-dimensional materials in a high-purity glovebox, design and build a stacking setup to fabricate heterostructures and optimize nanofabrication and measurement methods. \r\nFinally, we investigate the outlook for future devices by proposing NPN heterostructures. These findings open new discussions for further development of photovoltaic devices based on two-dimensional materials.","type":"literal","lang":"en"}],"http:\/\/www.europeana.eu\/schemas\/edm\/aggregatedCHO":[{"value":"https:\/\/circle.library.ubc.ca\/rest\/handle\/2429\/87251?expand=metadata","type":"literal","lang":"en"}],"http:\/\/www.w3.org\/2009\/08\/skos-reference\/skos.html#note":[{"value":"   OPTOELECTRONICS WITH TWO-DIMENSIONAL ATOMIC CRYSTALS  by PARHAM PASHAEI   B.Sc., Amirkabir University of Technology, 2013 M.Sc., University of Leuven-Chalmers University of Technology, 2015   A DISSERTATION 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 2023 \u00a9 Parham Pashaei, 2023ii  The following individuals certify that they have read, and recommend to the Faculty of Graduate and Postdoctoral Studies for acceptance, the dissertation entitled:   Optoelectronics with two-dimensional atomic crystals  submitted by Parham Pashaei  in partial fulfilment of the requirements for the degree of Doctor of Philosophy in Electrical and Computer Engineering  Examining Committee: Peyman Servati, Professor, Electrical and Computer Engineering, UBC Supervisor  John Madden, Professor, Electrical and Computer Engineering, UBC Supervisory Committee Member  Shahriar Mirabbasi, Professor, Electrical and Computer Engineering, UBC Supervisory Committee Member Mu Chiao, Professor, Mechanical Engineering, UBC University Examiner Michael Wolf, Professor, Department of Chemistry, UBC University Examiner  iii  Abstract  Two-dimensional materials are attractive choices for photovoltaic applications due to their unique material properties. Their extremely high absorption-to-thickness ratio supersedes conventional semiconducting crystals and their transparency and flexibility open new possibilities in photovoltaic applications. More importantly, van der Waals stacking allows the building of multi-stack solar cells.  Several reports have demonstrated the potential of these materials for photovoltaic applications. However, the total efficiency of monolayer photovoltaic devices remains low. In this report, properties of multi-layer photovoltaic devices based on a van der Waals heterostructure diode are investigated. Our annealing method in ultra-high vacuum shows up to ~4x increase in current and our low-vacuum measurements show improvement in current and noise. Low-temperature measurement at 77K is conducted to understand the underlying working mechanism and for further explanation of photovoltaic behavior. These results are compared with reports of single-layer-based heterostructure photovoltaic diodes with similar conditions showing higher current density. To achieve these results, we optimize a process to produce exfoliated two-dimensional materials in a high-purity glovebox, design and build a stacking setup to fabricate heterostructures and optimize nanofabrication and measurement methods.  Finally, we investigate the outlook for future devices by proposing NPN heterostructures. These findings open new discussions for further development of photovoltaic devices based on two-dimensional materials.   iv  Lay Summary  Providing a sustainable and environmentally friendly source of energy is considered as one of the most critical issues of the world. Solar cells are known as a potential solution to sustainable energy. However, fabrication challenges have halted their efficiency and the high cost of installation has limited their use. This research tests the promise of atomically thin materials for photovoltaic devices. These novel materials have extraordinary light absorbtion. Due to their extremely thin nature and unique stacking mechanism, they provide a new platform for flexible and high efficiency solar cells. The results of this work demonstrate that atomically thin materials can be effectively used as photovoltaic devices. It explains some of the challenges in their fabrication and demonstrates how to resolve them. Finally, it provides innovative approaches to cultivate their potential for multi-layer and multi-stack photovoltaic devices to transform solar cell engineering.   v  Preface  This dissertation is original, independent work by the author, Parham Pashaei.  Professor Peyman Servati supervised the project throughout its entire term.  Professor Joshua Folk advised on several parts of the project, such as contact optimization, device design and measurements. The exfoliation and heterostructure stacking in Chapter 3 are done with the help of Kevin Zhao.  The MATLAB simulation codes used in Chapter 2 are done in collaboration with Graham Allegretto. Part of the device fabrication in chapter 3.3, the measurement setup used in chapter 4, and the vacuum deposition in chapter 3.8.3. are done in Professor Joshua Folk\u2019s laboratory with the help of Dr. Ebrahim Sajjadi. Dr. Saeid Soltanian provided advice on metal deposition and the use of several tools: surface profiler in chapter 3.2.1., laser cutter, wet bench, doping, and the deposition system installed in the glove box. Dr. Silvia Folk advised on several experiments related to photolithography, deposition, and lift-off process. Raman spectroscopy was performed in Professor Maggie Xia\u2019s laboratory. Any figure obtained from published work is used with written permission from the publisher and is noted in figure captions. Professor John Madden and Professor Michael Wolf contributed to reviewing the dissertation and the project. Professor Andrea Damascelli advised on the scope of the project and spectroscopy methods at its early stages. Professor George Sawatzky advised on the project's scope at its early stage. Professor Pablo Jarillo-Herrero advised on some novel solutions for the projects in the outlook section.    vi  Table of Contents Abstract .......................................................................................................................................... iii Lay Summary ................................................................................................................................. iv Preface ............................................................................................................................................. v Table of Contents ........................................................................................................................... vi List of Tables ................................................................................................................................. ix List of Figures ................................................................................................................................. x Dedication ..................................................................................................................................... xv 1. Introduction ............................................................................................................................. 1 1.1. Overview of two-dimensional heterostructure photovoltaic devices .............................. 6 2. Simulations ............................................................................................................................ 10 2.1.1. Assumptions and simplifications .......................................................................... 12 2.1.2. Simulation results.................................................................................................. 13 2.2. Other Modeling and Simulations .................................................................................. 15 2.2.1. Modeling based on External Radiative Efficiency (ERE) .................................... 15 3. Experiments ........................................................................................................................... 16 3.1. 2D Materials .................................................................................................................. 17 3.1.1. Exfoliation ............................................................................................................. 17 3.1.2. Chemical Vapor Deposition .................................................................................. 19 3.1.3. Doped WSe2 .......................................................................................................... 21 3.2. Characterization ............................................................................................................ 21 3.2.1. Atomic Force Microscopy (AFM) ........................................................................ 22 3.2.2. Raman Spectroscopy ............................................................................................. 23 3.2.3. Infrared Atomic Force Microscope (AFM-IR) ..................................................... 23 3.2.1. Surface profiler ..................................................................................................... 24 3.3. Devices .......................................................................................................................... 27 3.3.1. Graphene\/MoS2\/Graphene Heterostructure .......................................................... 27 3.3.2. BN\/MoS2\/BN Heterostructure .............................................................................. 30 3.4. Etching .......................................................................................................................... 31 3.4.1. Etching Tools ........................................................................................................ 31 3.4.2. Boron Nitride etching ........................................................................................... 33 3.5. Experimental setup ........................................................................................................ 35 vii  3.5.1. 2D material stacking setup embedded in the glovebox ........................................ 35 3.5.2. Measurement setup ............................................................................................... 38 3.6. Other Tools ................................................................................................................... 41 3.7. Fabrication .................................................................................................................... 43 3.7.1. Preparing the stamps for assembly of two-dimensional material heterostructures43 3.7.2. Assembly of two-dimensional material heterostructures ...................................... 45 3.8. Metal contacts ............................................................................................................... 48 3.8.1. Metal Contacts for MoS2 ...................................................................................... 50 3.8.2. Pd Metal Contacts for WSe2 ................................................................................. 53 3.8.3. Deposition system ................................................................................................. 54 4. PN diode-based van der Waals heterostructures for photovoltaics ....................................... 59 4.1. Introduction ................................................................................................................... 59 4.2. Nanofabrication ............................................................................................................. 59 4.2.1. Fabrication of metal contacts ................................................................................ 59 4.3. Diode behavior results and discussion .......................................................................... 60 4.3.1. Diode behavior ...................................................................................................... 60 4.3.2. Vacuum effect on diode electrical response ......................................................... 61 4.3.3. High vacuum annealing effect on device performance ......................................... 62 4.4. Photovoltaic response ................................................................................................... 64 4.4.1. Photovoltaic results discussion ............................................................................. 68 4.4.2. Temperature-dependent electrical and photovoltaic response .............................. 75 4.4.3. Response time ....................................................................................................... 78 5. Novel van der Waals photovoltaic devices ............................................................................ 79 5.1. Multi-stack van der Waals photovoltaic devices .......................................................... 80 5.1.1. Introduction ........................................................................................................... 80 5.1.2. Theory and proposal ............................................................................................. 80 5.1.3. Experiment ............................................................................................................ 81 5.2. Flexible van der Waals photovoltaic devices ................................................................ 85 5.2.1. Introduction ........................................................................................................... 85 5.2.1. Fabrication of flexible devices .............................................................................. 85 6. Outlook .................................................................................................................................. 88 6.1. Van der Waals heterostructure solar cell ...................................................................... 88 viii  6.2. Thickness dependent study ........................................................................................... 90 6.3. Van der Waals heterostructure tandem solar cell ......................................................... 91 Bibliography ................................................................................................................................. 93 Appendix: Work done outside the scope of the thesis ................................................................ 101 A.1. Nanoparticle-decorated nanofibers for transparent conductors ....................................... 101      ix  List of Tables  Table 1-1. Summary of photovoltaic devices with atomically thin materials. The green cells show the areas of novelty and the red cells show the areas that need improvement. ..................... 4 Table 1-2. Summary of reported power conversion efficiency of 2DMs. ...................................... 8 Table 3-1. List of references for MoS2 devices with Al2O3, HfO2, Si3N4 and BN encapsulation for electronic and light-based studies. .......................................................................................... 31 Table 3-2. Results of measurements of the LED light source using Newport optical power\/energy meter model 842-PE. .................................................................................................................... 41     x  List of Figures   Figure 1-1. Solar spectrum matching. Cross section of 2D heterostructure solar cells consisting of various 2D materials stacked together can utilize a large part of the solar radiation spectrum resulting in high efficiency solar cells. ........................................................................................... 3 Figure 1-2. Summary of proposed structures for 2D heterostructure solar cells in cross section images. a) Schottky structure solar cells with semiconducting materials (MoS2, WSe2) and graphene. The bottom h-BN layer increases the efficiency of graphene b) Schottky structure enhanced by adding graphene conductive layer. c) Multilayer Schottky (top) and Multi-stack Schottky (bottom). d) p-n diode solar cell structure with two types of semiconducting 2D materials with graphene conductive layers at the top and the bottom. e) p-n diode structure solar cell in parallel mode. The holes are collected with top and bottom graphene conductive layer and the electrons are collected with the middle graphene conductive layer. f) p-n diode structure solar cell in series mode. The graphene (or h-BN) layer works as a tunnel barrier preventing formation of parasitic diode. ............................................................................................................................ 6 Figure 2-1. The simulations. a) The absorption spectrums of 2D materials extracted from experiments b) (green) solar irradiance and (blue) single layer MoS2 absorption. c) Transmitted and absorbed light by MoS2. d) (Blue) Band gap relaxation approximation. ............................... 11 Figure 2-2. (Left) EQE of the graphene-WSe2-MoS2-graphene heterostructure experimental results 19. Image used with permission from the publisher. (Right) Our simulation results for 2L-2L and the structure in the inset. ................................................................................................... 14 Figure 2-3. Estimation of PCE of TMDCs based on different EREs and comparison with other types of solar cells. 56 Reprinted (adapted) with permission from Jariwala, D., Davoyan, A. R., Wong, J. & Atwater, H. A. Van der Waals Materials for Atomically-Thin Photovoltaics: Promise and Outlook. ACS Photonics 4, 2962\u20132970 (2017). Copyright 2017 American Chemical Society........................................................................................................................................................ 15 Figure 3-1. MoS2 bulk crystal, WSe2 bulk crystal and tapes used to exfoliate 2D materials. ...... 18 Figure 3-2. AFM topography of MoS2 flakes. (a,b) The image library example to estimate flake thickness; (a. top) Optical microscopy image (a. bottom) AFM image. (b) Precise thickness of three lines, location of measured lines available in inset. (b) Contamination in the surface of the flake. .............................................................................................................................................. 22 Figure 3-3. Raman spectroscopy of atomically thin materials with optical microscope images in inset (Left) Raman graph of a 3-layer graphene. (Right) Three sets of Raman measurements of a 3-layer MoS2. Inset top left shows the typical graph for 1, 2 and 3 layers. .................................. 23 Figure 3-4. Nano-IR2-s Anasys instruments Atomic Force Microscope (AFM) located on an optical breadboard to avoid vibration. The glass opening is where the device and AFM tip are located. On the left side of the tool the tunable optical parametric oscillator and Q-switched pump (OPO) laser and CO2 laser systems are located. Below the optical table are power supplies and cooling pumps. ....................................................................................................................... 24 Figure 3-5. WSe2 flake sample. (Top) Profilometer image. (Bottom) AFM Images. .................. 25 Figure 3-6. WSe2 flake sample II. (Top) Profilometer image. (Bottom) AFM Images. ............... 26 xi  Figure 3-7. (Top) Cross-sectional schematic of the Graphene-MoS2-Graphene device. (Top left) Heterostructure after transfer of all layers. (Right) device after lithography and metal deposition. (Bottom left) device after several electrical measurements. ......................................................... 28 Figure 3-8. Room temperature electron transport measurements (a) Electron transport through Graphene\/MoS2\/Graphene: scan-down in red and scan-up in purple. OM image of the device in inset (b) Electron transport through graphene. The color of the graphs corresponds to the color of the markers of contacts in the inset. .............................................................................................. 29 Figure 3-9. Boron Nitride encapsulated MoS2 photodetector. (a) OM image of the heterostructure from top (b) AFM topography (c) AFM graphs of the lines indicated in figure b. (d) cross-sectional schematic of the final device. ........................................................................................ 30 Figure 3-10. PECVD system with load-lock used for etching. Image source: https:\/\/www.nanofab.ubc.ca\/ ........................................................................................................ 31 Figure 3-11. (Left) ECR machine. (Middle) knobs for various gases used in the machine. (Right) loading the sample using thermal paste. ....................................................................................... 32 Figure 3-12. AFM results of BN flakes (a) Table showing the etching rate of BN correlation with initial thickness (b) AFM topography (left) and profiles (right) before (top) and after (bottom) plasma treatment. .......................................................................................................................... 33 Figure 3-13. CAD schematic of the stamp holder. ....................................................................... 35 Figure 3-14. 2D material stacking setup mounted on an optical breadboard inside a glovebox. A monitor installed inside the glovebox connected to a CMOS camera allows viewing samples under the microscope. ................................................................................................................... 38 Figure 3-15. (Left) Custom-built measurement setup. (Right) LED photovoltaic measurement setups............................................................................................................................................. 