Florida State University Libraries

Transcription

1 Florida State University Libraries Electronic Theses, Treatises and Dissertations The Graduate School 2006 Nanotube and Nanofiber Buckypaper Cold Cathode Illumination: Experimental Investigation Yi Wen (Evin) Chen Follow this and additional works at the FSU Digital Library. For more information, please contact

2 THE FLORIDA STATE UNIVERSITY FAMU-FSU COLLEGE OF ENGINEERING NANOTUBE AND NANOFIBER BUCKYPAPER COLD CATHODE ILLUMINATION: EXPERIMENTAL INVESTIGATION By YI WEN (EVIN) CHEN A Thesis submitted to the Department of Industrial Engineering in partial fulfillment of the requirements for the degree of Master of Science Degree Awarded: Summer Semester 2006

3 The members of the Committee approve the thesis of Yi-Wen Chen defended on 06/30/2006. Ben Wang Professor Directing Thesis Zhiyong Liang Committee Member Chuck Zhang Committee Member Young-Bin Park Committee Member Approved. Chuck Zhang, Chair, Department of Industrial Manufacturing Engineering C. J. Chen, Dean, College of Engineering The Office of Graduate Studies has verified and approved the above named committee members. ii

4 ACKNOWLEDGEMENTS I would like to express my sincere appreciation for Dr. Ben Wang, not only for his extensive support and guidance throughout my graduate study, but also for inspiring me to think outside the box. Moreover, Dr. Zhiyong Liang, Dr. Chunk Zhang, and Young-Bin Park always gave their wonderful encouragement and excellent suggestions to my research at Florida State University. Additionally, I would like to give special thanks to Dr. Jin Gyu Park, Dr. Hsin-Yuan Miao, Ashley Liao, Edward Wang, Giang Pham, Yong Guo, Irene Yeh, Charlie Lin, Kenny Tsai, Craig Chin, Daphne Ku, Vavinlen Chen and all my co-workers at Florida Advanced Center for Composite Technologies (FACCT) for their help and contributions to this thesis. Furthermore, I would like to acknowledge Dr. Andrew Wang in Tunghai University and Taiwan ITRI ERSO for their valuable assistances and for the wonderful opportunity to participate in the development of CNT display technology. Lastly, I would like to thank my parents and families for their unconditional love and support all through my life. iii

5 TABLE OF CONTENTS LIST OF TABLES... vii LIST OF FIGURES....viii ABSTRACT... xii CHAPTER 1: INTRODUCTION General Introduction of Display Technology Introduction of display technology The principle of LCD Areas of Improvement Issues of current BLU Alternative BLU: CNT-BLU Issues of current CNT-BLU Research issues Research Motivation Research Objectives Design and investigation of CNT field emission apparatus Investigation of the effects of different compositions of BuckyPaper films properties on turn-on voltage, current density, and power efficiency Investigation of the effects of different compositions of open-ended CNT paste on turn-on voltage, current density, and power efficiency Preliminary study of BuckyPaper field emission durability Application suggestion CHAPTER 2: LITERATURE REVIEW Electron Field Emission Theory Field Emission Properties of CNT SWNT vs. MWNT Fabrication of CNT Aspect ratio verse field enhancement factor Emission current-voltage measuremen Emission stability of CNT Applications: Field Emission Display and Back Lighting Progress Made by Major Manufacturers Summary of Literature Review iv

6 CHAPTER 3: BUCKYPAPER FILMS AND CNT PASTE FIELD EMISSION PROPERTIES TESTING METHODOLOGY AND STANDARD Introduction Experiment Apparatus Experiment Criteria Specimen Preparation BuckyPaper specimen preparation procedure CNT Paste specimen preparation procedure Test Procedures CHAPTER 4: FIELD EMISSION AND DURABILITY PROPERTIES OF BUCKYPAPER Introduction Field Emission Theory and Durability Principle Field Emission Properties of BuckyPapers Comprehensive testing for all BuckyPaper categories Current-Voltage (I-V) properties Fowler-Nordheim (F-N) properties Summary Selected testing for high current output BuckyPaper Current-Voltage (I-V) properties Summary Durability and Repeatability Properties of BuckyPaper Repeatability of BuckyPaper for 1.5 hours Durability of BuckyPaper for 20 hours Durability of BuckyPaper with feedback control system Summary CHAPTER 5: FIELD EMISSION AND DURABILITY PROPERTIES OF CNT PASTE Introduction Field Emission Properties of Nanoparticle Pastes Current-Voltage (I-V) properties Fowler-Nordheim (F-N) properties Summary Durability and Repeatability Properties of Nanoparticle Pastes Repeatability of paste emitters for 1.5 hours Durability of paste emitter for 20 hours v