39 Figure 3-16. PCB where the device and the LED are mounted before and after turning the LED on................................................................................................................................................... 40 Figure 3-17. (Left) Relative Spectral Power Distribution. Figure adopted from www.cree-led.com. (Right) Newport optical power\/energy meter model 842-PE located near the tip of the dipstick. ......................................................................................................................................... 40 Figure 3-18. 10-ports glovebox system with multiple tools such as deposition system (on left). 42 Figure 3-19. PE-50 plasma cleaner. The digital system on the top of the door allows for adjusting plasma parameters. The manual knobs at the top right allow manual adjusting of gas flow. Image source: https:\/\/www.nanofab.ubc.ca\/ ............................................................................................ 42 Figure 3-20. (Top) Tergeo plasma cleaner system with the chamber opening at left side of the system and the user interface on the right side. (Bottom) Schematic of direct and indirect plasma treatment options of the system. Image source: https:\/\/piescientific.com. ................................... 43 Figure 3-21. Process of preparing the transfer stamp. From left to right: Preparation of PDMS domes, transfer to glass slide and covering with PC. ................................................................... 45 Figure 3-22. Image of an early version of the stacking setup with top stage, stamp, heater, and motorized stage. ............................................................................................................................ 45 Figure 3-23. Process of locating the dome-shaped stamp on the layered material. Followed by heating the substrate to allow the stamp to cover the area of the layered material. ...................... 46 xii  Figure 3-24. The final stage of the transfer process. (left) The PC containing the flake is attached to the substrate by using elevated heat. (Right) Magnified image of the stamp, confirming the transfer of the desired flake. .......................................................................................................... 46 Figure 3-25. Contact resistance of 2D materials in monolayer and multilayer 88. Image used with permission from the publisher. ..................................................................................................... 48 Figure 3-26. Energy levels of TMDCs and metals. Image adapted from 90. Image used with permission from the publisher. ..................................................................................................... 49 Figure 3-27. The work function of metals and their energy difference with the conduction band of MoS2. Image adapted from 90. Image used with permission from the publisher. ..................... 50 Figure 3-28. Titanium deposition on MoS2 in high vacuum and ultra-high vacuum. Image adapted from 94. Reprinted (adapted) with permission from McDonnell, S., Smyth, C., Hinkle, C. L. & Wallace, R. M. MoS2-Titanium Contact Interface Reactions. ACS Appl. Mater. Interfaces 8, 8289\u20138294 (2016. Copyright 2016 American Chemical Society. ............................................ 50 Figure 3-29. Comparison of contact resistance (Rc) for different metals at ultra-high vacuum (\u223c10\u20139 Torr) and vacuum (~10-6 Torr) on MoS2 95. Reprinted (adapted) with permission from English, C. D., Shine, G., Dorgan, V. E., Saraswat, K. C. & Pop, E. Improved contacts to MoS2 transistors by ultra-high vacuum metal deposition. Nano Lett. 16, 3824\u20133830 (2016). Copyright 2016 American Chemical Society. ............................................................................................... 51 Figure 3-30. The conductivity of Ti\/Au and Al contacts on MoS2 99. Image used with permission from the publisher. ........................................................................................................................ 52 Figure 3-31.Effect of Ti\/Au deposition on Al on Id-Vg curves. (Left) Al (100nm) before deposition of Ti\/Au. (Right) After deposition of Ti\/Au (15\/75nm) 99. Image used with permission from the publisher. ........................................................................................................................ 52 Figure 3-32. Achieving Ohmic contact for High-quality MoS2 devices on hBN. Single-layer devices are contact resistance dominated (several M.ohm.um). Multilayer devices have much lower contact resistance. (Reference: Fig. 2.3.2, 2.4.1. from 100). ................................................ 53 Figure 3-33. AJA Ultra-high vacuum hybrid evaporator. The sample transfers to the ultra-high-vacuum main chamber (on the left) using a load-lock. The e-beam sources with six crucibles are located at the bottom of the chamber, and the four sputtering sources are located at the top. The red mount on the chamber has a handle that allows rotating the sample to face the required deposition source.  Image Source: https:\/\/www.nanofab.ubc.ca\/ ................................................. 54 Figure 3-34. Image of the e-beam evaporator embedded in glovebox. The e-beam evaporator is located by the side of the heterostructure stacking setup. ............................................................. 55 Figure 3-35. Images of the thermal evaporator. (Left) Inside of the evaporator where the three Alumina coated tungsten evaporation coil baskets are installed. (Middle) Manually reading the current using a clamp current reader. (Right) Agilent VS series Leak detector connected to the vacuum chamber. The Pirani gauge (the orange box on the left) is simultaneously connected to the chamber to read the vacuum level. .......................................................................................... 57 Figure 3-36. Angstrom PVD sputtering system used for the deposition of metals. The sample holder is capable of handling substrates up to 150mm in diameter and flexible substrates with a maximum width of 120mm. .......................................................................................................... 58 Figure 4-1. Microscopic images of two van der Waals MoS2-WSe2 samples. (Left) an early generation of devices with smaller contacts. (Right) Devices with larger contacts. .................... 60 xiii  Figure 4-2. Diode behavior of the device for Vds -4v to 4v and Vbg -60v to 60v. (Left) Ids-Vds graph for Vbg -60v to 60v (Right) Vbg-Vds graph. ..................................................................... 61 Figure 4-3. (Left) Measurement in atmosphere conditions before annealing. (Right) Measurement in rough vacuum before annealing. ........................................................................ 61 Figure 4-4. Top Left: Rough vacuum before annealing. Top right: After annealing (Vbg: 60v). 63 Figure 4-5. PV response at 300K in dark and LED power of 270, 460, 481, 938 W\/m2. (Left) Vds from -4v to 4v. (Right) Vds from -0.3v to 0.65v. .................................................................. 65 Figure 4-6. Output power in pW for 270, 460, 681 and 937 W\/m2 intensities of white input light for voltages ranging from 0 to 400mV. ........................................................................................ 65 Figure 4-7. Output power in pW for 270, 460, 681 and 937 W\/m2 intensities of white input light for different range of voltages. (left) ranging from -300 to 700mV.  ........................................... 69 Figure 4-8. I-V characteristics of a conventional p-n diode under dark and illumination. Unlike this study, typically, the illuminated current is smaller than the dark current for voltages higher than Voc. ....................................................................................................................................... 69 Figure 4-9. Output power in pW for 270, 460, 681 and 937 W\/m2 intensities of white input light for voltages ranging from 0 to 400mV. ........................................................................................ 71 Figure 4-10. Short circuit current, open circuit voltage, and fill factor for illumination under white light at room temperature. ................................................................................................... 72 Figure 4-11. Ref 1 (black dashed line): Data extracted from Lee et al. 19. Ref 2 (grey dashed line): Data extracted from Furchi et al. 92. .................................................................................... 73 Figure 4-12. Microscopic images of the heterostructures. (left) The heterostructure used in this study. (Middle) Heterostructure used by Lee et al. scale bar is 3um. (Right) Heterostructure used by Mueller group. .......................................................................................................................... 74 Figure 4-13. (Top left) 77K measurement with Vds from -5v to 5v and (Top right) room temperature measurement with Vds from -4v to 4v. (Bottom left) PV response with white illumination for Vds from -0.3v to 0.65v at 77K (Bottom right) similar measurement for room temperature with removal of back gate current. ........................................................................... 76 Figure 4-14. Diode behavior measured before annealing for Vbg:0_60v for Vds: -4_4v at (left) room temperature (right) 77K. ...................................................................................................... 77 Figure 4-15. (left) Photo response of MoS2-WSe2 heterostructure with Pd contacts to white light. (Right) Photo response of WSe2 phototransistor with Ti and Pd contacts demonstrating higher photo gain for Pd and faster response time for Pd 109. Reprinted (adapted) with permission from Liu, W. et al. Role of metal contacts in designing high-performance monolayer n-type WSe2 field effect transistors. Nano Lett. 13, 1983\u20131990 (2013). Copyright 2014 American Chemical Society. .......................................................................................................................................... 78 Figure 5-1. Proposed model for the electrical circuit of the bipolar junction solar cell. .............. 81 Figure 5-2. Microscopic image of a multi-stack heterostructure. ................................................. 82 Figure 5-3. Side view of MoS2-WSe2-MoS2 multi-stack heterostructure. .................................... 82 Figure 5-4. Optical Microscopy images of the van der Waals heterostructure. (Top Left) Image of the device after exfoliation. (Top Right) Design of the Ti\/Au and Pd\/Au contacts to MoS2 and WSe2. (Bottom Left) Image of the device after depositions. (Bottom Right) Magnified image of the device. ..................................................................................................................................... 83 xiv  Figure 5-5. (Left) PEN substrates are cut into smaller pieces using a laser cutter. (Right) Image of two flexible substrates on the right side and the flexible substrate mounted on a custom-cut aluminium holder on the left. A conductive SEM glue stick is used to attach the flexible substrate to the aluminium holder. ............................................................................................................... 86 Figure 5-6. (Middle) Flexible substrate attached to a chip holder using glue. In spite of the flexibility of the substrate, the wire bonding between the substrate and chip holder shows a proper connection. (Right) Image of two flexible substrates on the right side and the flexible substrate mounted on a custom-cut aluminium holder on the left. A conductive SEM glue stick is used to attach the flexible substrate to the aluminium holder. ...................................................... 87 Figure 6-1. Graphical image of the proposed device with Pt bottom contact, multi-layer TMDCs and Graphene top contact (a) Cross-sectional image of the device (b) 3D image of the device (c) band diagram of the interface. ...................................................................................................... 89 Figure 6-2. Van der Waals heterostructure tandem solar cell. ...................................................... 92 Figure A-1. Nanofibers coated with ZnO and MoO3 nanoparticles are treated in room temperature up to 300 C and transparency is measured at 550nm wavelength. ......................... 102 Figure A-2. Sheet resistance of Nanofibers coated with ZnO and MoO3 nanoparticles treated from room temperature (23 C) to 300 C. .................................................................................... 103    xv  Dedication          To the one and only Earth we live in  1   1. Introduction    Providing sustainable and environmentally friendly energy sources is considered as one of the world's most critical issues. Solar cells are a potential solution to provide green and sustainable energy. However, this assumption does not include the materials, energy and money used and spent during the production of traditional solar cells 1. Due to solar panels' large size and weight, their installation is challenging and expensive 2. Hence, the energy produced in the first few years of the lifetime of a solar panel only compensates for the costs and energy wastes during its production, with an average payback period of six to ten years 2. Designing a solar cell with less production cost, environmental impact, and easier installation can change this $100 billion industry and how we use solar cells 3. Several alternatives have been developed to achieve this goal, such as GaAs and organic or dye-sensitized solar cells 4, 5. Some use toxic materials, and others are far from the efficiencies of traditional solar panels. In this project, we study a new generation of solar cells that require up to 100 times less material than conventional solar cells. This can significantly decrease the cost and environmental impact on production. The extremely thin and light structure of these solar cells will eliminate installation issues, a major hindrance for the commercial use of solar cells 2. The discovery of graphene in 2004 6, highlighted by a Nobel prize 7 in 2010 and supported with \u20ac1 billion graphene flagship 8, attracted the world's attention to this interesting material due to its unique properties, including atomic thickness, high electrical conductivity, exceptional mechanical strength, flexibility and transparency 9, 10. More importantly, graphene introduced a 2  new form of materials known as two-dimensional (2D) materials 11,12. These materials are naturally only one atom thick and have unique, attractive properties. Soon after the introduction of graphene, several new 2D materials were discovered. A 2013 study showed that some of these 2D materials have extraordinary light absorption capability 13. They can absorb light 100 times better than the absorbers in conventional solar cells such as Silicon or GaAs. Light-matter interaction and photonics studies from 2014 to 2015 proved the exceptional capability of 2D materials in photovoltaic effect, generating electricity by photon absorption 14, 15, 16, 17. These pieces of evidence paved the way for introducing a new generation of solar cells. In this study, we demonstrate the photovoltaic effect in these materials. While building solar cells with one type of 2D materials is possible, efficiency could be increased by making more advanced structures. Chemically Vapor Deposited (CVD) Graphene on a copper substrate is grown as a part of this work. The multilayer CVD graphene is then transferred with PMMA coating to the target optoelectronic device. However, due to the lack of a band gap, graphene cannot be used as the light absorber of a solar cell. In contrast, 2D transition metal dichalcogenides (TMDCs) such as MoS2 are suitable for this purpose thanks to their intrinsic band gap. However, layered TMDCs lack some of the unique properties of graphene, such as exceptional electrical conductivity.  In order to make an efficient photovoltaic device, a combination of 2D materials that take advantage of their properties is needed. By combining these materials in the form of heterostructures, like a Lego structure with only one atom thick layers, a new field in the electronic industry is introduced as graphene Nobel laureates suggest 12. These heterostructures are only a few nanometers thick, and depending on the selection and combination of the 3  materials, they can be used for various applications such as sensors, transistors, memories, spintronics, and photonics 18, 19, 20, 21.   Figure 1-1. Solar spectrum matching. Cross section of 2D heterostructure solar cells consisting of various 2D materials stacked together can utilize a large part of the solar radiation spectrum resulting in high efficiency solar cells.  Each of these materials can absorb a specific wavelength of light with high efficiency 22, 23. Hence, by combining them, we can tune to absorb efficiently over a wider range of solar radiation spectrum. Making similar heterostructures or multi junctions with conventional semiconducting materials is extremely complicated, energy-consuming and expensive 24, 25. Hence, conventional semiconductor heterostructures are only used for limited applications and are limited to a few junctions. On the contrary, 2D materials naturally stack on top of each other with Van der Waals forces. Due to their extremely high surface-to-thickness ratio, these weak forces are enough to keep them together 12. With advancements in transfer technologies, heterostructures can be stacked on top of each other with mass production methods such as roll-to-roll transfer. One of the first van der Waals heterostructure solar cell proposals was based on a Schottky barrier configuration made of single layers of graphene and MoS2 similar to Fig 2. a. The Schottky barrier at the interface separates the electron-hole pairs, causing exciton dissociation without the requirement of an applied electric field. Although the power conversion efficiencies demonstrated 4  were below one percent, due to its extreme thinness, it was shown to have orders of magnitudes higher power densities than conventional solar cell technologies 13. A year later, Lee et al. demonstrated the first P-N diode solar cell via a van der Waals heterostructure. Graphene was used to collect holes and electrons from MoS2 and WSe2 monolayers, being intrinsically N-type and P-type, respectively. They measured External Quantum Efficiency (EQE) of 2.4%, demonstrating a very high efficiency-to-thickness ratio 19. They also identified tunneling between graphene layers as a significant loss mechanism. Our simulation study in Chapter 2 builds on Lee et al.\u2019s work by proposing and simulating structures that could further increase potential EQE.  Table 1-1 summarizes some of the work done on atomically thin structures with potential for solar cell applications, highlighting the areas that need more research in red. Table 1-1. Summary of photovoltaic devices with atomically thin materials. The green cells show the areas of novelty and the red cells show the areas that need improvement. Ref.  year Structure Scale Efficiency Diode Gate Materials Doping Method Size 26 2014 Lateral P-N  Dual gate monolayer WSe2 Electrostatic Exfoliate 10um light\u2013power conversion 0.5% electroluminescence 1% 26 2014 Vertical VdW back gate monolayer WSe2 Electrostatic Exfoliate 5um EQE \u223c 1.5% monolayer MoS2 light\u2013power conversion 0.2% 27 2013 Vertical VdW back gate  BN\/Gr\/WS2\/Gr\/BN Electrostatic Exfoliate WS2 10um EQE \u223c 30% Flex PET & SiO2 Gold nanoparticles CVD\/Flake Graphene Photocurrent 3.5uA 28 2014 MoS2\/SiO2\/Si heterojunction Back gate                       new design n-type MoS2          SiO2\/p-Si Electrostatic Exfoliated 10um EQE 4.5% 29 2014 p-Si\/MoS2\/Al back gate CVD MoS2, Al fingers,  P-Si backgate Electrostatic CVD  1 cm power conversion 5.23%                 EQE  ~60% 30 2013 Vertical VdW Top & Bottom SiO2\/G\/MoS2\/G\/HfO2\/G Glass\/ITO\/Al2O3\/G\/MoS2\/Ti  Electrostatic   top and bottom  Exfoliated ~5um EQE 55%                                               IQE 85% 5  31 2014 Lateral PN HfO2\/WSe2 Dual back gate          monolayer WSe2          HfO2 backgate Electrostatic Exfoliated ~5um EQE ~0.2%   Figure 1-2 depicts the proposed heterostructures. The Schottky barrier solar cell explained earlier can be further improved by adding a hexagonal Boron Nitride (h-BN) layer on the substrate to increase the conductivity of the graphene layer. Since graphene has higher conductivity than TMDCs, adding a graphene layer at the top of the semiconducting 2D material can also increase efficiency (Fig.1-2. b.).  By adding extra layers of monolayers, the band gap of the structure can be tuned to the solar spectrum. Stacks with different thicknesses and band gaps can build a solar cell with improved overall absorption (Fig.1-2.c). 6   A p-n junction solar cell consists of a p-type and n-type TMDC heterostructure with graphene at the top and bottom of the structure (Fig.1-2. d). Since these materials have different band gaps, by appropriate material selection, high efficiencies can be achieved. Stacking several p-n diodes using materials with different band gaps can increase absorption. These stacks can be connected in parallel (Fig.1-2 e) or series (Fig.1-2. f).  