7 5.3.3 Durability of paste emitter with feedback control system Summary CHAPTER 6: DATA ANALYSIS AND COMPARISON Vertically Grown CNT on a Substrate CNT Pastes CNT BuckyPaper Films Performance and Property Comparison for Emitters Made in Various Methods CHAPTER 7: CONCLUSIONS Summary of Observations and Conclusions Proposed Future Research REFERENCES BIOGRAPHICAL SKETCH vi

8 LIST OF TABLES 1.1 Dominant display technologies in various display size Effect of bundle length on field emission of SWNTs Turn-on voltage comparison of all BuckyPaper types The β values of four selected BuckyPaper types Turn-on voltage comparison of three nanoparticle pastes The β values of various CNT paste types Comparison of I-V performance of different grow methods Comparison of I-V performance of different CNT mixture emitters Comparison of different CNT emitters vii

9 LIST OF FIGURES 1.1 The global FED market by technology category A concise scheme of LCD A rotation schematic of liquid crystal Cost analysis: (a) 32 LCD materials, (b) BLU module materials Schematic of BLU replacement Comparison figure of threshold voltage of various CNTs Electron emitted current versus applied voltage of a SWNT film measured at different anode cathode distances using a hemispherical anode I-V and F-N plot of diode configuration carbon nanotube film Emitted current as a function of time Stability of the emission current measured under a constant applied electric field from a 0.2 cm 2 area SWNT film Photos of CNT emitters illuminating three primary colors-green, red, blue-on an anode current of 200 µa, and a voltage of 10 kv The color image from (a) 4.5 three color display (b) 9 full-color FED The images of (a) 7 CNT-BLU by Dialight (b) 5 CNT-BLU by LG (c) 20 CNT-BLU & LCD by ERSO Categories and criteria of BuckyPaper films and CNT pastes Schematic of field emission apparatus (a) Bell jar chamber; (b) Varian Inc. turbo cart pumping system; (c) Kethley 2410 source meter Schematic of BuckyPaper field emission BuckyPaper field emission trail samples Schematic of open-ended CNT paste field emission Open-ended CNT paste sample Schematic of field emission test procedures Randomly dispersed BuckyPaper sample viii

10 4.2 Fowler-Nordheim Plot for a nanotube matrix sample BuckyPaper types and identification codes I-V Curves of randomly dispersed SWNT BuckyPaper I-V Curves of aligned SWNT BuckyPaper I-V Curves of VGCNF/SWNT in ratio 1:1 BuckyPaper I-V Curves of VGCNF/SWNT in ratio 3:1 BuckyPaper I-V Curves of 58/61 in ratio 1:1 BuckyPaper I-V Curves of e-beam irradiated SWNT BuckyPaper I-V Curves of e-beam irradiated SWNT BuckyPaper The I-V curves of all BuckyPaper types F-N curve of randomly dispersed SWNT BuckyPaper F-N curve of N61-1 BuckyPaper F-N curve of N61-2 BuckyPaper F-N curve of M61 BuckyPaper High current output I-V curve of randomly dispersed BuckyPaper High current output I-V curve of N61-1 BuckyPaper High current output I-V curve of 58/61 BuckyPaper Repeatability of randomly dispersed SWNT BuckyPaper Repeatability of N61-1 BuckyPaper Repeatability of 58/61 BuckyPaper Repeatability of aligned SWNT BuckyPaper Repeatability of M61 BuckyPaper: (a) Backside; (b) Frontside Durability of N61-1 BuckyPaper: (a) Backside; (b) Frontside; (c) Logarithm of frontside sample; (d) Logarithm of backside sample Durability of N61-2 BuckyPaper: (a) Backside; (b) Frontside; (c) Logarithm of frontside sample; (d) Logarithm of backside sample Durability of 58/61BuckyPaper: (a) Backside; (b) Frontside; (c) Logarithm of frontside sample; (d) Logarithm of backside sample Durability of M61 BuckyPaper (a) Backside; (b) Frontside; (c) Logarithm of ix

11 frontside sample; (d) Logarithm of backside sample Durability of aligned SWNT BuckyPaper (a) Backside; (b) Frontside Durability of randomly dispersed SWNT BuckyPaper with feedback control system Durability of N61-1 BuckyPaper with feedback control system Durability of M61 BuckyPaper with feedback control system Nanoparticle paste types and identification codes I-V Curve of 400 nm open-ended SWNT paste I-V Curve of 50 nm open-ended SWNT paste I-V Curve of carbon nanosphere paste F-N curve of 400 nm open-ended SWNT paste F-N curve of 50 nm open-ended SWNT paste F-N curve of carbon nanosphere paste Repeatability of 400 nm open-ended SWNT paste Repeatability of 50 nm open-ended SWNT paste Repeatability of carbon nanosphere paste Durability of 400 nm open-ended SWNT paste (a) Durability of sample; (b) Logarithm current of sample Durability of 50 nm open-ended SWNT paste: (a) Durability of sample; (b) Logarithm current of sample Durability of Carbon Nanosphere paste (a) Durability of sample; (b) Logarithm current of sample Durability of 400 nm open-ended SWNT paste with feedback control Durability of 50 nm open-ended SWNT paste with feedback control Durability of carbon nanosphere paste with feedback control (a) I-V curves of various vertically grown CNT lengths; (b) SEM image of the CNT emitters (a) I-V curves of grown vertically CNT in different conditions; (b) SEM image of the CNT emitters x

12 6.3 (a) I-V curves of thermal CVD grown vertically aligned CNT; (b) SEM image of the CNT emitters (a) I-V curves of laser ablation grown vertically CNT; (b) SEM image of the CNT emitters (a) I-V curves of various CNT paste; (b) SEM image of the CNT paste emitter (a) I-V curves of LTCC CNT paste; (b) SEM image of the CNT paste emitter I-V curve of sprinkled open-end SWNT paste (a) I-V curves of various BuckyPaper films; (b) SEM image of the BuckyPaper emitter xi

13 ABSTRACT Since the first Cathode Ray Tube (CRT) TV was invented in 1927, display technology has progressed at a rapid speed, attracting tremendous attention and enormous resources. Liquid Crystal Display (LCD) is the most mature and popular technology in flat panel displays compared to CRT. Carbon nanotube backlight unit (CNT-BLU) was regarded as a strong contender to replace the cold cathode fluorescence lamp backlight unit (CCFL-BLU) in LCD. CNTs have been spotlighted as one of promising alternatives for new electron sources; small tips and large aspect ratios of CNTs allow for a large electric field enhancement that makes them ideal field emitters. In this research, nanotube network BuckyPaper was proposed to act as surface luminary source for BLU. The major objective of this research is to systematically characterize various field emission properties of BuckyPapers and various lengths of open-ended CNTs. Particularly, low turn-on voltage, high average current, high luminance, low power consumption, uniformity and longer life-span will be explored. VGCNF/SWNT in ratio 1:1 BuckyPaper film demonstrated the lowest turn-on voltage (0.623 V/um) and randomly dispersed SWNT BuckyPaper film showed the largest enhancement factor value (1062) among all BuckyPaper samples. The structure of emitters might be the major reason for the varying results from different compositions of BuckyPaper films. The 400 nm open-ended SWNT paste showed the lowest turn on voltage at V/um and the largest enhancement factor value (3470). Open-ended SWNT paste exhibited the best I-V properties, but did not demonstrate acceptable durability. The effect of varying aspect ratio of open-ended SWNT will be investigated in the future. xii