1.1. Overview of two-dimensional heterostructure photovoltaic devices Applications of crystalline semiconductors such as Silicon introduced a new era in the way we harvest light. Similarly, using 2D materials for photovoltaic applications can create new opportunities in energy harvesting. Since the discovery of graphene, two-dimensional materials have been used for electronic and optoelectronic applications such as transistors and sensors. Since 2014, several studies have shown the potential of 2D materials for photovoltaic Figure 1-2. Summary of proposed structures for 2D heterostructure solar cells in cross section images. a) Schottky structure solar cells with semiconducting materials (MoS2, WSe2) and graphene. The bottom h-BN layer increases the efficiency of graphene b) Schottky structure enhanced by adding graphene conductive layer. c) Multilayer Schottky (top) and Multi-stack Schottky (bottom). d) p-n diode solar cell structure with two types of semiconducting 2D materials with graphene conductive layers at the top and the bottom. e) p-n diode structure solar cell in parallel mode. The holes are collected with top and bottom graphene conductive layer and the electrons are collected with the middle graphene conductive layer. f) p-n diode structure solar cell in series mode. The graphene (or h-BN) layer works as a tunnel barrier preventing formation of parasitic diode. 7  applications by demonstrating single-type material devices 27, 29, 16. The maximum efficiency of these devices is capped by the Shockley-Queisser limit for single-type solar cells.  Photovoltaic devices, by incorporating two types of 2D materials, have been realized with the aim of increasing the efficiency of Field 32 and 19. By inserting a layered material in the middle of the p-n junction, they could reduce recombination by separating photo-excited holes and electrons 33, 34.  With developments in 2D materials fabrication, this improving trend can continue. Theoretically, achieving up to 68.2% efficiency is the maximum limit by having an infinite number of stacks for one sun illumination 35. In the past 40 years of crystalline semiconductor solar cells history, building up to three to four junctions has been hardly achieved. With 2D van der Waals heterostructures, the future seems to be much brighter. Two critical fabrication challenges are the main criteria for defining the future of this new category of solar cells: large-area material growth and vertical stacking of multi-junctions. Large area material growth typically refers to materials that have a larger area than conventional research grade exfoliated materials with micrometer scale area. Both of these challenges are rapidly improving. Large area growth of graphene ~400\u2009cm2 has already been achieved with reasonable quality 36 and several layered TMDCs have recently been grown in large area 37, 38. Making a large area single junction solar cell is doable with the current materials and technology and is considered one of the side goals of this work. The second challenge, making a large stack and transfer of 2D materials, is currently the major barrier for the evolution of 2D materials based solar cells. Recent advancements such as the hot pickup technique 39 and large area transfer 40 have solved this challenge to some extent. Yet, more investigations can directly affect the progress of 2D solar cells in gaining higher efficiencies.  8  In this table, we review some of the notable works on two-dimensional heterostructure photovoltaics.  Table 1-2. Summary of reported power conversion efficiency of 2DMs. Type Material Thickness Efficiency (%) References Materials Structure   nm layers Power conversion EQE Date Author Ref 2D  Single material Lateral PN MoSe2 11 10 6.3 (14)   2015 Memaran et al 41 Schottky (Metal) WSe2 100   (6.7)2   2015 Wi et al 42 Schottky (Graphene) WS2 37   3.3   2014 Shanmugam et al 43 Lateral Schottky  MoS2 ~50   2.5   2013 Fontana et al 44 Schottky (Metal-Au) MoS2 110-220   0.7-1.8   2012 Shanmugam et al 45 Vertical PN (Doped) MoS2 120   2.8   2014 Wi et al 46 Schottky (Cu-MoS2-In) MoS2 20u   (~1)1   1982 Fortin et al 47 Graphite-MoS2 MoS2 230-4u   -   1968 Evans & Thompson 48 Lateral PN WSe2   1 0.5   2014 Mueller et al 26 Lateral PN Phosphorene ~6-7   ~0.0005*   2014  Buscema et al 16 Multi Material Vertical vdW MoS2-WSe2   1 0.2 ~1.5 2014 Furchi et al 17 Vertical vdW MoSe2-WSe2   3-3 0.1* 0.1 2015 Flory et al 49 Vertical vdW Phosphorene-MoS2 11-1   ~0.006*   2014 Deng et al 50 Hybrid Vertical heterojunction MoS2-Si   1-Bulk 5.23 60 2014 Tsai et al 29 Nanocomposite MoS2\/TiO2    1.3   2012 Shanmugam et al 51 *Estimated based on the graphs presented in the paper 1 Measured at 120K. Not clear if ''intrinsic efficiency'' refers to power conversion efficiency 2 Measured under 532nm laser  In this work, we use a multi-material device structure to benefit from their variety of band gaps to increase effective light absorption. We chose a PN diode vertical van der Waals structures for several reasons. First, it allows using two materials with different bandgaps. Second, the PN diode structure is common in classical solar cell structures and is effective for charge separation. Third, the structure design aligns materials vertically allowing to absorb light by both layers. Forth, the structure design is scalable, and upon increasing the area of the device, the PN junction will expand with the area of this device. Other designs that have lateral structure have challenges in scaling as the PN junction area is too small compared to the area of the device and it doesn\u2019t scale proportionally. Fifth, this structure design allows expansion to larger stacks with more variety of materials. Experiments on stacks beyond two materials are explored in the last chapter of this thesis. 9  Since with increase in thickness the total light absorption of the material increases, and potentially the total efficiency could increase, we investigate designs with multilayer materials. One of the challenges with thicker layer materials is limitations with electrostatic doping due to screening. We have tested few methods to resolve these challenges. We demonstrate that higher current density is achieved compared to similar devices with single layer materials. During these experiments we found annealing and vacuum measurement methods that improve device performance significantly. In order to achieve competitive results, we designed and built a stacking setup in a high purity glove box environment. The low level of moisture and oxygen in the stacking setup installed in glovebox environment, the treatment of substrates prior to stacking, optimization of stacking process, and optimization of metal deposition and band gap engineering, enabled us to get these competitive results.     10  2. Simulations    To simulate the functioning of a two-dimensional solar cell and, ultimately, its theoretical EQE, a layer-by-layer approach is taken. The simulations are conducted via MATLAB using monolayer absorption spectrum data and monolayer bandgap data from 52. The model used as a basis for the simulations is shown below:  \ud835\udc38\ud835\udc44\ud835\udc38 = \u2211 \ud835\udc50! \u222b \ud835\udc34(\ud835\udc38)\ud835\udc38!,#(\ud835\udc38)\ud835\udc3f(\ud835\udc38)\ud835\udc51\ud835\udc38$%!&!'( \u222b \ud835\udc38),#(\ud835\udc38)\ud835\udc51\ud835\udc38$*$  For each layer, the incident solar spectral irradiance, Ei, \u03bb (E), is multiplied by the layers absorption profile, A (E). The incident solar spectral irradiance was taken from NREL and is the circumsolar derivation. The portion of the light that is transmitted by the layer is then applied to the following layer, which is multiplied by its own absorption profile and so on. The photons that are absorbed by the layer are then multiplied by the bandgap and relaxation-loss function, L (E). Any photons absorbed that are below the bandgap are assumed to be losses, while a portion of the energy of photons that are greater than the bandgap is also lost due to band-gap relaxation.  11   Figure 2-1. The simulations. a) The absorption spectrums of 2D materials extracted from experiments b) (green) solar irradiance and (blue) single layer MoS2 absorption. c) Transmitted and absorbed light by MoS2. d) (Blue) Band gap relaxation approximation. The final loss mechanism applied to the layer, ci, is due to the energy required to dissociate the photogenerated excitons to free-charge carriers via the p-n junction. This mechanism isn\u2019t inherent to the layer but the p-n junction itself. P-n junctions with a smaller built-in potential will have a lesser associated loss. Since the p-n junction is atomically thin, it is theorized that 12  materials arranged in a p-n junction can have a much smaller built-in potential compared to conventional solar cells while still achieving a high level of exciton dissociation efficiency. In the simulations, it was assumed that all excitons were dissociated to free-charge carriers. The same calculations are performed on each layer with a solar spectral irradiance attenuated by preceding layers. The portion of the incident photons that contribute to free-charge carriers for each layer is summed and divided by the applied irradiance spectrum, resulting in an estimated EQE. The results are discussed in subsequent sections. 2.1.1. Assumptions and simplifications Due to the infancy of two-dimensional materials, various mechanisms pertinent to the study have yet to be characterized. This has led to many assumptions and simplifications that would ultimately affect the estimated EQE. This section discusses these assumptions and how they would affect the results. The most significant assumption was the omission of reflectance. It was assumed that all incident light is either transmitted or absorbed, inflating the estimated EQE measurements. Future iterations of the simulation will take into consideration the complex dielectric constant of the monolayers to incorporate reflection. Once the light has been transmitted through the first layer, due to the high number of boundaries, internal reflection can be utilized to create a more efficient solar cell. Another critical assumption was that stacking monolayers does not affect their individual absorption profiles. It has been proven through simulations and experimental results that this is not true 13,52; stacking monolayers does not result in a superposition of the two absorption 13  profiles but as an intermediate between the two profiles. We also did not consider the change in bandgap that occurs as monolayers are stacked.  Tunneling that occurs between the sheets of graphene was omitted to simplify the simulation. The p-n diode structure allowed us to assume that 100% of all excitons are dissociated into holes and electrons and that no recombination occurs in the TMDC materials. We neglected to consider current matching and their associated losses.  2.1.2. Simulation results To validate the simulation, a solar cell consisting of single monolayers of MoS2 and WSe2 (one-layer-one-layer or 1L-1L) sandwiched between layers of graphene was created and compared to 19. The EQE results across the visible spectrum are shown in Fig.4. The plot obtained from the simulation of the 1L-1L cell is relatively close to the experimental results of Lee et al.\u2019s 2L-2L solar cell. Intuitively, the absorption of the 2L-2L cell should be close to double that of the 1L-1L, resulting in an EQE that\u2019s nearly double as well. The added benefit the extra layers provide to absorption seems to be completely counteracted by losses from tunneling, a loss mechanism suggested in 19, and, perhaps to a lesser extent, the change that occurs to absorption profiles when monolayers are stacked, which was not mentioned in Lee et al. 19 but had been studied by Wang et al. 53. Further work could be done to discriminate the two effects. The 1L-1L cell that was simulated neglecting tunneling and absorption profile changes has a similar EQE to the experimental results of 2L-2L by Lee et al. Due to the similarities between the two structures, the 1L-1L simulation is used to model 2L-2L structures in subsequent multi-junction simulations. The accuracy of the 1L-1L simulation to 2L-2L limits its application to strictly multi-junction solar cells using the 2L-2L p-n junction as a building block. 14    Figure 2-2. (Left) EQE of the graphene-WSe2-MoS2-graphene heterostructure experimental results 19. Image used with permission from the publisher. (Right) Our simulation results for 2L-2L and the structure in the inset. Once the initial p-n junction was validated, two different stacks\u2014a stack being either two layers of MoS2 and WS2 or MoSe2 and WSe2 sandwiched between graphene\u2014were used to simulate the solar cell. It is limited to two junction types due to a lack of experimental data and band gap engineering considerations. Once more data is produced on other two-dimensional TMDC, the simulation can be used for multi-junction structures. The most optimum solar cell tested that is relatively reasonable to be produced found to be two stacks made of WS2 and MoS2 followed by eight stacks of WSe2 and MoSe2 for a total of ten stacks resulting in an overall EQE of 36.68%. It was limited to ten stacks because anything higher would be highly unlikely to produce. The number of monolayers used for ten stacks is 51 resulting in a thickness of less than 100nm.   15  2.2. Other Modeling and Simulations 2.2.1. Modeling based on External Radiative Efficiency (ERE) External radiative efficiency (ERE) can be used to describe, compare and evaluate photovoltaic potentials of 2D material based solar cells. ERE is defined as the portion of the recombination current, mostly as dark current, that produce radiative emission from the device. 54 ERE depends on extrinsic parameters such as device design and intrinsic parameters such and materials properties, interface quality and materials quality. 55 Accurate experimental ERE values for 2D materials are not widely available at the time of this work, however, existing photoluminescence quantum yield (PLQY) values for 2D materials can be used to estimate the ERE. The maximum ERE is capped by Internal Radiative Efficiency (IRE), the portion of the recombination radiated internally. Due to similarities between the nature of IRE and PLQY, it is possible to the estimate IRE, and as a result ERE, based on PLQY values. Jariwala et al. 55 used these assumptions to estimate power conversion efficiencies of TMDCs based on different EREs and compared them with other types of solar cells such as perovskites, GaAs and CdTe.   Figure 2-3. Estimation of PCE of TMDCs based on different EREs and comparison with other types of solar cells. 56 Reprinted (adapted) with permission from Jariwala, D., Davoyan, A. R., Wong, J. & Atwater, H. A. Van der Waals Materials for Atomically-Thin Photovoltaics: Promise and Outlook. ACS Photonics 4, 2962\u20132970 (2017). Copyright 2017 American Chemical Society. 16   3. Experiments  In this chapter, different aspects of experiments, such as materials, characterization, fabrication, and measurements are discussed in several sections. By experimenting various exfoliation and annealing methods inspired by the literature as well as new recipes a process to produce exfoliated 2D materials in glovebox is optimized to achieve targeted size, thickness, and quality of flakes. Unlike most studies, we focus on multi-layer devices. Multi-layer flakes have higher absorption of light which could result in better photovoltaic performance. However, it is harder to electrostatically dope them. We use chemically doped materials and high back gate voltage to obtain diode effect with multi-layer materials and achieve competitive results explained in the next chapters. A home-made stacking setup is designed and built to fabricate heterostructures in high-purity glovebox condition. The stacking setup allows high-precision stacking at four degrees of freedom with high-purity, unique at the time of its implementation. This setup allows stacking of 2D materials with very high quality and precision. Thanks to its angle rotation experiments on angle of rotation of flakes could be possible. The high purity of the glovebox allows experiments on sensitive 2D materials. Ultra-high vacuum deposition setups are explained and tested to improve the quality of metal contacts.  High-work function metal deposition systems are tested and discussed to achieve better contact to WSe2.  17  3.1. 2D Materials The base material used in the device structures of this work is 2D materials with less than tens of nanometers thickness. Graphene, Molybdenum Disulfide (MoS2), Tungsten diselenide (WSe2), and Phosphorene are 2D materials that were explored in this work.  There are several methods to obtain single-layer or multi-layer 2D materials. Among these methods, only some provide the quality needed for electronic applications. Epitaxial growth 57, Chemical Vapour Deposition (CVD) 58 , and exfoliation are some of the methods that provide high-quality 2D materials. Exfoliation and CVD are two common methods that were experimented on because of their relevance to the applications needed in this work.  3.1.1. Exfoliation Mechanical exfoliation (also known as micromechanical cleavage and the Scotch tape method) has advantages that make it an attractive option for generating 2D materials among many research groups 59. Materials generated with this technique typically have a high quality 60, and a large variety of materials such as graphene, MoS2, WSe2 and phosphorene could be generated with a relatively similar procedure. This technique allows generating materials with various thicknesses, from monolayers to multilayers, without complications of other deposition techniques. However, it has some disadvantages. The thickness, size and sometimes the quality of the flakes are relatively random and non-reproducible. Acquiring a specific size, thickness, and quality of flakes may require several tries and spending a lot of time searching for flakes and characterizing them.  In order to optimize the exfoliation procedure and to get the highest possible quality flakes with desirable thickness and size, several variations of sample preparation and exfoliation were tried. 18  Cleanroom-grade blue tape, Scotch tape and PDMS were used for cleaving the bulk materials. PDMS was used from both commercial sources (Gel-packs) and made in the lab by mixing and curing. Normally a scotch tape (mother tape) is used to take flakes directly from the bulk material. Flakes on this tape usually are very thick and cover a large area of the tape. Then, either blue tape, scotch tape or PDMS was used on the initial mother tape to get thinner flakes. This process is repeated either on the same tape or by using a new tape for each step until the required thickness and size is reached. The rigidity of the exfoliation film and substrate both play a role in this process 61. A combination of exfoliation by PDMS on Scotch tape, blue tape on scotch tape and PDMS on PDMS was tried for the mechanical cleavage. To have more control over the flexibility and rigidity of the substrate and film, PDMS with a variety of thicknesses and curing temperatures were made. The number of repetitions of the cleavage differs for each material and flake size and thickness. For example, for MoS2 exfoliation with scotch tape, by cleaving more than seven times the number of large thin flakes decreased and lot of thin flakes looked broken or had lots of cracks. Six rounds of exfoliations with separate tapes without heating produced the most suitable results in our experiments in this example. Once adequate repetitions of the exfoliations are reached, the film is transferred to the substrate. SiO2 (285nm)\/Si, flexible PEN and PET were used as substrates.  Figure 3-1. MoS2 bulk crystal, WSe2 bulk crystal and tapes used to exfoliate 2D materials.  19  Heating the substrate, fast ultrasonication, oxygen plasma and a combination of them were tried to prepare the substrate to improve the exfoliation results. Two systems were tried for oxygen plasma treatment of the substrates. PE-50 plasma etcher with 15 cubic centimeters per meter oxygen flow and pressure of 250mTorr was used for 15 minutes. 3.1.2. Chemical Vapor Deposition To fabricate graphene, both Chemical Vapor Deposition (CVD) and Exfoliation methods were performed. Growing graphene with chemical deposition has several advantages, such as a larger surface area compared to exfoliation and control over the thickness. Single crystal graphene in the centimeter scale is fabricated repeatably by Hao et al. 62. Larger areas of the fabricated 2D materials enable more flexibility in the design and nanofabrication of the devices. For example, it is possible to design and fabricate multiple devices on a single graphene crystal. Since all the devices are built on the same flake, this method helps with better systematic comparison of the devices. The other advantage of this technique is the ability to produce multiple devices in one fabrication. This technique is used for graphene and other 2D materials such as MoS2 63.   To grow graphene using CVD method, a low-pressure chemical vapor deposition system is used. A 3-zone Carbolite TZF furnace capable of reaching a uniform growth temperature of 1000C is used for the growth process. Two gas cylinders containing Praxair 3.7 ultrahigh purity 99.97% of methane and Praxair 5.0 ultra-high purity 100.00% hydrogen are used as gas sources. Both cylinders are connected to CCR-400 MFC flow meters.  For the purpose of this fabrication, several pieces of 25um thick copper foils (Alfa-Aesar) are cut in 1x1 cm size to fit properly in the quartz tube. Based on the work by Polat el al. 36 the smoothness of the cooper foils is a very effective and low-cost method to fabricate large area 20  graphene with high-quality. Hence, ultra-smooth copper foils with 18um thickness (Taiwan Copper Foil, Co. Ltd, B1-SBS) were obtained. These foils have similar smoothness to the 20um foils (Mitsui mining and smelting co., LTD, B1-SBS) used by Polat et al. The copper foils are located in a quartz boat and inserted in the CVD furnace. Then, the furnace was pumped to a pressure of less than 10 mTorr. To remove contaminants and oxides from the copper surface, annealing with hydrogen flow was performed. The furnace temperature was raised to 1000 C under hydrogen gas flow for 30 minutes. Methane gas was used as the carbon source with 460mTorr pressure for 30 minutes. Finally, the camber was brought back to room temperature with 90 mTorr hydrogen pressure.  Graphene grown on copper can be characterized using various techniques such optical microscopy, scanning electron microscopy and Raman spectroscopy. For evaluating the electrical properties of the produced graphene or using it in electronic devices, the graphene needs to be transferred to a non-conductive substrate. Si with 285nm SiO2 thickness is a common choice for graphene substrate because of the dielectric properties of silicon oxide and the good visibility of graphene under an optical microscope. Transfer of graphene from copper to Si\/SiO2 substrate was done using a sacrificial PMMA layer. The purpose of the PMMA layer is to support graphene and avoid crumbling during the process of removal from copper and transfer to another substrate.  To achieve a smooth 500nm thick film, C4 PMMA (Microchem 950 PMMA C4) was spin coated at 3500rpm for 45 seconds on the copper\/graphene films. The films were then annealed for 1 minute at 180 \u00baC to remove the remaining solvents. To enable the transfer of the graphene, the copper needs to be etched away. A copper etchant solution was prepared, and the copper film 21  covered with graphene and PMMA was soaked in the etchant solution with the PMMA side facing up. When the copper film was completely etched, the graphene supported by the PMMA film was carefully transferred to a DI water bath. The Si\/SiO2 substrate is then inserted in the water, and the graphene\/PMMA film is gently captured with the substrate. The substrate is then dried up in vacuum overnight and annealed at 180 \u00baC for 30 minutes. Finally, the PMMA was washed away in warm acetone at 50 \u00baC for 2 hours. By this stage, graphene is transferred to Si\/SiO2 substrate. The substrate is then rinsed with acetone and IPA and blown dried with Nitrogen gas. 3.1.3. Doped WSe2  There are several methods reported for doping WSe2. Pudasaini et al. 64 perform hole doping by using remote oxygen plasma using an RF power of 400 W and O2 flow rate of 60 sccm with 15 mTorr chamber pressure at 150 \u00baC for 60s.  Zhao, P. et al. reached up to \u223c1019 cm\u20133 hole concentration by covalent functionalization using NO2 gas. 65  In this work WSe2 in bulk is provided commercially by 2dsemiconductors. The dopant is Nb and the doping range is 1017 cm\u20133.  3.2. Characterization The (opto)-electronic properties of 2D materials highly depends on their thickness, quality and contamination22. Hence, characterizing these materials is important in understanding their (opto) electronic behavior. In this chapter, we study these materials with various characterization techniques such as atomic force microscopy, Raman Spectroscopy, photoluminescence and optical microscopy.  22  Figure 3-2. AFM topography of MoS2 flakes. (a,b) The image library example to estimate flake thickness; (a. top) Optical microscopy image (a. bottom) AFM image. (b) Precise thickness of three lines, location of measured lines available in inset. (b) Contamination in the surface of the flake. 3.2.1. Atomic Force Microscopy (AFM) In order to understand the thickness and study the contamination of the exfoliated materials AFM measurements is performed. For instance, in the discussion chapter, the etch rate of Boron Nitride is studied with AFM. To simplify estimating the thickness of flakes, a library of optical microscopy (OM) images matched with AFM mapping is created. By using this library, it is possible to quickly estimate the thickness of the material by correlating the color and transparency in OM pictures with AFM data in the library. By studying a wider range of flakes with various thicknesses, it is possible to precisely identify the thickness of 2D materials with OM 66, 67. High-resolution AFM scanning can also show contamination in the form of residues on the top surface. Previous studies show thermal annealing can reduce these contaminations 68, 69. Hence, we included thermal annealing in the fabrication, which is further explained in the fabrication chapter.                         23  3.2.2. Raman Spectroscopy Raman spectroscopy can provide information about the quality, crystallinity, and thickness of 2D materials 70. We performed Raman measurement on the exfoliated flakes of graphene and MoS2 with an HR-800 Horiba Raman tool. The measurements are performed before device fabrication to minimize the effect of metal contacts in the measurement. In graphene, the ratio of 2D to G Raman peaks and their sharpness defines the thickness of the layer 70. Based on the 2D\/G ratio available from the previous works the number of layers in the flakes are estimated. In MoS2, the distance of the peaks and their intensity show the number of MoS2 layers. In this experiment, the MoS2 is expected to be over three layers. These measurements will be useful in proofing the thickness, quality, and crystallinity of the flakes.    Figure 3-3. Raman spectroscopy of atomically thin materials with optical microscope images in inset (Left) Raman graph of a 3-layer graphene. (Right) Three sets of Raman measurements of a 3-layer MoS2. Inset top left shows the typical graph for 1, 2 and 3 layers. 3.2.3. Infrared Atomic Force Microscope (AFM-IR)  A Nano-IR2-s Anasys Instruments Atomic Force Microscope (AFM) with Nanoscale Infrared (IR) Spectroscopy and Scanning Near-field Optical Microscopy (SNOM). The tool can perform simultaneous high-resolution AFM imaging and spectroscopy with spatial resolution in the nm range. It includes a CO2 laser for scattering SNOM applications, as well as a tunable infrared 24  Optical Parametric Oscillator and Q-switched pump (OPO) laser source with a tuning range of 900 \u2013 3600 cm-1.  Figure 3-4. Nano-IR2-s Anasys instruments Atomic Force Microscope (AFM) located on an optical breadboard to avoid vibration. The glass opening is where the device and AFM tip are located. On the left side of the tool the tunable optical parametric oscillator and Q-switched pump (OPO) laser and CO2 laser systems are located. Below the optical table are power supplies and cooling pumps. 3.2.1. Surface profiler The use of an optical surface profiler for 2D materials is studied in this section. A Filmetrics 3D profilometer is used for evaluating the thickness of 2D material flakes. The advantage of this system is enabling fast and easy surface spectroscopy. Ease of installing and locating the sample, fast imaging and tip-less operation compared to AFM are other advantages of this system, which makes it attractive for 2D materials rapid spectroscopy.  WSe2 flakes produced with the tape method are used for testing the setup. The thickness of the flakes is measured with an AFM for comparison. In a test on a multi-layer WSe2 with up to 30x10um dimension, the Profilometer shows an average thickness of 25nm. However, the AFM measurement performed on three areas of the device resulted in 8-10nm thickness.  25       Figure 3-5. WSe2 flake sample. (Top) Profilometer image. (Bottom) AFM Images. In another test on a WSe2 flake, the profilometer measures 52nm average thickness. The AFM measurement on two lines located near the same area measured by the profilometer show 15nm average thickness.  26   Figure 3-6. WSe2 flake sample II. (Top) Profilometer image. (Bottom) AFM Images. The profilometer results are significantly different from AFM spectroscopy in several measurements on samples with different sizes and thicknesses. In some measurements there is more than 3 times larger thickness measurement in profilometer results compared to AFM. AFM spectroscopy has established reliable results for thin layers and is considered as a standard measurement for thin materials. Due to the major differences in profilometer thickness measurement results with standard AFM spectroscopy for thin layers, its inconsistency with optical imaging results and existing studies in the literature, the surface profiler was not used for measuring the thickness of 2D flakes in this work. 27  Measurements on thin films with larger than 30um dimension on each side with thickness of 25nm and larger show much more reliable results. Note that the flakes in this study are as narrow as 5um in certain areas. Hence, AFM measurement is used in this study due the smaller thickness and surface area of the flakes. 3.3. Devices 3.3.1. Graphene\/MoS2\/Graphene Heterostructure The first successfully stacked heterostructure consists of Graphene-MoS2-Graphene (GMG) layers. This device enabled testing stacking mechanism and the measurement setup by performing electron transport measurements. The fabrication process is as below: \u2022 Cleaning with Acetone and IPA \u2022 Spin coating PMMA C4 & A4 at 4000rpm for 55sec \u2022 Electron beam lithography \u2022 Development of resist \u2022 30s oxygen plasma \u2022 5nm Cr-100nm Au deposition with a thermal evaporator 28   Figure 3-7. (Top) Cross-sectional schematic of the Graphene-MoS2-Graphene device. (Top left) Heterostructure after transfer of all layers. (Right) device after lithography and metal deposition. (Bottom left) device after several electrical measurements.  The electron transport measurements show the graphene contacts working properly, having the typical graphene Dirac curve by applying back gate voltage. The MoS2\/graphene interface, however, was not appearing in the measurements. Further investigations showed graphene flakes on both sides are connected, resulting in electrically bypassing the MoS2 flake. Another run of lithography and etching was performed to remove the excess graphene which resulted in damaging the device. Hence a new device with a modified design was fabricated. Unlike the first device, where graphene was in contact with MoS2 from top and bottom, in the second GMG heterostructure, both graphene flakes are located at the bottom to avoid uneven doping effect from the substrate. 29    Figure 3-8. Room temperature electron transport measurements (a) Electron transport through Graphene\/MoS2\/Graphene: scan-down in red and scan-up in purple. OM image of the device in inset (b) Electron transport through graphene. The color of the graphs corresponds to the color of the markers of contacts in the inset. Room temperature back-gated R-V curves are shown in figure 9. The back gate is applied from the bottom of the chip with 285nm SiO2 as dielectric. Figure 9.a shows the R-V curve for current applied through Graphene\/MoS2\/Graphene. The increase in resistance at low back gate voltages matches the natural doping of MoS2. Figure 9.b. shows the R-V curve for both graphene flakes. The Dirac peak location and sharpness is different for the flakes since they have different thickness. Photoresponsivity is one of the main characteristics of a photodetector which is the electrical output current divided by input optical power expressed in A\/W unit. Photocurrent measurement of this device shows Photoresponsivity of 2700 A\/W at 500nm wavelength illumination. This result is roughly three orders of magnitude higher than the first MoS2 photodetector report by Yin et al. 72, but it is still roughly four orders of magnitude lower than graphene encapsulated device claimed by Zhang et al. 20. However, due to the dominance of the noise in the results, more measurements are needed to confirm the accuracy of the reported Photoresponsivity.  30  3.3.2. BN\/MoS2\/BN Heterostructure Encapsulation of 2D materials can improve their electron transport properties. While encapsulation effects on electron transport is studied in devices encapsulated with Al2O3 73,74, HfO2 75, Si3N4 76 and BN 77, the optoelectronic studies are less developed. Kufer and Konstantatos studied photodetection properties of encapsulated MoS2 with Atomic layer deposition (ALD) of HfO2, showing better response, vanishing hysteresis and better mobility 78. However, the optoelectronic properties of a 2D material encapsulated device are unknown to our knowledge as of wiring this thesis. In this structure, we will make a MoS2-based photodetector fully encapsulated in BN as shown in Fig 10. While ALD method only covers the top part of the active area, our device protects MoS2 from both sides. Furthermore, the bottom BN can improve electron transport 77. Since the MoS2 is covered with BN from the top, in order to deposit electrical contacts on MoS2, BN layer on the top needs to be etched.  Figure 3-9. Boron Nitride encapsulated MoS2 photodetector. (a) OM image of the heterostructure from top (b) AFM topography (c) AFM graphs of the lines indicated in figure b. (d) cross-sectional schematic of the final device.  31    Al2O3 HfO2 Si3N4 BN MoS2 Electron 74, 73 79 76 77 Light - 78 - - Table 3-1. List of references for MoS2 devices with Al2O3, HfO2, Si3N4 and BN encapsulation for electronic and light-based studies.  3.4. Etching 3.4.1. Etching Tools PECVD   A Trion Orion III Plasma Enhanced Chemical Vapor Deposition (PECVD) system with load-lock is used for some of the etching experiments. The system can be used as a RIE etcher for CF4\/O2 and for O2 ashing. To obtain consistent results, it is important to maintain a clean chamber. Hence, the reactor was cleaned with cleanroom wipes and DI water, dried with cleanroom wipes and IPA. The cleaning was followed by plasma cleaning the chamber with CF4 and O2 for 30 minutes.  Figure 3-10. PECVD system with load-lock used for etching. Image source: https:\/\/www.nanofab.ubc.ca\/   32  ECR (Electron Cyclotron Resonance) Plasma Etcher  A PLASMAQUEST: ECR plasma etcher was used for etching BN flakes. The system uses CF4, CHF3, O2 and Ar. For etching BN, O2 and CHF3 gases were used. The system can also be used to etch SOI, metals and other materials.        Figure 3-11. (Left) ECR machine. (Middle) knobs for various gases used in the machine. (Right) loading the sample using thermal paste. The system is used with the following process. First, the system cooler was activated. Then the knobs for O2 and CHF3 flow were manually opened. The plasma fuse is turned on, followed by turning on the plasma. The lock-in is then vented to load the sample. A syringe is used to put a very small drop of thermal paste to attach the sample. Microwave reflection needs to be adjusted to less than 10% using the three knobs on the system. The RF is turned on and the power knob is adjusted to reach 200v. The CHF3-O2 recipe with 40sccm CHF3 and 4sccm O2 and 200watts microwave power for 10 seconds is used for etching the BN. Finally, the process is finished by purging nitrogen, cleaning the sample holder with IPA, turning the cooler off, and turning the main switches of the plasma and microwave off.    33  3.4.2. Boron Nitride etching Boron nitride (BN) can be used to encapsulate atomically thin devices, protect them from the environment and reduce their charge traps 77. One approach is to cover the device on the top side with a BN flake. In order to access the electrical contacts to the active area of the device, the top BN flake needs to be etched in some areas. This should be done with minimum etching on the active area of the device. Hence, a good understanding of the etching rate of BN and the underlying active material is essential. BN can be slowly etched with O2 plasma or rapidly etched with a combination of CHF3 and O2 plasma 77. Since precise etching is needed for the purpose of this work, oxygen plasma with 200W power and 40sccm O2 is used with the Trion PECVD machine. Several BN flakes are isolated on a SiO2\/Si substrate and plasma treated in the same condition, followed by AFM measurement to identify the thickness.  Figure 3-12. AFM results of BN flakes (a) Table showing the etching rate of BN correlation with initial thickness (b) AFM topography (left) and profiles (right) before (top) and after (bottom) plasma treatment. Plasma treatment was applied on flakes from 8 up to 160 nanometers for 60, 45 and 20 seconds. The results show that the etching rate is related to the initial thickness of the flake, i.e., the 34  thicker the flake, the higher the etch rate. However, the reason behind this behavior and a clear correlation between thickness and etching rate needs more investigation. The experiments also show the etching rate significantly slows down in very short plasma times of 20 seconds. As a result, decreasing plasma power is suggested to achieve very low etching rates.   35  3.5. Experimental setup  3.5.1. 2D material stacking setup embedded in the glovebox In order to fabricate 2D material heterostructures, a stacking setup that meets the requirements of this project is built. The stacking setup has the following characteristics: - An optical breadboard for mounting the stages and the microscope with low vibration, which includes: o Newport M-PG-22-4-ML Modal damped 600x600x110mm M6 optical breadboard with microlocks. o Four Newport ND60-A microlock compatible elastomeric isolation mount passive dampers to minimize vibrations. o Custom-made machined shopped steel brackets and 80\/20 brackets for securing the microscope and controller.    Figure 3-13. CAD schematic of the stamp holder. - A stage for mounting the sample with the first 2D material which includes: o High precision X-Y motors for sub-micron control over the location of the first material. 36  o Z direction motor for focusing.  o Two Thorlabs 19W ring shape ceramic heaters with 23mm external diameter used for heating the stage to 200 C for the transport mechanism. o A heat isolation design with mica layers limits the heat transfer from the stage to the motors. o A custom-made CNC cut circular cupper stage with two 23mm circular openings for ceramic heaters designed for uniform distribution of heat all over the sample.  - A stage that can hold the transfer stamp located at the top of the mounting stage which includes: o  High precision X-Y-Z motors for sub-micron control over the location of the stamp holder o Machined steel frame for holding the glass stamp.  o Magnets for easy attachment of the stamp in the glovebox. o Polymer-based support at the back of stamp holder to damp vibrations of the stamp. - A microscope with high resolution and high stability for viewing flakes as small as a few microns, which includes: o A set of 5x, 20x, 50x and 100x high focal length apochromatic bright field objectives with 34mm, 20mm, 13mm and 6mm working distances.  The wider 50x and 20x objectives are for locating flakes and the higher magnification 50x and 100x objectives are used for fine alignment of the flakes. 37  All microscope objectives are high focal distance to allow space for the transparent stamp between the sample and the objective. o A 0.5x adapter is used for attaching the imaging sensor. o A 3.1 Mpixels CMOS sensor with 3.2\u03bcm x 3.2\u03bcm for digital imaging. o A computer for remote viewing and control. o The microscope does not have a binocular to minimize weight and hence vibration. - A power supply and controller for powering and controlling the temperature of the heater. o Omega CN63200-R1-AL PID temperature and process controller. A glovebox custom-designed glovebox is used for storing the stacking setup. The stacking setup is mounted in a glove box for several reasons. First, some 2D materials such as Phosphorene and WTe2 are very sensitive to ambient conditions, and they degrade when exposed to air 80. Second, the air and moisture in the environment can deposit in the heterostructure interface and degrade the quality of the interface. Third, even 2D materials that are not known to be very sensitive to ambient conditions degrade when exposed to air for longer periods of time 81. Hence, mounting the stacking setup in an environment where there is low oxygen, water and other contaminants is preferable for high quality devices and necessary for some 2D materials. In order to provide these conditions, we prepared a glovebox with less than 0.1ppm Oxygen level and 0.5ppm Moisture levels. Two sections of the glovebox are dedicated to 2D materials; one is prepared for exfoliation and storage and the other where the stacking setup is located.  To minimize exposure to air, equipment needed for the fabrication of the devices is also located in the glove box where possible. This equipment includes an electron beam evaporation system for the deposition of gold. This system is located in the same glove box that hosts the stacking 38  setup. The second glovebox includes an Atomic Layer Deposition (ALD) and a spin coater. The second glovebox has the same atmospheric conditions and is interconnected via interlocks with the first glovebox. The ALD system is useful for the deposition of oxides such as Al2O3 and Hf2O3. As explained in the earlier sections these oxides are used for the encapsulation of 2D materials to protect them from contaminants in the air.   Figure 3-14. 2D material stacking setup mounted on an optical breadboard inside a glovebox. A monitor installed inside the glovebox connected to a CMOS camera allows viewing samples under the microscope.  