14 CHAPTER 1 INTRODUCTION 1.1 General Introduction of Display Technology Introduction of display technology Since the first Cathode Ray Tube (CRT) TV was invented in 1927, display technology has progressed at a rapid speed and attracted tremendous attention. However, more flat panel display technologies have been developed due to several drawbacks in CRT displays such as large size, heavy weight, high radiation, and bad image quality. Currently, Liquid Crystal Displays (LCDs), Plasma Displays (PDPs), Field Emission Displays (FEDs), Light Emitting Diodes (LEDs), Organic Light Emitting Diodes (OLEDs), and Projection TV are major products in the display industry [1]. Table 1.1 shows the major technologies in various sizes. LCD technology is being looked upon as a strong candidate to replace CRT in large sized display. LCD is the most mature and popular technology in flat panel display (FPD) products compared to CRT because of its advantage of lighter weight, less power consumption, faster heat releasing, portability, and bright color images. The global sales of display have kept growing from $ 26 billion in 2002 to $ 48.2 billion in The forecast of LCD sales indicated that it will increase to $ 55.1 billion in 2006 [2]. 1

15 Table 1.1 Dominant display technology in various display size Display Size Dominant Display Technology 4 TN-LCD, STN-LCD, TFT-LCD, OLED 10 ~30 TFT-LCD 25 ~40 CRT 35 ~80 PDP >80 Projection Display Figure 1.1 The global FED market by technology category Most of the major FPD players are in Asia: Japan, Korea, and Taiwan. Many corporations such as LG Corporation, Philips and Samsung in Korea; Sharp, NEC, and Toshiba in Japan; and AUO, and Chi Mei in Taiwan invested large resources in manufacturing capability of FPD. Not only the manufacturing capability but also the design and technology innovation are the major keys to being the winner. It is reasonable to predict that the technology innovation will speed up the manufacturing process and shorten the product life cycle. Besides, not only are FDPs used for computer display, but they also are increasingly being used in many other aspects of life, for instance, PDAs, GPS navigations, and televisions [1]. 2

16 1.1.2 The principle of LCD A LCD uses bending of liquid crystal to control the passage of light. The basic structure (Figure 1.2) of a LCD panel may be thought as two glass substrates sandwiching a layer of liquid crystal. The front glass substrate is fitted with a color filter, while the back glass substrate has transistors fabricated on it. When voltage is applied to a transistor, the liquid crystal is bent, allowing light to pass through to form a pixel. A light source is located at the back of the panel and is called a backlight unit. The color filter gives each pixel its own color. The combination of these pixels in different colors forms the image on the panel [3]. Figure 1.2 A concise schematic of LCD Figure 1.3 The rotation schematic of liquid crystal 3

17 1.2 Area of Improvement Issues of current BLU Flat display technologies include LCD, PDP, FED, Projection TV, LED, and OLED. Among all of them, LCD has become the major technology in the display market. However, there are many issues associated with manufacturing large-size LCD. One of the limitations is Cold Cathode Fluorescent Lamps Backlight Unit (CCFL-BLU) in LCD. There are a few restrictions among the conventional CCFL-BLU for current commercial LCD technology compared to CNT-BLU. The heat from the phosphor layer is difficult to be released because of the lighting mechanism of CCFL. Moreover, the phosphor layer affects the liquid crystal or polarizer films and results in LCD damage. Secondly, CCFL costs more, weights more, and is less efficient than CNT-BLU due to those additional enforced lightened films such as diffuser, light guide, and reflector. Those films make up 70% of the BLU cost in the LCD, and the light leakage of lighten films reduces the light emitting efficiency. CCFL-BLU has additional issues such as high patent royalty fees and adversarial environmental impact which reduce the competitiveness of LCD industry compared to emerging technologies. Currently, CCFL is the main light source in LCDs. As it can be seen in Figure 1.4 (a), BLU occupies almost 40% in the total cost of LCD for a 32-inch display. The BLU cost increases rapidly as the size of LCD enlarges. When the size of LCD increases, the cost rises quickly due to the increasing cost in the BLU which includes brightness enhancement film (BEF), prism sheet, and diffusion film. Therefore, BLU would be a key limiting factor for LCD to be a dominant technology in large sized display. Figure 1.4 (b) shows the cost analysis of CCFL and those brightness enhancement films in a BLU module [4]. 4