3.5.2. Measurement setup Photovoltaic measurement setups  Several measurement setups were tested for the photovoltaic measurements of the devices. The first one consists of a white light source with a 150W Xenon lamp. It is equipped with an AM 1.5G filter that provides light output similar to the solar spectrum. A monochromator allows breaking the light into different frequencies to measure frequency response. The system is 39  installed on an optical breadboard for stability. A Newport optical power\/energy meter model 842-PE is installed on the sample holder for calibration and light measurement.   The second measurement setup uses 19 individual LEDs as a power source. The tunability of power is one of the advantages of this system.  Figure 3-15. (Left) Custom-built measurement setup. (Right) LED photovoltaic measurement setups.  Custom-built measurement setup  The measurement setup has the following requirements: \u2022 Electrical measurement of the current and voltage down to a few nAs. \u2022 Vacuum condition to compare the results in ambient condition and vacuum. \u2022 Integration of light source for light response measurements. \u2022 Capability for low-temperature measurement. A dipstick, also known as a dunker, is used as the main platform for electrical measurement. The dipstick allows vacuum and low temperature measurement with low electrical noise. Two Keithley 4200 source meters are used as the back gate and drain-source voltage sources. The 40  chip is located at the end of the dipstick on a PCB that connects the chip to the measurement setup wirings.    Figure 3-16. PCB where the device and the LED are mounted before and after turning the LED on. High-power white light LEDs are used as the light source for this experiment. An LED with 810 (780lm ~ 840lm) Flux @ 85\u00b0C, 2700K CCT, and forward voltage of 36v with 113 lumens\/Watt from Cree Inc are used 82, 83.       Figure 3-17. (Left) Relative Spectral Power Distribution. Figure adopted from www.cree-led.com. (Right) Newport optical power\/energy meter model 842-PE located near the tip of the dipstick.   A PCB is attached to the end of the dipstick for vacuum measurements. Surface-mount LED light sources are installed on the PCB near the chip. In the initial measurements, red and green light LEDs were tested. For the measurements presented in this article, a white LED has been used. To measure the power of the LED more accurately, the output power is measured using a 41  Newport optical power\/energy meter model 842-PE. The Input voltage, current and power are presented in the table below. The power density is based on 113x10-6 m2 sensor area. Table 3-2. Results of measurements of the LED light source using Newport optical power\/energy meter model 842-PE.  LED V LED mA Input power mW power reading mW W\/m2 * 30 0.19 5.7 1.5 13.274 31 0.494 15.314 3.5 30.973 32 0.87 27.84 8 70.796 33 1.28 42.24 10 88.495 36 2.6 93.6 20 176.991 39 3.99 155.61 30.5 269.911 40 4.48 179.2 38.2 338.053 44 6.34 278.96 52 460.176 50 9.18 459 77 681.415 60 13.96 837.6 106 938.053 * Based on 113x10-6m^2 sensor area  3.6. Other Tools Glovebox  To handle air sensitive 2D materials and stack them with higher quality by avoiding oxygen and moisture, a controlled environmental condition may be required. N2 and Ar glovebox systems have been used by research groups to provide such conditions 84, 85. For this study, a custom-made glovebox system was developed, which consists of 10-ports glovebox system in two interconnected sections with a dedicated 2D materials exfoliation area, a locally made motorized heterostructure stacking setup, vacuum deposition system, spin coater, and an atomic layer deposition system. The system uses N2 gas with below 0.1ppm Oxygen level and below 0.5ppm moisture level which provides excellent condition for working with 2D materials. 42   Figure 3-18. 10-ports glovebox system with multiple tools such as deposition system (on left).  Plasma cleaner  A PE-50 plasma etcher is used for cleaning and preparing the samples. The system uses Ar and O2 gases. O2 gas is used for cleaning the substrates in this study. Based on the sample, typically 100W of power is used for 10-15 minutes with an oxygen flow rate of 15 cubic centimeters per minute, providing 250mTorr pressure. The gas flow is controlled manually by using adjustment knobs.  Figure 3-19. PE-50 plasma cleaner. The digital system on the top of the door allows for adjusting plasma parameters. The manual knobs at the top right allow manual adjusting of gas flow. Image source: https:\/\/www.nanofab.ubc.ca\/ Another system that is used for sample cleaning in this study is Tergeo plasma cleaner. The indirect, downstream mode option of the system is suitable for gentle surface cleaning of 2D materials and PDMS preparation. In this study, the direct immersion mode is used for cleaning 43  Si\/SiO2 substrates. The indirect mode is used for removing resist residues and creating oxygen on the surface of 2D materials. For cleaning 2D materials, two recipes were used. The first one is 50w plasma power with 15sccm oxygen flow for 5 seconds and the second recipe is 10 seconds at fluctuating mode with 2sccm oxygen flow. For cleaning Si\/SiO2 samples, direct mode with 6sccm oxygen flow for 1-5 minutes is used.      Figure 3-20. (Top) Tergeo plasma cleaner system with the chamber opening at left side of the system and the user interface on the right side. (Bottom) Schematic of direct and indirect plasma treatment options of the system. Image source: https:\/\/piescientific.com. 3.7. Fabrication 3.7.1. Preparing the stamps for assembly of two-dimensional material heterostructures After the exfoliation of flakes, they need to be stacked carefully to form the desired van der Waals heterostructures. The stacking process should be done with minimal contamination between the layers. It is also important to apply a process that prevents significant damage to the layers and minimizes the formation of cracks on the flakes. To stack the exfoliated flakes, we use 44  the dry transfer technique 86. The dry transfer technique uses a transparent polymer to align and pick up the first flake and locate and release it on the second flake using heat and pressure.  Multiple methods with different recipes for polymers and shapes were experimented as described here. One of the methods relies on using a dome-shaped PDMS covered with polycarbonate 87. In this method, a SylgardTM 184 silicone elastomer kit is used by mixing the agents with 10:1 molar ratio is used to prepare the PDMS for domes shaped base layer. The process of mixing the polymers introduces bubbles that can influence the final shape of domes in undesirable ways. Hence, the bubbles should be removed to ensure a smooth surface and shape of the final domes. To remove the bubbles, the mixture is vacuumed by locating it at a desiccator connected to a vacuum until the bubbles disappear from the mixture. The mixture is then poured on a Petri dish using a pipette to form dome shaped polymers. The PDMS domes are then heated at 40o C overnight. Next, Polycarbonate (PC) films were made by making a 7% by weight solution of Poly (Bisphenol A carbonate) in Chloroform. The solution is mixed with a magnetic stirrer in a chloroform safe wetbench overnight. Using a pipette, a small amount the solution is transferred to a glass slide. Another glass is pressed and slid on the top of the film to make a thin film of PC. The film is left in the wetbench for 30 minutes to dry.   Once the PDMS domes and the PC films are ready, they need to be assembled on the final stacking stamp. A hole-punched double-sided Kapton tape is attached to a glass slide and a PDMS dome is moved to the center of the tap. The PC film is cut into smaller pieces, carefully removed from the glass substrate, and located at the top of the PDMS dome. The tape helps with attaching the PC film on the top of the dome. Finally, the excess amount of the PC film is cut and 45  removed. This process might introduce wrinkles between the PC film and the dome. To remove these wrinkles, the stamp was annealed for few minutes at 210o C.  Figure 3-21. Process of preparing the transfer stamp. From left to right: Preparation of PDMS domes, transfer to glass slide and covering with PC. 3.7.2. Assembly of two-dimensional material heterostructures The transfer stamp discussed in the previous section was used to pick up and transfer flakes using the setup shown in the figure below. The setup has an in-house made machined top stage that secures the glass slide that contains the stamp. The motorized stages help with precisely locating the flakes. The microscope allows seeing through the stamp and transferring the flakes at the desired target location and the adjustable heater helps with releasing or picking-up flakes, as discussed below.  Figure 3-22. Image of an early version of the stacking setup with top stage, stamp, heater, and motorized stage. The first step in the assembly is picking up the desired flake. Under the microscope, an area close to the center of the dome is selected and, using the motorized stage, located near the flake. Then, the substrate is heated to 90o C for most types of flakes. The transfer stamp is slowly 46  moved down until the dome gently touches the substrate near the flake. To cover the flake with PC, the temperature of the substrate is slowly increased up to 120o C while monitoring the expansion of the PC until the flake is fully covered by the stamp. This process is followed by a roughly 1-minute wait to assure that the flake is properly covered with PC. To pick up the flake with the stamp, the stamp needs to reverse back, so the dome shape allows picking up the flake. Hence, the temperature is set back to room temperature by turning off the heater and using a fan to allow cooling down the stamp and reversing it back to its initial shape. By reaching this stage, the flake should be picked up by the stamp and is ready for transfer to the target substrate or picking up another flake using the same procedure.    Figure 3-23. Process of locating the dome-shaped stamp on the layered material. Followed by heating the substrate to allow the stamp to cover the area of the layered material. Once all the layers of the heterostructure are picked up by the stamp, they were transferred to a silicon (SiO2\/Si) substrate by locating the stamp on the target substrate and leaving it at 180o C until the PC is attached to the substrate as shown in the figure below.     Figure 3-24. The final stage of the transfer process. (left) The PC containing the flake is attached to the substrate by using elevated heat. (Right) Magnified image of the stamp, confirming the transfer of the desired flake. 47  Finally, the PC is washed away by dipping the sample in chloroform for roughly 30 minutes, and the sample is checked for remaining PC residues and the condition of flakes. Sometimes during this process, the flakes were removed from the sample, and the process had to be repeated from the beginning. Once the condition of the flakes was confirmed, the sample was rinsed in IPA for 5 minutes to remove the remaining chloroform and potential contamination.    48  3.8. Metal contacts In this section, we discuss the choices for metal contacts to 2D materials used in this study. First, we start with an introduction to the effect of metal contacts on 2D materials.  The majority of studies on optoelectronic and photovoltaic properties of 2D materials experiment on single-layer materials. In this study, the Multi-layer van der Waals materials have experimented. There is a significant difference in the contact properties of multi-layer and single-layer 2D materials.  Figure 3-25. Contact resistance of 2D materials in monolayer and multilayer 88. Image used with permission from the publisher.  Contact through bulk TMD is known as a promising approach for contact to 2D materials 89. A Schottky junction between the channel and contact area is caused by the metal screening. The screening effect is weaker with bulk TMD contact, which results in a smaller Schottky barrier 89.  In one approach, bulk TMD could be used as an intermediate step to form contact between metal and monolayer. The advantage of this approach is the close work function matching between 49  bulk and monolayer TMD and less screening and renormalizing of the work function compared to the conventional metal approach, which causes worse contacts.  Figure 3-26. Energy levels of TMDCs and metals. Image adapted from 90. Image used with permission from the publisher.    50  3.8.1. Metal Contacts for MoS2 Lee et al.91 reports using Al with a work function of 4.1ev followed by deposition of Cr and Au with the following thicknesses Al\/Cr\/Au (40\/1\/50 nm). Furchi et al.92 reports using Pd\/Au for contacts both to MoS2 and WSe2 without specifying the thickness. Roy et al. 93 reports using Ni with 50nm thickness.  Figure 3-27. The work function of metals and their energy difference with the conduction band of MoS2. Image adapted from 90. Image used with permission from the publisher. Experiments by McDonnell et al. 94 demonstrate the importance of vacuum levels during the deposition of Ti on MoS2. According to their report, ultrahigh vacuum condition (~1x10-9 mbar) is required to achieve pure MoS2-Titanium contact. Depositions under high vacuum (~1x10-6 mbar) result in the formation of TiO2, according to this study.  English et al. have demonstrated that depositing Au on MoS2 in ultra-high vacuum (\u223c10\u20139 Torr) results in 3x smaller Rc compared to deposition in normal conditions 95. Their experiments reveal that Au contacts result in lower contact resistance compared to Ti, Ni and Sc.   Figure 3-28. Titanium deposition on MoS2 in high vacuum and ultra-high vacuum. Image adapted from 94. Reprinted (adapted) with permission from McDonnell, S., Smyth, C., Hinkle, C. L. & Wallace, R. M. 51  MoS2-Titanium Contact Interface Reactions. ACS Appl. Mater. Interfaces 8, 8289\u20138294 (2016. Copyright 2016 American Chemical Society. To avoid potential TiO2 formation on the MoS2-Ti interface during deposition, depositions of Ti are performed in an ultra-high vacuum setup. Au is deposited in the same chamber after the deposition of Ti. The deposition thickness of Ti and Au are 10nm and 80nm, respectively. Further information about the deposition system used in this study is available in the deposition systems section.   Figure 3-29. Comparison of contact resistance (Rc) for different metals at ultra-high vacuum (\u223c10\u20139 Torr) and vacuum (~10-6 Torr) on MoS2 95. Reprinted (adapted) with permission from English, C. D., Shine, G., Dorgan, V. E., Saraswat, K. C. & Pop, E. Improved contacts to MoS2 transistors by ultra-high vacuum metal deposition. Nano Lett. 16, 3824\u20133830 (2016). Copyright 2016 American Chemical Society. Oxygen plasma can remove resist residues and creates oxygen on the surface suitable for TiO2 formation, which prevents MoTi formation and creates a smaller tunnel barrier 96. Hence, oxygen plasma for 5s at 50W power is applied on the substrates prior to Ti\/Au 10nm\/80nm deposition. The work function of Ti (4.3eV 97) and Al (4.54eV 97) are both close to the reported range of MoS2 work function (4.5 to 5.2eV 98). This makes Ti and Al suitable candidates for contact to MoS2. A study of Ti and Al contacts to MoS2 in Field Effect Transistors (FETs) by Shimazu et al. 99 demonstrates significant differences in conductivity relative to the back gate voltage. 52  The Gs-Vg measurement shows that the devices with Al contact remain at the conductivity of above 20uS across -40v to 40v back gate voltage 99. For Ti\/Au contacts, the conductivity is almost zero for negative back gate voltages and raises to above 20uS at 40v.   Figure 3-30. The conductivity of Ti\/Au and Al contacts on MoS2 99. Image used with permission from the publisher.   The work by Shimazu et al. also demonstrates that deposition of Ti\/Au (15\/75nm) on Al (100nm) significantly changes the conductivity properties. They argue that the deposition of Al on Ti\/Au creates a resistive alloy.   Figure 3-31.Effect of Ti\/Au deposition on Al on Id-Vg curves. (Left) Al (100nm) before deposition of Ti\/Au. (Right) After deposition of Ti\/Au (15\/75nm) 99. Image used with permission from the publisher.  As described earlier, based on the experiments by Shimazu et al. 99, the electron doping effect of Al contacts is only evident when Al is not in contact with other metals with high work function.  53      Figure 3-32. Achieving Ohmic contact for High-quality MoS2 devices on hBN. Single-layer devices are contact resistance dominated (several M.ohm.um). Multilayer devices have much lower contact resistance. (Reference: Fig. 2.3.2, 2.4.1. from 100). Experiments on MoS2 devices on hBN demonstrate that Single layer devices are contact resistance dominated with contact resistance of several M.ohm.um. Three layer devices, however, show much lower contact resistance below one M.ohm.um. 100. Kaushik et al. demonstrate that introducing ultrathin TiO2 interfacial layer between MoS2 and various metals such as Ti, Au and Pd can reduce contact resistance by 24x times 101. 3.8.2. Pd Metal Contacts for WSe2 Pd as a high work function metal provides proper hole injection into WSe2, which makes it a good candidate as a metal contact 102. Lee et al. use Pd\/Au with respectively 20\/30nm thickness for metal contacts to MoS2-WSe2 p-n junction devices 91. Zhao et al. use e-beam evaporation of Pd\/Au with respectively 10\/30 nm thickness on WSe2 65. Roy et al. use 50nm Pd for contact to WSe2 93. Flory et al. use the evaporation of  Pd\/Au with 20\/40\u2009nm thickness for a MoSe2\/WSe2 junction 49. In this work, 10nm of Pd followed by 85nm Au deposition is used. More information about the deposition of Pd is provided in the deposition systems section.   54  3.8.3. Deposition system In this section, we briefly study the deposition systems used in this work.  Ultra-High vacuum hybrid evaporator  We use the electron beam deposition system of the AJA Ultra-High vacuum hybrid evaporator. This system has several features that make it preferable for metal deposition of Au and Ti for this project. It is capable of reaching high vacuum levels at a short amount of time. The small load-lock system cuts the vacuum time from several hours to less than an hour. The high distance of the sample holder to the evaporator provides uniform deposition. The limited metals - Au, Ti, Al, Al2O3 - available in the system prevent issues due to cross-contamination. Due to the complexity of the process of changing the available metals, another system is used for experiments that require metals that are not already available in this system. It is worth noting that an unpublished AFM study of metal deposited on 2D materials from collaborators of this work indicated evidence of low-quality deposition. Further study of surface roughness and potential patching could shed light on the deposition quality.  Figure 3-33. AJA Ultra-high vacuum hybrid evaporator. The sample transfers to the ultra-high-vacuum main chamber (on the left) using a load-lock. The e-beam sources with six crucibles are located at the 55  bottom of the chamber, and the four sputtering sources are located at the top. The red mount on the chamber has a handle that allows rotating the sample to face the required deposition source.  Image Source: https:\/\/www.nanofab.ubc.ca\/  E-beam and thermal evaporator embedded in glovebox  One of the issues with this e-beam evaporator is that the minimum amount of metal needed to run a deposition is higher than thermal evaporation. Hence, for experiments with metals such as Palladium (Pd), the minimum cost for beginning an experiment is relatively high. With thermal evaporation systems, it is possible to run deposition with smaller amounts of metals.  To test the thermal evaporation of the system, an Alumina coated tungsten evaporation basket is installed in the chamber. The basket is loaded with a new 2.5-gram Palladium pallet as well as small amounts of metal left from previous experiments. The chamber is then vacuumed. While increasing the voltage applied to the evaporation basket, the system shut down due to heating. The manufacturer suggested using a system with water cooling for the evaporation of metals that require high temperatures.  Figure 3-34. Image of the e-beam evaporator embedded in glovebox. The e-beam evaporator is located by the side of the heterostructure stacking setup.    56  Thermal evaporator  A thermal evaporator that is equipped with a rough pump and turbo pump is used. An external fan is used to cool down the turbo pump and prevent shutdown.  A thermal evaporator with easy access to boats and baskets allows rapid installation of new deposition metals. The fully manual operation of this old system gives full control over all the steps of the deposition process. The water-cooling capability is suitable for the evaporation of metals that require high temperatures.  To minimize contamination, the walls of the chamber were scrubbed and vacuumed. A laboratory-grade vacuum cleaner with HEPA filters and masks was required for safety reasons. After fixing the assembly of the chamber, a test vacuum using a Pirani gauge was conducted. Insufficient vacuum levels indicated a potential leak in the chamber. Cleaning the gaskets with isopropanol alcohol improved the vacuum level while still not reaching the previously recorded levels. To identify leak spots in the chamber, an Agilent VS series leak detector was used. Detected leaky areas were cleansed with isopropanol alcohol and tightened. The chamber reached 3.4x10-7 Torr in less than 24 hours of vacuum. Since this system is not capable of reaching high currents, regular boats that drain hundreds of amps of current are not suggested. Alumina-coated tungsten evaporation coil baskets are ordered instead. These coil baskets can reach temperatures high enough for evaporating palladium in a vacuum with relatively low power. The disadvantage of these baskets is their fragility. They can break and may need to be changed after a few depositions. It is suggested to order a few of these baskets to be prepared for changing them in case of damage.  57  To evaporate metals that require high temperatures, a cooling system may be required. The water-cooling system of the evaporator is tested and fixed. It uses a water pump that runs water through the stands holding the basket.               Figure 3-35. Images of the thermal evaporator. (Left) Inside of the evaporator where the three Alumina coated tungsten evaporation coil baskets are installed. (Middle) Manually reading the current using a clamp current reader. (Right) Agilent VS series Leak detector connected to the vacuum chamber. The Pirani gauge (the orange box on the left) is simultaneously connected to the chamber to read the vacuum level.  Sputtering system  A physical vapor deposition (PVD) sputtering system (Nexdep Base Angstrom Engineering) is used for some of the depositions, such as the deposition of Aluminium on the resist. The system is capable of depositing metals, dielectrics, magnetic materials, and semiconductor thin films using three independent circular magnetron sputter sources. It has a 1.2kW DC sputter power supply and a 1kW pulsed DC power supply. For the deposition of 7nm of Al on resist on a SiO2\/Si substrate, the pulsed DC mode at 20% power is used for 20 seconds.  58   Figure 3-36. Angstrom PVD sputtering system used for the deposition of metals. The sample holder is capable of handling substrates up to 150mm in diameter and flexible substrates with a maximum width of 120mm.           59  4. PN diode-based van der Waals heterostructures for photovoltaics 4.1. Introduction The combination of a p-type and n-type semiconductor material results in the formation of a p-n junction with rectifying diode behavior 103. The internal electric field of the p-n junction can separate electron-hole pairs, thus providing a suitable condition for a photovoltaic device. In this chapter, we review the van der Waals p-n junctions for photovoltaic applications.  We successfully observe diode effect in a vertical van der Waals heterostructure. Our custom-made low-vacuum measurement setup shows an improved electrical response. We conduct experiments on ultra-high vacuum annealing, which results in up to ~4x increase in current. To evaluate the photovoltaic effect, we install a white LED in a low-vacuum setup and observe photovoltaic behavior at room temperature with different light intensities. To understand the underlying working mechanism and explanation of photovoltaic behavior, we perform low-temperature measurements at 77K. Compared with other monolayer works, we produce devices ~30x larger than Furchi et al. and ~3x larger than the devices reported by Lee et al. with higher current density at similar incident light. 4.2. Nanofabrication 4.2.1. Fabrication of metal contacts The first series of devices were made with 5\/75nm Ti\/Au because of good adhesion of Ti and the conductivity of gold. In the next series of devices, the Si\/SiO2 (285nm) substrate was heated to 300o C prior to exfoliation to clean the substrate from any contamination. The flakes were exfoliated from tape to PDMS to silicon. The transfer was done at 220o C. In this device, the 60  contacts to 2D materials were designed wider to provide lower contact resistance. Some devices stopped working after ramping the back gate voltage to higher than -15v, indicating the probability of damage to the surface of SiO2.        Figure 4-1. Microscopic images of two van der Waals MoS2-WSe2 samples. (Left) an early generation of devices with smaller contacts. (Right) Devices with larger contacts. 4.3. Diode behavior results and discussion  4.3.1. Diode behavior In this part, we review the diode behavior of the p-n junction. The combination of a p-type and n-type semiconductor material results in the formation of a p-n junction with rectifying diode behavior. In classical p-n junction diodes, a built-in potential is built across the junction due to the movement of carriers from one side of the junction to the other. This depletion region, typically in the order of a few micrometers, influences diffusion and drift processes resulting in carrier transport across the junction. In 2D materials heterostructures, the p-n junction is only a few nanometers. 104, 13 61     Figure 4-2. Diode behavior of the device for Vds -4v to 4v and Vbg -60v to 60v. (Left) Ids-Vds graph for Vbg -60v to 60v (Right) Vbg-Vds graph.  4.3.2. Vacuum effect on diode electrical response  To test the effect of vacuum pump during the measurement, diode performance is compared while measuring the device in ambient conditions and in low-vacuum using a rough pump.   Figure 4-3. (Left) Measurement in atmosphere conditions before annealing. (Right) Measurement in rough vacuum before annealing.   62  4.3.3. High vacuum annealing effect on device performance During this study, it was observed that annealing could improve the device performance significantly. An earlier study has demonstrated the effect of high vacuum annealing through analysis of their AFM topography 68. The electrical measurement of the device in this study shows ~4x increase in forward bias current after annealing.  To achieve the optimum results from the annealing, a setup with the capability of reaching high vacuum was used. An evaporator installed in a high-purity glovebox was used as the vacuum chamber for the annealing process. The evaporator is equipped with a heater located under the sample holder. A cryogenic pump is used to reach vacuum levels as low as 6x10-8 torr. The device is annealed using the following process. The chip is mounted on the sample holder of the evaporator facing down. The heater was set to 120 oC and the vacuum started simultaneously. In 45 minutes, the vacuum reached 2.6x10-6 torr and the temperature was raised to 108 oC. The device was annealed for 24 hours at 120 oC, reaching 6x10-8 torr vacuum by the end of the process. It was cooled down for 3 hours in vacuum reaching 70 oC before purging the chamber. The device was finally removed from the chamber and transferred to the measurement setup, followed immediately by running a low vacuum pump. The diode performance of the device measured in similar conditions prior to annealing is used to evaluate the effect of annealing. Up to 60 volts back gate is applied while sweeping the drain-source voltage from -4v to +4v in both measurements before and after the annealing. The forward bias current at 60v back gate voltage and 4v Vds has raised from ~0.85nA to ~3.5nA after annealing. This indicates an over 4x increase in current. Furthermore, the unusual response of the device at Vds>3v at Vbg 40v and 30v has been corrected after annealing. This correction 63  could be the result of the removal of unwanted contaminants during the annealing process at vacuum.        Figure 4-4. Top Left: Rough vacuum before annealing. Top right: After annealing (Vbg: 60v).     64  4.4. Photovoltaic response The electrical response of the diode to various intensities of light is measured and reported in this section. All the measurements are performed under a white LED. The electrical response is measured under the white LED light with 156uW, 278uW, 460uW and 837uW input powers. Experiments in this section show the I-V measurement of the diode in the -4V to 4V drain-source voltage range. These measurement results can be described in several sections. The section with large reveres bias ranging from -4v to -1v shows significant impact of light with reverse current raising continuously. Later we will discuss that in the forward bias region, the situation is different. This increase in the reverse current corresponds to electron and hole pairs generated by the absorption of light and separated with the drift voltage.  In the section from -1v to 0v the dark current quenches at around -1v reverse voltage. This behaviour demonstrates that the separation of electrons and holes highly depends on the drift voltage. When the reverse bias is small, there is not enough internal drift voltage to separate electrons and holes. To minimize this current drop, a stronger intrinsic internal field is needed. This internal field can separate generate electrons and holes by separating excitons without strong dependence on the external bias. The internal field can be increased by selecting 2D materials with a larger voltage difference between the valance band of the p-type material and the conduction band of the n-type material. The field required to separate electrons and holes can also be enhanced by using metal contacts with uneven work functions.     65   Figure 4-5. PV response at 300K in dark and LED power of 270, 460, 481, 938 W\/m2. (Left) Vds from -4v to 4v. (Right) Vds from -0.3v to 0.65v.  The section from 0v to 400mV bias voltage is where the device generates current in the reverse direction of the bias voltage. In this area the device works as a photovoltaic device and generates power. The device generates maximum 4.5 pW power at 681 W\/m2 input light. The active part of the device, the heterostructure area, is 31.1 um2. This will correspond to a total of 21,179 pW input light power for the area of the device. Using a simple power efficiency equation results in 0.0212% power conversion efficiency.   Figure 4-6. Output power in pW for 270, 460, 681 and 937 W\/m2 intensities of white input light for voltages ranging from 0 to 400mV.   66  The report on the efficiency of 2D material heterostructure photovoltaic devices has multiple criteria to consider. To bring light to the efficiency reports, we provide the following explanation. Firstly, the reported area is typically the overlap region of the p-type and n-type regions visible under the microscope. However, cracks and gaps in the 2D materials caused by exfoliation, stacking process and nanofabrication, may result in a smaller effective area. For example, Mueller group report of 0.2% is driven based on a 1.1um2 effective area of a device with roughly 9um2 overlap area under microscope. Hence, the reliance of efficiency on the effective area makes the reports unreliable for devices where cracks and gaps are not extensively studied, typically resulting in efficiencies lower than the actual value. Secondly, the effective area of the device plays a multidimensional role in the efficiency. The larger device area provides more exposure to light and results in a higher number of absorbed photons. This should contribute positively to higher efficiency by generating more carriers. However, the larger device area means more loss caused by lateral and vertical recombination. These two mechanisms simultaneously result in more carriers due to higher absorption and more loss due to recombination. Depending on the device design, material quality and device area, and distance from contacts, this could result in higher or lower efficiency. Existing reports dominantly show a lower efficiency for larger devices, indicating the superiority of the recombination over higher absorption for larger devices. The high dark current could correspond to the sensitivity of the device to infrared emission from the heated LED light. In order to confirm this effect, the dark current measurement immediately after turning off the LED and with a few minutes wait after turning off the LED is compared. The dark current at 100mV seems to go from 25pA to 10pA with a few minutes of delay after the 67  LED is turned off, which confirms that waiting for the LED to cool down brings down the dark current.  With this dark current with 10pA at 100mV and 35pA light current at the same voltage, the device has an effective peak power of  2.5pW and 460mW light power the device generates.  From 400mV to 1900mV, the effect of forward bias takes over with higher light intensity measurements resulting in a higher drain-source current. This behaviour changes at Vds above 1900mV. In this region, dark current increases and reaches values higher than high intensity light measurements.        68  4.4.1. Photovoltaic results discussion In this section we discuss the photovoltaic results for room temperature, low temperature (77K), and for different intensities of incident light. We particularly study the response of the device in the photovoltaic region where we observe a positive output power from the device. We discuss how these results reveal the underlying working mechanism of the photovoltaic effect in the multilayer van der Waals heterostructures. We finally compare our results with single layer heterostructure reports in the literature and we show how our device area and short circuit current compares. The following graph shows the output electrical power for white light illumination intensities of 270, 460, 681 and 937 W\/m2. It shows the output electrical power for forward voltages from -300 to 700mV. Averaging has been used to minimize the noise in the figures. The results from the 450 to 700mV range show that the current and the value of output power increase by increasing the input light intensity. This is contrary to what is observed in conventional semiconducting p-n diodes, where the illuminated current is smaller than the dark current and goes lower with an increase in light intensity.  This could be explained by the role of free carriers generated by light. Typically, the highly doped semiconductor provides the majority of free carriers required for diffusion current in the forward region. In this case, the multi-layer device screens the electrostatic doping effect provided by the back gate, and the doping effect is small for higher layers. The smaller doping effect limits the number of free carriers needed for diffusion current in the forward region. When light is projected to the device, the number of free carriers generated by light is comparable to 69  the existing free carriers in the junction, which results in a higher diffusion current. A similar effect is observed in other works for voltages higher than Voc 32.  Figure 4-7. Output power in pW for 270, 460, 681 and 937 W\/m2 intensities of white input light for different range of voltages. (left) ranging from -300 to 700mV.  Figure 4-8. I-V characteristics of a conventional p-n diode under dark and illumination. Unlike this study, typically, the illuminated current is smaller than the dark current for voltages higher than Voc. The figure below shows the output electrical power for white light illumination intensities of 270, 460, 681 and 937 W\/m2 for the area that the device is generating power or showing the photovoltaic response from 0 to over 350mV forward voltage. Averaging has been used to minimize the noise in the figures. 70  These measurement results show that the input light is effective in generating photovoltaic output power. The results also demonstrated that increasing the input light intensity from 460W\/m2 to 681W\/m2 increases the output power almost with the same increase ratio (~50%). The measurement for 937W\/m2 input light, however, is inconsistent with this trend and the trend observed for V>450mV provided in previous figures.  The drop in output power at higher input light intensities could correspond to the competing current generation mechanism. The increase in input light, on one hand, increases the photovoltaic current and, on the other hand, increases the free carriers contributing to diffusive forward current. This results in inconsistent output power with increased input power in the photovoltaic region. This behavior is not common in p-n diode-based photovoltaic devices. In a conventional semiconducting p-n diode photovoltaic device, the free carriers of the doped semiconducting material are in a majority providing enough carriers for the diffusion current. Hence, the photogenerated current doesn\u2019t have a significant influence on the diffusive current. As discussed earlier, in this case, the multi-layer device screens some of the electrostatic doping effects from the back gate, resulting in a smaller doping effect. As a result, the carriers generated by light seem to be on a competing scale for higher intensities of light. This behavior is not observed in single-layer p-n diode heterostructure devices 91,105. We believe this difference is due to stronger electrostatic doping and less screening effect for single-layer devices used in other reports, which creates a higher carrier density. Our low temperature measurements help to elucidate this explanation. Further information about the low temperature measurements is available in the next section. The low temperature measurements are performed at 77k. At low temperatures the carrier density in semiconductors drops. This drop in carrier density results in lack of diode effect as shown in measurement 71  results. The low temperature measurements help to evaluate the contribution of current from carriers generated by incident light while minimizing the effect of current due to free carriers in the semiconductor activate by thermal effect. The measurement at 77K for 270, 460, 681 and 937 W\/m2 show consistent increase in the current flow with increase in intensity of light. This is contrary to the room temperature measurement. This can validate the theory of competing current generation mechanism at room temperature. In room temperature the incident light increases the photovoltaic current and the free carriers contributing to diffusive forward current. The competition between these currents results in a drop in output power when incident light is increased from 681W\/m2 to 937W\/m2. At 77K, there is minimal free carriers in the semiconductor due to lack of thermal activation. Hence, the carriers generated by incident light are not in a competing scale with free carriers. This results in a consistent increase in the current with increase in incident light.  Figure 4-9. Output power in pW for 270, 460, 681 and 937 W\/m2 intensities of white input light for voltages ranging from 0 to 400mV. To better understand the output power, we compare our results with existing reports by other groups. Furchi et al. achieve ~1.3pW maximum power for a white light illumination from a halogen lamp with 670W\/m2 power. We achieve ~3.7pW max power for a white LED light with 72  681W\/m2 power. This is close to 3 times (~2.85x) more power than the results obtained by Furchi et al. for similar input power. One of the critical requirements for building a photovoltaic device is the scalability of the area. For building devices for this experiment, we seek large flakes to build larger overlap areas. The device presented in this report has a 31.1um2 area which is ~30x times larger compared to the 1.1um2 device area by Furchi et al. and ~3x larger than the device reported by Lee et al. (~10um2). It is worth mentioning that because of cracks and gaps in the flakes, it is possible that the effective area of the device is smaller than the overlap observable with optical imaging.  To have a better understanding of the correlation between illumination power and short circuit current, open circuit voltage, and the fill factor, we present the following graphs. The short circuit current increases by increasing the input light. The open circuit voltage, which is a function of device energy levels, remains unchanged, as expected. The fill factor shows a slight decrease with illumination power which could be because of the competing current generated by an increase in carrier density.  Figure 4-10. Short circuit current, open circuit voltage, and fill factor for illumination under white light at room temperature. To have a better understanding of the device performance, we compare our results with the data reported by other groups using a MoS2\/WSe2 heterostructure. Ref. 1 (Lee et al.) data is extracted 73  from Figure 2.a of their report 19. The data corresponds to a monolayer-monolayer MoS2\/WSe2 heterostructure with Pd and Al contacts. It is measured under -10v back gate voltage under white-light illumination with unspecified power. The -10v back gate voltage data is used because of its diode like behaviour, as discussed in ref. 3. Ref. 2 (Furchi et al.) data is extracted from Figure 3. a of their report 17. They use a monolayer-monolayer MoS2-WSe2 heterostructure with Pd contacts. The data shows the response with -50v back gate voltage.   Figure 4-11. Ref 1 (black dashed line): Data extracted from Lee et al. 19. Ref 2 (grey dashed line): Data extracted from Furchi et al. 92. 74   Figure 4-12. Microscopic images of the heterostructures. (left) The heterostructure used in this study. (Middle) Heterostructure used by Lee et al. scale bar is 3um. (Right) Heterostructure used by Mueller group.    75  4.4.2. Temperature-dependent electrical and photovoltaic response In this section, we study the effect of temperature on the electrical and photovoltaic response of the device. This information can help with understanding the role of thermal activation in the mechanism behind diode effect and the photovoltaic response. The measurements were done at 77K, cooled down using liquid nitrogen, and at 300K (room temperature). The response to white light LED source with four intensities as well as dark current was measured. Two sets of measurements were performed, one to observe the diode behaviour and the other to carefully investigate the photovoltaic response. The former at higher Vds ranges from -6v to 6v for 77K and -4v to 4v at 300K and the latter from -250mV to 650mV.  The results from the 77K measurement prior to annealing demonstrate that the diode behaviour vanishes across all back gate voltages. The diode response changes significantly when temperature reaches 77K. The reverse current is close to 0 at RT, while at 77K, we observe a reserve current in a similar scale as the forward current.  