18 Figure 1.4 Cost analyses: (a) 32 LCD materials, (b) BLU module materials Alternative BLU: CNT-BLU LCD looks firmly entrenched as the dominant flat-panel technology, but in fact, the display industry is keen to find a technology that delivers better performance at a lower cost. CNT-FED technology could be the candidate for the LCD BLU because of its unique properties compared with traditional backlight technologies. CNT-BLU has many advantages such as low surface temperature and low processing costs. Light guide, reflector film, diffuser film, BEF film will not be necessary and can be eliminated to save cost. Moreover, mercury and other toxic gas will be removed. Therefore, a large number of researchers plunged enormous resources into this plane illumination of CNT-BLU to replace linear illumination of CCFL [4]. (Figure 1.5) 5

19 Figure 1.5 The schematic of BLU replacement Issues of current CNT-BLU CNT-BLU has been promoted as a large size plane luminary to avoid those problems of CCFL. In recent years, resources have poured into CNT-based luminary. However, current CNT-based luminary still has several problems in the fabrication process. For instance, unstable current makes CNT-based luminary unreliable, a low pressure field emission mechanism produces a lower level of brightness, and CNT paste and sealing process of CNT-based luminary result in bad quality uniformity Research issues Current progress in the production of CNTs in luminary cold cathode is investigated by chemical vapor deposition (CVD) technique as an alternative means of fabricating nanotubes pastes [5]. However, disadvantages such as high cost, low performance, and complexity and premature status for mess production still exist. This research involves investigation and analysis of a process and various properties of making a larger size surface illumination source using BuckyPaper which is a thin 6

20 sheet formed by nanotubes with controlled dispersion, and open-ended CNTs in various lengths. We have demonstrated cold cathode illumination by using BuckyPaper instead of using screen printed CNTs. BuckyPaper films produced as a thin sheet and formed with well-dispersed carbon nanotubes was investigated by Florida Advanced Center for Composite Technologies (FACCT) researchers [6]. Therefore, in this research, we will characterize and analyze the field emission properties of different BuckyPaper films to further reveal their performance and establish a database for future applications. Moreover, alternative methods of manufacturing CNT cathode prepared by mixing CNT powder and some materials such as silver conductive adhesive, frit, and other inorganic and organic compounds are also studied. The field emission properties of open-ended CNTs in paste form also will be analyzed. Finally, a CNT field emission apparatus has been designed, and installed and application suggestions are discussed. 1.3 Research Motivation CNTs have been spotlighted as one of promising alternatives for new electron sources due to their superior emission properties and relatively simple fabrication processes. Small tips and large aspect ratios of CNTs allow for a large electric field enhancement near the tip that makes them ideal field emitters [7]. CNTs are expected to make it possible to produce emission resources for display, lighting, x-ray, or plasma ignition with low cost, low power consumption, scalability, and etc [8]. With CNTs superior field emission electronic properties, they can be used as a new luminary source at low cost, low surface temperature, large size plane luminary, low power consumption, and pulse-type driving in displays for televisions or computers. One promising technology is CNT backlight unit (CNT-BLU), which is now considered by many as a perfect replacement for the contemporary CCFL-BLU 7

21 in LCD [4]. In this research, nanotube BuckyPaper films are used as surface luminary source for flat panel displays. A BuckyPaper film has properties such as uniform nanotube dispersion, high nanotube loading, easy modification, strong interfacial bonding, and controlled nanotube orientation [6]. In order to replace current limited linear and point luminary such as CCFL, BuckyPapers are expected to be a plane field emitter due to the excellent field emission properties of CNT. Furthermore, electronic effects on emission properties of open-ended SWNTs have been studied in the research by Zhou et al. [9]. Nevertheless, research on the effects of the field emission properties of open-ended CNTs at various lengths is lacking. We are analyzing the alternative plane field emitter candidate, CNT paste, to conclude the optimal length of CNT in paste and provide application suggestions. 1.4 Research Objectives The major objective of this research is to demonstrate the feasibility of using BuckyPaper as backlight source. Particularly, low turn-on voltage, high average current, high luminance, low power consumption, uniformity and long life-span will be investigated. In order to do so, several tasks must be accomplished: Design and investigation of a CNT field emission apparatus We will develop an experimental apparatus that can accurately characterize the field emission of BuckyPaper and CNT paste. In preliminary studies, field emission experiments were performed in a field emission measurement system (FEMS) with a base pressure below 10-7 torr. Current voltage (I-V) curve tests were performed under continuous (nonpulsed) bias conditions. Emission current measurements were recorded in a computer, allowing us to track current versus time at each voltage level. Continued illumination will involve observing in a visible chamber and measure the luminance by a luminance meter [10, 11, and 12]. 8