This phenomenon could correspond to suppression of the diffusion current and domination of drift current due to the low temperature. The lack of enough carriers at low temperature could be the cause of suppressed diffusion current. The carrier density of semiconductors drops by decrease in temperature based on these equations:      Where KB is the Boltzmann constant, T is the absolute temperature of intrinsic semiconductor, Nc is the effective density of states in the conduction band, and Nv is the effective density of 76  states in the valence band. Lower carrier concentration results in fewer available carriers for diffusion current.    Figure 4-13. (Top left) 77K measurement with Vds from -5v to 5v and (Top right) room temperature measurement with Vds from -4v to 4v. (Bottom left) PV response with white illumination for Vds from -0.3v to 0.65v at 77K (Bottom right) similar measurement for room temperature with removal of back gate current. The diode behaviour of the device is measured by applying and sweeping a voltage between the WSe2 and MoS2 while changing the back gate voltage for each measurement. The Vds is swept from -4v to 4v for each measurement and the back gate voltage is applied from 0v to 60v in 7 steps with 10v increase in each step. 77     Figure 4-14. Diode behavior measured before annealing for Vbg:0_60v for Vds: -4_4v at (left) room temperature (right) 77K.         78  4.4.3. Response time The response of the device to light is measured by turning on the light and measuring the current with a fixed applied voltage. It takes over 50 seconds to reach the saturation current. Similar behavior is reported by Zhang et al. in a WSe2-based phototransistor and is suggested to be caused by the Pd contacts 106 due to excitation through defect or charge impurities 107. The role of adsorbates in ambient air and temperature-dependent persistent photoconductance in slower response time of MoS2 phototransistors is demonstrated by comparison of measurements in vacuum, air and at room and low temperatures (66K and 4.2K) by Zhang et al. 108. The prolonged response time demonstrates that the choice of Pd as a metal contact is not suitable for time-sensitive sensing applications. Despite lower photo gain, using Ti as a metal contact might be more appropriate for sensing applications of the MoS2-WSe2 heterostructures due to faster response time.  Figure 4-15. (left) Photo response of MoS2-WSe2 heterostructure with Pd contacts to white light. (Right) Photo response of WSe2 phototransistor with Ti and Pd contacts demonstrating higher photo gain for Pd and faster response time for Pd 109. Reprinted (adapted) with permission from Liu, W. et al. Role of metal contacts in designing high-performance monolayer n-type WSe2 field effect transistors. Nano Lett. 13, 1983\u20131990 (2013). Copyright 2014 American Chemical Society.    79   5. Novel van der Waals photovoltaic devices In the previous section, we demonstrated a photovoltaic device based on van der Waals heterostructures. In this section, we explore novel photovoltaic devices that can benefit from van der Waals heterostructures. This goal is particularly focused on multi-stack photovoltaics and flexible photovoltaics. Both have not been explored with junction-based van der Waals structures at the time of conducting this research to our knowledge.   80  5.1. Multi-stack van der Waals photovoltaic devices 5.1.1. Introduction Since the invention of the first solar cell based on a p-n junction in 1954 by Chapin, Fuller and Pearson the search for methods to improve their efficiency started 110. From the early days of this discovery, the pursuit of multi-junction solar cells was reflected in literature 111. Multiple p-n junctions with different energy band gaps were considered as one of the future solutions to achieve higher efficiencies. However, the impossibility of having electrical current circulating across a p-n junction biased in reverse was identified as the challenge for building such devices 111. Tunnel diodes with highly doped p++\u2013n++ junctions 112 helped resolve this challenge. Multijunction solar cells 113, which benefit from tunnel diodes, and wafer-bonded solar cells 114 are some existing methods to attach different solar cells with p-n junctions. Based on the estimations by Mart\u0131\u00b4 and Luque115 maximum efficiency of 54.7% is expected for these three stack heterostructures. Despite theoretical proposals for such structures 116, 117 limited or no presence of experimental demonstration of p\/n\/p or n\/p\/n structure solar cells in literature during this study, at least for 2D materials, came to our surprise. 5.1.2. Theory and proposal Thanks to their van der Waals stacking, 2D materials allow the making of heterostructures that were hardly possible with conventional crystalline materials using methods such as epitaxy. Exploiting this unique benefit of 2D materials, we propose a 2D material based bipolar junction solar cell, a new approach to building multi-junction solar cells.   Our structure is composed of an n-type (bottom), p-type (middle), n-type (top) setup, which generates two diodes running in reverse. The n-type (bottom) layer is composed of a thick 81  multilayer MoS2, which allows maximum absorption of light, and the higher thickness results in a lower bandgap material that absorbs light with lower energies more effectively. The p-type (middle) layer is a p-doped WSe2. The p-doped WSe2 eliminates the need for electrostatic doping and allows for building a three-stack setup. The top n-type is a few layers MoS2, which provides high transparency and the higher band gaps absorb photons with higher efficiency. The circuit schematic below shows an electrical model for the device. The diodes in reverse correspond to the reverse-facing diodes created by the n-p-n structure. The current sources correspond to the photogenerated current in the two junctions. The Rc corresponds to the contact resistance for each layer.  Figure 5-1. Proposed model for the electrical circuit of the bipolar junction solar cell. 5.1.3. Experiment In this section, the design and fabrication of the devices built to demonstrate and test the three-terminal heterojunction bipolar transistor solar cells are explained. The device is composed of three-layer stacked semiconducting 2D materials. MoS2 is used as the n-type material. Since electrostatic doping of WSe2 is hardly possible in this structure, doped Rc RcRc82  WSe2 is used as the p-type material. An NPN structure is used by stacking MoS2-WSe2-MoS2 using the stacking setup built as a part of this project.   Figure 5-2. Microscopic image of a multi-stack heterostructure.  Figure 5-3. Side view of MoS2-WSe2-MoS2 multi-stack heterostructure. To provide optimum contact with the layered materials, separate contact metals are used for the MoS2 and WSe2. This required multiple sets of lithography. Pd\/Au contact with 10\/85nm thickness was used for contact to WSe2 layer and Ti\/Au with contact was used for MoS2 layers.  83         Figure 5-4. Optical Microscopy images of the van der Waals heterostructure. (Top Left) Image of the device after exfoliation. (Top Right) Design of the Ti\/Au and Pd\/Au contacts to MoS2 and WSe2. (Bottom Left) Image of the device after depositions. (Bottom Right) Magnified image of the device.   Once the exfoliation and stacking of the NPN structure is completed, the structure is washed in chloroform for 45 minutes to remove any remaining residues from the polymer-based stacking stamp. Two series of contacts are then designed based on the microscopic picture of the stacked heterostructure. The sample was prepared for lithography by spin-coating resist and lithography was performed using e-beam lithography. First, the Pd\/Au (10\/85nm) contacts are deposited on WSe2 using a metal evaporator at 3x10-7 vacuum pressure. Markers are also deposited close to the heterostructure to make the next lithography step alignment more accurate. Lift-off is then performed to remove excess deposited metal. The device is then annealed in the glovebox for 20 minutes at 110o C. To remove remaining resist residues, the device is treated with remote oxygen plasma for 10sec at fluctuating mode at 2sccm. The device is then spin coated with resist to 84  prepare for the second round of e-beam lithography. Once the device is developed, it is moved to the AJA evaporator to deposit Ti\/Au at 10\/85nm thickness. Finally, the device is bonded and annealed at glovebox at 1100 C for 8 hours before moving to the measurement setup.    85  5.2. Flexible van der Waals photovoltaic devices 5.2.1. Introduction One of the advantages of using 2D materials is their flexibility. This is particularly important for solar cell applications as it has the potential for making flexible solar cells. Another application of flexible 2D materials devices is a platform for studying strain-induced properties such as band gap engineering 118.  Stacked 2D materials heterostructures provide a more interesting platform for studying strain-induced effects. Modeling of MoS2\/WS2, MoSe2\/WSe2, Graphene\/Graphene and MoS2\/MoS2 using a 2D non-linear shear-lag mode are used for studying transferred strain 119. Raman spectroscopy of MoS2\/WS2 heterostructures under strain is studied and the resulting shift in the Raman spectroscopy are presented by Susarla et al. 120. One of the recent areas of interest in 2D materials is the study of Moire patterns in twisted bi-layer 121 or multi-layer devices 122. The mechanism for building such devices usually involves cutting a larger flake and accurately aligning it with the desired angle. Studying various angles with such procedures requires building at least one device for each angle. However, it could be possible to study such patterns via stretch 123. Building 2D material heterostructures on flexible substrate allow such studies.   5.2.1. Fabrication of flexible devices Two flexible substrates were selected for this experiment. Initially, 125um thick heat-stabilized Melinex was used. It is important to use substrates with minimal surface roughness. Hence, an ultra-smooth with ~2nm excepted roughness PEN substrate with 50um thickness named TEONEX\u00ae Q65FA was used. The substrate is ultra-heat stabilized with heat shrinkage 0.3% at 86  200o C and 0% at 150o C. The substrate was cut into pieces suitable for lithography using a laser cutter. SEM conductive double-sided tapes were used to glue the flexible substrate to a metallic base for easier handling.   Figure 5-5. (Left) PEN substrates are cut into smaller pieces using a laser cutter. (Right) Image of two flexible substrates on the right side and the flexible substrate mounted on a custom-cut aluminium holder on the left. A conductive SEM glue stick is used to attach the flexible substrate to the aluminium holder. One of the issues with fabrication on non-conductive flexible substrates is the electron beam lithography process. The non-conductive substrate does not allow the free flow of electrons used in electron beam lithography. So, a conductive resist or coating is required to be able to perform electron beam lithography on these substrates. One method suggests depositing a metal on the flexible substrate. This can be achieved by sputtering around 10nm of Aluminium. 7nm Al was deposited using an Angstrom PVD sputtering in pulsed DC mode at 20% power for 20 seconds. The other method suggests using conductive resists, such as Aquasave. The following recipe was used for the deposition of Aquasave. Spin coating of Aquasave at 2500rpm with 50rpm\/sec acceleration for 50 seconds followed by 5 minutes of baking in an oven at 150\u00b0C. After the lithography, the sample is cleaned with DI water at room temperature for 1 minute. Finally, the sample is developed using an IPA:DI mix at 15\u00b0C for 30 seconds, followed by 30 seconds dip in DI water. 87  To transfer the flakes to the flexible substrate, the same stacking setup as described in the previous sections is used. For the transfer of WSe2 the transfer started at 120\u00b0C and the temperature was raised to 150\u00b0C followed by cooling down to 110\u00b0C. For MoS2 transfer, we started with 130\u00b0C on the transfer setup substrate and raised the temperature to 170\u00b0C; finally, it was cooled down to 120\u00b0C. The flexible substrate needs optimized configurations for bonding. The bonding was performed using the 300 pressure setting.    Figure 5-6. (Middle) Flexible substrate attached to a chip holder using glue. In spite of the flexibility of the substrate, the wire bonding between the substrate and chip holder shows a proper connection. (Right) Image of two flexible substrates on the right side and the flexible substrate mounted on a custom-cut aluminium holder on the left. A conductive SEM glue stick is used to attach the flexible substrate to the aluminium holder.    88  6. Outlook  6.1. Van der Waals heterostructure solar cell The combination of an n-type and p-type 2D material creates a p-n junction19. Some research groups have demonstrated a 2D material based p-n diode with potential for solar cell applications 19,32. While these works focus on monolayer materials, multilayer heterostructures are expected to have higher efficiencies mainly due to their higher light absorption 32,19. In addition, a more in-depth understanding of the impact of thickness on the device characteristics of atomically thin p-n diodes is critical for solar cells and other optoelectronic applications. Covering the p-n diode with a multilayer graphene from top, acting as a transparent electrode, can enhance the collection generated carrier. Multilayer graphene has better electrical conductance than single layer and approximately absorbs less than 2% of light per layer. 3-5 layers of graphene should achieve a reasonable conductance to transparency ratio. Furthermore using multilayer graphene has shown better solar cell performance than monolayer and bilayer graphene 43.  WSe2 is intrinsically weakly p-doped and can easily get n-doped. Because of the weak doping of WSe2, the MoS2-WSe2 heterostructure may not intrinsically be in the p-n regime. One approach to confirm that the heterostructure is in the p-n regime is by applying a back gate voltage to the device and performing electrostatic doping 19, 26.  By sweeping the back gate voltage and inducing doping effect to WSe2 and MoS2, the heterostructure goes into p-p, n-n, and p-n regimes. This approach has several problems. First of all, due to screening effect of 2DMs, the electrostatic doping effect is only functional for the very few bottom layers. In structures with thicker materials, electrostatic doping effect is totally screened. Secondly, this design requires applying back gate voltage, which is not suitable for several optoelectronic and photovoltaic applications.  89  Here, we propose a different approach. By using a high work function metal as the bottom contact, it is possible to induce doping effect on the bottom TMDCs. WSe2 has a valence band edge energy of 5.5ev (\u03c7WSe2 + Eg \u2248 5.5 eV) 124. Platinum with a work function of 6ev is 0.5ev below the valence band of WSe2. This energy difference creates an Ohmic contact and has a doping effect by shifting the energy diagram of WSe2 at the interface 125.  This metal layer also enhances the lateral transport of carriers.     Figure 6-1. Graphical image of the proposed device with Pt bottom contact, multi-layer TMDCs and Graphene top contact (a) Cross-sectional image of the device (b) 3D image of the device (c) band diagram of the interface.  The fabrication process of the device is similar to the devices presented earlier in this work. The additional platinum layer could be deposited with the Lesker E-beam evaporator. Photoluminescence measurement with Horiba HR-800 could provide information about the quality of the heterostructure interface. Afterward, the devices need to be bonded to a chip 90  carrier and electrically measured to confirm the diode behavior and extract the diode factors. To understand the photovoltaic response of the device, they need to be measured with the 100w Xenon lamp solar simulator equipped with a monochromator. These measurements will reveal PCE, short circuit current, open circuit voltage and EQE properties of this device for photovoltaic applications.  6.2. Thickness dependent study One of the main questions in photovoltaics in 2DMs is the rule of thickness. Due to unique properties of single-layer 2MDs, several types of research have been done on their optoelectronic properties 26, 32. However, for better absorption of light, using thicker materials is necessary. Photovoltaic measurements on multilayer TMDCs demonstrate higher power conversion efficiency (PCE) than single-layer devices, as shown in Table 2. But, increasing the thickness does not always increase the efficiency. As shown in the table, there is no clear correlation between the thickness of the materials and their efficiency. Also, these reports are performed with different device designs, so it is not possible to compare them. In order to address this question and increase our understanding of the effect of thickness on optoelectronic properties and PV efficiency, we perform thickness dependent study on these structures.   Heterostructures consisting of MoS2-WSe2 with 5nm-5nm (10nm total) & 20nm-20nm (40nm total) & 40-40nm (80nm total) could be made with the similar device design presented in the previous part. The thickness of devices is chosen with a reasonable difference so the effect of the thickness will be dominant. To minimize dependence of the performance on fabrication conditions, devices could be made on the same chip. This method will also decrease device 91  fabrication time. Similar to the previous sections, electrical and optoelectrical measurements could be performed to understand the diode behavior and photovoltaic response of the device. PCE, short circuit current, open circuit voltage and EQE properties of devices with different thicknesses provide more information about the devices. This data reveals more information about the rule of thickness in the photovoltaic response of 2DMs.         6.3. Van der Waals heterostructure tandem solar cell High-efficiency solar cells can be achieved by using semiconducting materials with different bandgaps in the form of a multi-junction or tandem solar cell 126. A similar approach can be used to make an atomically thin tandem solar cell. Van der Waal heterostructures have a great advantage for the fabrication of a multi-cell structure. Unlike regular 3D crystalline semiconductor heterostructures that require a complicated growth process limited by lattice matching of layers, van der Waal heterostructures simply stack together with weak Van der Waals forces 12. The complexity of material growth in conventional covalently bonded crystalline heterostructure limits the number of stacks to 4. In contrast, in van der Waals heterostructures, larger stacks are achievable due to their easier stacking mechanism. The single-cell device presented in the previous chapter paves the way for multi-cell structures. Currently, the main challenge to fabricating this structure is the difficulty of stacking more than four layers. Once this heterostructure is fabricated, by depositing metal contacts, it is possible to measure the electrical response of the device to the incident light. Characterizing the EQE at different frequencies, current-voltage and light-power conversion of the device can reveal the potentials of this structure as a future solar cell.  92       Figure 6-2. Van der Waals heterostructure tandem solar cell. 93  Bibliography 1. Fundamentals Of Solar Cells: Photovoltaic Solar Energy Conversion - Alan Fahrenbruch, Richard Bube. 2. Solar Energy International. Photovoltaics\u2009: design and installation manual\u2009: renewable energy education for a sustainable future. (New Society Publishers, 2004). 3. Solar Photovoltaic (PV) (CPV, Hybrid and Other systems) Market for Residential, Non-residential and Utility Applications: Global Industry Perspective, Comprehensive Analysis, Size, Share, Growth, Segment, Trends and Forecast, 2014 \u2013 2020. (2016). doi:MRS - 52702 4. W??hrle, D. & Meissner, D. Organic Solar Cells. Adv. Mater. 3, 129\u2013138 (1991). 5. Gr\u00e4tzel, M. Dye-sensitized solar cells. J. Photochem. Photobiol. C Photochem. Rev. 4, 145\u2013153 (2003). 6. Novoselov, K. S. Electric Field Effect in Atomically Thin Carbon Films. Science 306, 666\u2013669 (2004). 7. Nobelprize.org. The 2010 Nobel Prize in Physics - Press Release. Nobel Media A B 2014 (2014). Available at: http:\/\/www.nobelprize.org\/nobel_prizes\/physics\/laureates\/2010\/press.html.  8. European Union. Graphene flagship. (2013). doi:http:\/\/cordis.europa.eu\/fp7\/ict\/programme\/fet\/flagship\/ 9. Geim, A. K. & Novoselov, K. S. The rise of graphene. Nat. Mater. 6, 183\u2013191 (2007). 10. Zhu, Y. et al. Graphene and graphene oxide: Synthesis, properties, and applications. Adv. Mater. 22, 3906\u20133924 (2010). 11. Xu, M., Liang, T., Shi, M. & Chen, H. Graphene-Like Two-Dimensional Materials. Chem. Rev. 113, (2013). 12. Geim,  a K. & Grigorieva, I. V. Van der Waals heterostructures. Nature 499, 419\u201325 (2013). 13. Bernardi, M., Palummo, M. & Grossman, J. C. Extraordinary sunlight absorption and one nanometer thick photovoltaics using two-dimensional monolayer materials. Nano Lett. 13, 3664\u20133670 (2013). 14. Britnell, L. et al. Strong Light-Matter Interactions Thin Films. Science (80-. ). 340, 1311\u20131314 (2013). 15. Tsai, M. L. et al. Monolayer MoS2 heterojunction solar cells. ACS Nano 8, 8317\u20138322 (2014). 16. Buscema, M. et al. Photovoltaic effect in few-layer black phosphorus PN junctions defined by local electrostatic gating. Nat. Commun. 5, 1\u20136 (2014). 17. Furchi, M. M., Pospischil, A., Libisch, F., Burgdorfer, J. & Mueller, T. Photovoltaic effect 94  in an electrically tunable van der Waals heterojunction. Nano Lett 14, 4785\u20134791 (2014). 18. Georgiou, T. et al. Vertical field-effect transistor based on graphene\u2013WS2 heterostructures for flexible and transparent electronics. Nat. Nanotechnol. 8, 100\u2013103 (2012). 19. Lee, C.-H. H. et al. Atomically thin p\u2013n junctions with van der Waals heterointerfaces. Nat. Nanotechnol. 9, 676\u2013681 (2014). 20. Zhang, W. et al. Ultrahigh-Gain Photodetectors Based on Atomically Thin Graphene-MoS2 Heterostructures. Sci. Rep. 4, 3826 (2015). 21. Withers, F. et al. Light-emitting diodes by band-structure engineering in van der Waals heterostructures. Nat. Mater. 14, 301\u2013306 (2015). 22. Lee, H. S. et al. MoS2 Nanosheet Phototransistors with Thickness-Modulated Optical Energy Gap. Nano Lett. 12, 3695\u20133700 (2012). 23. Peng, B., Ang, P. K. & Loh, K. P. Two-dimensional dichalcogenides for light-harvesting applications. Nano Today 10, 128\u2013137 (2015). 24. Yamaguchi, M. III-V compound multi-junction solar cells: Present and future. Sol. Energy Mater. Sol. Cells 75, 261\u2013269 (2003). 25. Cressler, J. D. Silicon heterostructure handbook:Materials, Fabrication, Devices, Circuits, and Applications of SiGe and Si Strained-Layer Epitaxy. Materials Today 9, (2006). 26. Pospischil, A., Furchi, M. M. & Mueller, T. Solar-energy conversion and light emission in an atomic monolayer p\u2013n diode. Nat. Nanotechnol. 9, 257\u2013261 (2014). 27. Britnell, L. et al. Strong Light-Matter Interactions in heterostructures of atomically Thin Films. Science (80-. ). 340, 1331\u20131314 (2013). 28. Lopez-Sanchez, O. et al. Light generation and harvesting in a van der waals heterostructure. ACS Nano 8, 3042\u20133048 (2014). 29. Tsai, M. L. et al. Monolayer MoS2 heterojunction solar cells. ACS Nano 8, 8317\u20138322 (2014). 30. Yu, W. J. et al. Highly efficient gate-tunable photocurrent generation in vertical heterostructures of layered materials. Nat. Nanotechnol. 8, 952\u2013958 (2013). 31. Baugher, B. W. H., Churchill, H. O. H., Yang, Y. & Jarillo-Herrero, P. Optoelectronics with electrically tunable PN diodes in a monolayer dichalcogenide. 9, (2013). 32. Furchi, M. M. et al. Photovoltaic effect in an electrically tunable van der Waals heterojunction. Nano Lett 14, 4785\u20134791 (2014). 33. Mueller, T., Furchi, M. M., Zechmeister, A., Schuler, S. & Pospischil, A. Atomically-thin van der Waals Heterostructure Solar Cells. 8\u20139 (2015). 34. Long, M. et al. Broadband Photovoltaic Detectors Based on an Atomically Thin Heterostructure. Nano Lett. 16, 2254\u20132259 (2016). 95  35. Vos, A. De. Detailed balance limit of the efficiency of tandem solar cells. J. Phys. D. Appl. Phys. 13, 839\u2013846 (1980). 36. Polat, E. O. et al. Synthesis of Large Area Graphene for High Performance in Flexible Optoelectronic Devices. Sci. Rep. 5, 16744 (2015). 37. Huang, J.-K. et al. Large-Area and Highly Crystalline WSe2 Monolayers: from Synthesis to Device Applications. arXiv Prepr. arXiv \u2026 (2013). doi:10.1021\/nn405719x 38. Zhan, Y., Liu, Z., Najmaei, S., Ajayan, P. M. & Lou, J. Large-area vapor-phase growth and characterization of MoS 2 atomic layers on a SiO 2 substrate. Small 8, 966\u2013971 (2012). 39. Pizzocchero, F. et al. The hot pick-up technique for batch assembly of van der Waals heterostructures. Nat. Commun. 7, 11894 (2016). 40. Li, X. et al. Transfer of large-area graphene films for high-performance transparent conductive electrodes. Nano Lett. 9, 4359\u20134363 (2009). 41. Memaran, S. et al. Pronounced Photovoltaic Response from Multilayered Transition-Metal Dichalcogenides PN-Junctions. Nano Lett. 15, 7532\u20137538 (2015). 42. Wi, S. et al. Photovoltaic response in pristine WSe<inf>2<\/inf> layers modulated by metal-induced surface-charge-transfer doping. Appl. Phys. Lett. 107, (2015). 43. Shanmugam, M., Jacobs-Gedrim, R., Song, E. S. & Yu, B. Two-dimensional layered semiconductor\/graphene heterostructures for solar photovoltaic applications. Nanoscale 6, 12682\u20139 (2014). 44. Fontana, M. et al. Electron-hole transport and photovoltaic effect in gated MoS2 schottky junctions. Sci. Rep. 3, 1\u20136 (2013). 45. Shanmugam, M., Durcan, C. A. & Yu, B. Layered semiconductor molybdenum disulfide nanomembrane based Schottky-barrier solar cells. Nanoscale 4, 7399 (2012). 46. Wi, S. et al. Enhancement of photovoltaic response in multilayer MoS2 induced by plasma doping. ACS Nano 8, 5270\u20135281 (2014). 47. Fortin, E. & Sears, W. M. Photovoltaic effect and optical absorption in MoS2. J. Phys. Chem. Solids 43, 881\u2013884 (1982). 48. Evans, B. L. & Thompson, K. T. The photovoltage in thin crystals of MoS2. J. Phys. D. Appl. Phys. 1, 1619 (1968). 49. Flory, N. et al. A WSe2\/MoSe2 heterostructure photovoltaic device. Appl. Phys. Lett. 107, (2015). 50. Deng, Y. et al. Black phosphorus-monolayer MoS2 van der Waals heterojunction p-n diode. ACS Nano 8, 8292\u20138299 (2014). 51. Shanmugam, M., Bansal, T., Durcan, C. A. & Yu, B. Molybdenum disulphide\/titanium dioxide nanocomposite-poly 3-hexylthiophene bulk heterojunction solar cell. Appl. Phys. Lett. 100, 153901\u201312111 (2012). 96  52. Li, Y. et al. Measurement of the optical dielectric function of monolayer transition-metal dichalcogenides: MoS2, Mo S e2, WS2, and WS e2. Phys. Rev. B - Condens. Matter Mater. Phys. 90, 205422 (2014). 53. Wang, Q. H., Kalantar-Zadeh, K., Kis, A., Coleman, J. N. & Strano, M. S. Electronics and optoelectronics of two-dimensional transition metal dichalcogenides. Nat. Nanotechnol. 7, 699\u2013712 (2012). 54. Green, M. A. Radiative efficiency of state-of-the-art photovoltaic cells. Prog. Photovoltaics Res. Appl. (2012). doi:10.1002\/pip.1147 55. Jariwala, D., Davoyan, A. R., Wong, J. & Atwater, H. A. Van der Waals Materials for Atomically-Thin Photovoltaics: Promise and Outlook. ACS Photonics 4, 2962\u20132970 (2017). 56. Jariwala, D., Davoyan, A. R., Wong, J. & Atwater, H. A. Van der Waals Materials for Atomically-Thin Photovoltaics: Promise and Outlook. ACS Photonics 4, 2962\u20132970 (2017). 57. De Heer, W. A., Berger, C. & Wu, X. Epitaxial graphene. Solid State Commun. 143, 92\u2013100 (2007). 58. Cecilia Mattevi, Hokwon Kim & Manish Chhowalla. A review of chemical vapour deposition of graphene on copper. J. Mater. Chem. 21, 3324\u20133334 (2011). 59. Novoselov, K. S. et al. Two-dimensional atomic crystals. Proc. Natl. Acad. Sci. 102, 10451\u201310453 (2005). 60. Huang, Y. et al. Reliable Exfoliation of Large-Area High-Quality Flakes of Graphene and Other Two-Dimensional Materials. ACS Nano 9, 10612\u201310620 (2015). 61. Gao, E. et al. Mechanical exfoliation of two-dimensional materials. J. Mech. Phys. Solids 115, 248\u2013262 (2018). 62. Hao, Y. et al. The role of surface oxygen in the growth of large single-crystal graphene on copper. Science (80-. ). 342, 720\u2013723 (2013). 63. Dankert, A. et al. Spin-Polarized Tunneling through Chemical Vapor Deposited Multilayer Molybdenum Disulfide. ACS Nano acsnano.7b02819 (2017). doi:10.1021\/acsnano.7b02819 64. Pudasaini, P. R. et al. High-performance multilayer WSe2 field-effect transistors with carrier type control. Nano Res. 2018 112 11, 722\u2013730 (2017). 65. Zhao, P. et al. Air stable p-doping of WSe2 by covalent functionalization. ACS Nano 8, 10808\u201310814 (2014). 66. Li, H. et al. Rapid and reliable thickness identification of two-dimensional nanosheets using optical microscopy. ACS Nano 7, 10344\u201310353 (2013). 67. Castellanos-Gomez, A., Agrat, N. & Rubio-Bollinger, G. Optical identification of atomically thin dichalcogenide crystals. Appl. Phys. Lett. 96, 213116 (2010). 97  68. Furchi, M. M. et al. Photovoltaics in Van der Waals Heterostructures. IEEE J. Sel. Top. Quantum Electron. 23, 1\u201311 (2017). 69. Cheng, Z. et al. Toward intrinsic graphene surfaces: A systematic study on thermal annealing and wet-chemical treatment of SiO2-supported graphene devices. Nano Lett. 11, 767\u2013771 (2011). 70. Ferrari, A. C. et al. Raman spectrum of graphene and graphene layers. Phys. Rev. Lett. 97, 187401 (2006). 71. Website. Available at: https:\/\/www.filmetrics.com\/profilometers\/profilm3d?gclid=Cj0KCQjwvr6EBhDOARIsAPpqUPFI-FJn3P1M2HpwnXWbb6iq2HH70KTsN9MLUvUQpdhzIfQnmW9GfmwaAqCbEALw_wcB.  72. Yin, Z. et al. Single-layer MoS 2 phototransistors. ACS Nano 6, 74\u201380 (2012). 73. Das, S., Chen, H. Y., Penumatcha, A. V. & Appenzeller, J. High performance multilayer MoS2 transistors with scandium contacts. Nano Lett. 13, 100\u2013105 (2013). 74. Song, J. G. et al. Effect of Al2O3 Deposition on Performance of Top-Gated Monolayer MoS2-Based Field Effect Transistor. ACS Appl. Mater. Interfaces 8, 28130\u201328135 (2016). 75. Radisavljevic, B., Radenovic, A., Brivio, J., Giacometti, V. & Kis, A. Single-layer MoS2 transistors. Nat. Nanotechnol. 6, 147\u2013150 (2011). 76. Late, D. J., Liu, B., Matte, H. S. S. R., Dravid, V. P. & Rao, C. N. R. Hysteresis in single-layer MoS 2 field effect transistors. ACS Nano 6, 5635\u20135641 (2012). 77. Xu, S. et al. Universal low-temperature Ohmic contacts for quantum transport in transition metal dichalcogenides. 2D Mater. 3, 021007 (2015). 78. Kufer, D. & Konstantatos, G. Highly Sensitive, Encapsulated MoS2 Photodetector with Gate Controllable Gain and Speed. Nano Lett. 15, 7307\u20137313 (2015). 79. Radisavljevic, B., Radenovic, A., Brivio, J., Giacometti, V. & Kis, A. Single-layer MoS2 transistors. Nat. Nanotechnol. 2011 63 6, 147\u2013150 (2011). 80. Lee, C. H. et al. Tungsten Ditelluride: a layered semimetal. Sci. Reports 2015 51 5, 1\u20138 (2015). 81. Li, Q., Zhou, Q., Shi, L., Chen, Q. & Wang, J. Recent advances in oxidation and degradation mechanisms of ultrathin 2D materials under ambient conditions and their passivation strategies. J. Mater. Chem. A 7, 4291\u20134312 (2019). 82. No Title. Available at: https:\/\/cree-led.com\/media\/documents\/ds-MHDE.pdf.  83. No Title. Available at: https:\/\/www.digikey.ca\/en\/products\/detail\/MHDEWT-0000-000N0HG227G\/MHDEWT-0000-000N0HG227GCT-ND\/5087394?itemSeq=312765676.  84. Cao, Y. et al. Quality Heterostructures from Two-Dimensional Crystals Unstable in Air by Their Assembly in Inert Atmosphere. Nano Lett. 15, 4914\u20134921 (2015). 98  85. Bandurin, D. A. et al. High electron mobility, quantum Hall effect and anomalous optical response in atomically thin InSe. Nat. Nanotechnol. 2016 123 12, 223\u2013227 (2016). 86. Wang, L. et al. One-Dimensional Electrical Contact to a Two-Dimensional Material. Science (80-. ). 342, 614\u2013617 (2013). 87. Kim, K. et al. Van der Waals Heterostructures with High Accuracy Rotational Alignment. Nano Lett. 16, 1989\u20131995 (2016). 88. Allain, A., Kang, J., Banerjee, K. & Kis, A. Electrical contacts to two-dimensional semiconductors. Nat. Mater. 2015 1412 14, 1195\u20131205 (2015). 89. Utama, M. I. B. et al. A dielectric-defined lateral heterojunction in a monolayer semiconductor. Nat. Electron. 2, 60\u201365 (2019). 90. Schulman, D. S., Arnold, A. J. & Das, S. Contact engineering for 2D materials and devices. Chemical Society Reviews 47, 3037\u20133058 (2018). 91. Lee, C.-H. et al. Atomically thin p\u2013n junctions with van der Waals heterointerfaces. Nat. Nanotechnol. 9, 676\u2013681 (2014). 92. Furchi, M. M. et al. Photovoltaic effect in an electrically tunable van der Waals heterojunction. Nano Lett 14, 4785\u20134791 (2014). 93. Roy, T. et al. Dual-Gated MoS 2 \/WSe 2 van der Waals Tunnel Diodes and Transistors. ACS Nano 9, 2071\u20132079 (2015). 94. McDonnell, S., Smyth, C., Hinkle, C. L. & Wallace, R. M. MoS2-Titanium Contact Interface Reactions. ACS Appl. Mater. Interfaces 8, 8289\u20138294 (2016). 95. English, C. D., Shine, G., Dorgan, V. E., Saraswat, K. C. & Pop, E. Improved contacts to MoS2 transistors by ultra-high vacuum metal deposition. Nano Lett. 16, 3824\u20133830 (2016). 96. Bolshakov, P. et al.  Contact Engineering for Dual-Gate MoS 2 Transistors Using O 2 Plasma Exposure . ACS Appl. Electron. Mater. 1, 210\u2013219 (2019). 97. Skriver, H. L. & Rosengaard, N. M. Surface energy and work function of elemental metals. Phys. Rev. B 46, 7157\u20137168 (1992). 98. Fontana M., Deppe T., Boyd A.K., Rinzan M., Liu A.Y., Paranjape M., B. P. Photovoltaic effect in gated MoS2 schottky junctions. Arxiv Prepr. (2012). 99. Shimazu, Y., Arai, K. & Iwabuchi, T. Contact-induced doping in aluminum-contacted molybdenum disulfide. Jpn. J. Appl. Phys. 57, (2017). 100. Cui, X. Achieving Ohmic Contact for High-quality MoS2 Devices on Hexagonal Boron Nitride. (2018). doi:10.7916\/D84Q86J6 101. Kaushik, N., Karmakar, D., Nipane, A., Karande, S. & Lodha, S. Interfacial n-Doping Using an Ultrathin TiO2 Layer for Contact Resistance Reduction in MoS2. ACS Appl. Mater. Interfaces 8, 256\u2013263 (2016). 102. Fang, H. et al. High-performance single layered WSe 2 p-FETs with chemically doped 99  contacts. Nano Lett. 12, 3788\u20133792 (2012). 103. Ohl, R. S. Light-sensitive electric device. (1946). 104. Sze, S. M. & Lee, M. K. Semiconductor devices\u2009: physics and technology. (2012). 105. Furchi, M. M., Pospischil, A., Libisch, F., Burgdo, J. & Mueller, T. Photovoltaic E ff ect in an Electrically Tunable van der Waals Heterojunction. (2014). 106. Zhang, W. et al. Role of metal contacts in high-performance phototransistors based on WSe2 monolayers. ACS Nano 8, 8653\u20138661 (2014). 107. Moazzami, K., Murphy, T. E., Phillips, J. D., Cheung, M. C.-K. & Cartwright, A. N. Sub-bandgap photoconductivity in ZnO epilayers and extraction of trap density spectra. Semicond. Sci. Technol 21, 717\u2013723 (2006). 108. Zhang, W. et al. High-gain phototransistors based on a CVD MoS2 monolayer. Adv. Mater. 25, 3456\u20133461 (2013). 109. Liu, W. et al. Role of metal contacts in designing high-performance monolayer n-type WSe2 field effect transistors. Nano Lett. 13, 1983\u20131990 (2013). 110. Chapin, D. M., Fuller, C. S. & Pearson, G. L. A new silicon p-n junction photocell for converting solar radiation into electrical power [3]. J. Appl. Phys. 25, 676\u2013677 (1954). 111. Wolf, M. Limitations and Possibilities for Improvement of Photovoltaic Solar Energy Converters: Part I: Considerations for Earth\u2019s Surface Operation. Proc. IRE 48, 1246\u20131263 (1960). 112. Esaki, L., Esaki & Leo. New Phenomenon in Narrow Germanium p-n Junctions. PhRv 109, 603\u2013604 (1958). 113. Guter, W. et al. Current-matched triple-junction solar cell reaching 41.1% conversion efficiency under concentrated sunlight. Appl. Phys. Lett. 94, 223504 (2009). 114. Dimroth, F. et al. Wafer bonded four-junction GaInP\/GaAs\/\/GaInAsP\/GaInAs concentrator solar cells with 44.7% efficiency. Prog. Photovoltaics Res. Appl. 22, 277\u2013282 (2014). 115. Mart\u00ed, A. & Luque, A. Three-terminal heterojunction bipolar transistor solar cell for high-efficiency photovoltaic conversion. Nat. Commun. 6, 6902 (2015). 116. Fang, H. et al. Strong interlayer coupling in van der Waals heterostructures built from single-layer chalcogenides. Proc. Natl. Acad. Sci. 111, 6198\u20136202 (2014). 117. Mart\u00ed, A. et al. Operation of the Three Terminal Heterojunction Bipolar Transistor Solar Cell. Phys. status solidi c 14, 1700191 (2017). 118. Conley, H. J. et al. Bandgap engineering of strained monolayer and bilayer MoS2. Nano Lett. 13, 3626\u20133630 (2013). 119. Kumar, H., Dong, L. & Shenoy, V. B. Limits of Coherency and Strain Transfer in Flexible 2D van der Waals Heterostructures: Formation of Strain Solitons and Interlayer Debonding. Sci. Reports 2016 61 6, 1\u20138 (2016). 100  120. Susarla, S. et al. Deformation Mechanisms of Vertically Stacked WS 2 \/MoS 2 Heterostructures: The Role of Interfaces. ACS Nano 12, 4036\u20134044 (2018). 121. Yankowitz, M. et al. Tuning superconductivity in twisted bilayer graphene. Science (80-. ). 363, 1059\u20131064 (2019). 122. Park, J. M., Cao, Y., Watanabe, K., Taniguchi, T. & Jarillo-Herrero, P. Tunable strongly coupled superconductivity in magic-angle twisted trilayer graphene. Nat. 2021 5907845 590, 249\u2013255 (2021). 123. Pochet, P., McGuigan, B. C., Coraux, J. & Johnson, H. T. Toward Moir\u00e9 engineering in 2D materials via dislocation theory. Appl. Mater. Today 9, 240\u2013250 (2017). 124. Kim, K. et al. Band Alignment in WSe2 \u2013Graphene Heterostructures. ACS Nano 9, 4527\u20134532 (2015). 125. Movva, H. C. P. et al. High-Mobility Holes in Dual-Gated WSe2 Field-Effect Transistors. ACS Nano 9, 10402\u201310410 (2015). 126. Yamaguchi, M. Multi-junction solar cells and novel structures for solar cell applications. Phys. E Low-Dimensional Syst. Nanostructures 14, 84\u201390 (2002). 127. Soltanian, S. et al. Highly stretchable, sparse, metallized nanofiber webs as thin, transferrable transparent conductors. Adv. Energy Mater. 3, 1332\u20131337 (2013). 128. Lin, S. Y., Wang, C. M., Kao, K. S., Chen, Y. C. & Liu, C. C. Electrochromic properties of MoO3 thin films derived by a sol-gel process. J. Sol-Gel Sci. Technol. 53, 51\u201358 (2010). 129. Girotto, C., Voroshazi, E., Cheyns, D., Heremans, P. & Rand, B. P. Solution-Processed MoO 3 Thin Films As a Hole-Injection Layer for Organic Solar Cells. ACS Appl. Mater. Interfaces 3, 3244\u20133247 (2011).            101  Appendix: Work done outside the scope of the thesis  While the main scope of this thesis is focused on atomically thin materials for optoelectronics, other nanotechnology-related projects are also explored. Some of them are directly and some are indirectly related to the goals of the thesis. Exploring these projects was helpful in expanding the knowledge, skills, and ideas needed to achieve the proposed goals.  For instance, experience with flexible electronics was helpful with the idea proposed in section 3.1 and will make its fabrication and characterization easier once reached. Experiments with growing Organic Light Emitting Diodes (OLEDs) in the glove box brought up the idea of exploring 2D material heterostructures for light-emitting device applications. This experience was also helpful in considering using glove box for experiments on air-sensitive 2D materials such as Phosphorene. The growth of graphene is investigated for potential use in the devices presented in this work. Chemically vapor deposited graphene has been grown on Copper and transferred to SiO2 substrate. In order to achieve large crystals of graphene, ultra-smooth Copper foil has been used as suggested by Polat et al.36. Due to the time-consuming procedure of study of the growth of 2D materials, this research has been limited to several growth and transfers. As an example, one of these side projects is briefly presented below. A.1. Nanoparticle-decorated nanofibers for transparent conductors Indium tin oxide (ITO) films are currently the most widely used transparent conductors for displays, lighting devices, and solar cells. Due to ITO films\u2019 brittle nature, they are not suitable for flexible and stretchable applications. Soltanian et al.127 have shown gold metalized nanofiber webs as transparent conductors, scalable for the manufacturing of large area transparent 102  conductors for flexible and stretchable applications. In this work, we build up on electrospinned Polyacrylonitrile (PAN) nanofibers and coat them with conductive nanoparticles, aiming to achieve high conductivity and high transparency with affordable materials. The nanoparticles are processed based on Lin et al. 128 and Girotto et al. 129 work and spin-coated on nanofibers. The figures below show a set of measurements for the transparency and sheet resistance of these films.    Figure A-1. Nanofibers coated with ZnO and MoO3 nanoparticles are treated in room temperature up to 300 C and transparency is measured at 550nm wavelength. 103   Figure A-2. Sheet resistance of Nanofibers coated with ZnO and MoO3 nanoparticles treated from room temperature (23 C) to 300 C.  ","type":"literal","lang":"en"}],"http:\/\/www.europeana.eu\/schemas\/edm\/hasType":[{"value":"Thesis\/Dissertation","type":"literal","lang":"en"}],"http:\/\/vivoweb.org\/ontology\/core#dateIssued":[{"value":"2024-05","type":"literal","lang":"en"}],"http:\/\/www.europeana.eu\/schemas\/edm\/isShownAt":[{"value":"10.14288\/1.0438661","type":"literal","lang":"en"}],"http:\/\/purl.org\/dc\/terms\/language":[{"value":"eng","type":"literal","lang":"en"}],"https:\/\/open.library.ubc.ca\/terms#degreeDiscipline":[{"value":"Electrical and Computer Engineering","type":"literal","lang":"en"}],"http:\/\/www.europeana.eu\/schemas\/edm\/provider":[{"value":"Vancouver : University of British Columbia Library","type":"literal","lang":"en"}],"http:\/\/purl.org\/dc\/terms\/publisher":[{"value":"University of British Columbia","type":"literal","lang":"en"}],"http:\/\/purl.org\/dc\/terms\/rights":[{"value":"Attribution-NonCommercial-NoDerivatives 4.0 International","type":"literal","lang":"*"}],"https:\/\/open.library.ubc.ca\/terms#rightsURI":[{"value":"http:\/\/creativecommons.org\/licenses\/by-nc-nd\/4.0\/","type":"literal","lang":"*"}],"https:\/\/open.library.ubc.ca\/terms#scholarLevel":[{"value":"Graduate","type":"literal","lang":"en"}],"http:\/\/purl.org\/dc\/terms\/contributor":[{"value":"Servati, Peyman","type":"literal","lang":"en"}],"http:\/\/purl.org\/dc\/terms\/title":[{"value":"Optoelectronics with two-dimensional atomic crystals","type":"literal","lang":"en"}],"http:\/\/purl.org\/dc\/terms\/type":[{"value":"Text","type":"literal","lang":"en"}],"https:\/\/open.library.ubc.ca\/terms#identifierURI":[{"value":"http:\/\/hdl.handle.net\/2429\/87251","type":"literal","lang":"en"}]}}