22 1.4.2 Investigation of the effects of different compositions of BuckyPaper films on turn-on voltage, current density, and power efficiency In adequate experiment conditions, we will implement a series of tests to examine the various properties of BuckyPaper to characterize I-V properties and power consumption. BuckyPaper properties for both aligned and random, and oriented at the front side and membrane side will be examined. E-beam treated BuckyPaper will be tested by the assembled field emission measurement system (FEMS) and characterized for their threshold voltage, average current, and power consumption. Based on the analysis of BuckyPaper properties, a more comprehensive database of BuckyPaper field emission properties and various application suggestions will be compiled Investigation of the effects of different compositions of open-ended CNT paste on turn-on voltage, current density, and power efficiency Moreover, another objective of this project is examining and characterizing I-V properties and power consumption of chopped, open-ended CNT paste by FEMS. Based on the analysis of BuckyPaper properties, a more comprehensive property database of BuckyPaper field emission properties and various application suggestions will be compiled Preliminary study of BuckyPaper field emission durability Jo et al. has investigated one parameter of stability of CNT field emission properties. Although the field emission capability can be affected by the apparent differences of the length and the density, they proposed that the stability might be affected by other factors such as surface condition [13]. In our research, a preliminary study of BuckyPaper field emission durability will be examined to verify that BuckyPaper films are excellent field emission emitter candidates. 9

23 1.4.5 Application suggestion Based on this research, we will provide application potentials analysis of the BuckyPaper and the CNT paste with properly chopped open-ended CNTs. 10

24 CHAPTER 2 LITERATURE REVIEW Since the discovery of carbon nanotubes (CNTs) by Iijima in the NEC Fundamental Research Laboratory in Tsukuba, Japan in 1991 [14], the research in CNT continues to expand. CNTs have remarkable mechanical, electronic, and magnetic properties and attracted much attention [14]. CNTs have the right combination of properties: nanometer-size diameter, structural integrity, high electrical and thermal conductivity, and chemical stability, which make them excellent electron emitters for a variety of applications, such as electronic and display devices [15-20]. Electron field emission from CNTs was first demonstrated in 1995 [15], and has since been studied extensively. In recent years, enormous efforts have been made to produce field emission devices using CNTs, such as flat lamp and display cold cathode [21]. 2.1 Electron Field Emission Theory Electron field emission or Fowler-Nordheim quantum mechanical tunneling is the process of releasing electrons from metal surfaces via barrier penetration in a high electric field [22]. A high electric field near the emitter can lower the barrier strength to increase emission substantially [23]. The electron current density can reach 10-7 A/cm 2 and does not delay in emission [24]. In 1937, Muller first used this phenomenon to devise a new way to do microscopy [25]. Muller was successful in using the extracted electrons from a field-emitting metal tip to create an image of the tip, revealing its surface structure. Experimentally, cold field emission requires a sharp tip, high aspect ratio, and ultra high vacuum (UHV) conditions. Field emission occurs when a voltage difference 11

25 exists between cathode and anode. The Fowler-Nordheim model for field emission describes the electron current density, J, emitting from a surface into vacuum as a function of the applied field E. The model parameters that govern the emission are the material (effective work function,φ), the area available for emission, Ae, and a geometric enhancement factor, β. B 1 and B 2 are known fundamental constants; I is the prebreakdown current; and d is gap distance. The enhancement factor describes how electric fields can be enhanced by protrusions from the emitting surface [26]. Emission current density with local surface field E is given by J = (B 1 E 2 /φ) exp(-b 2 φ 3/2 / E) (1) Under the assumption that there is one major site of field emission with its emitting area denoted by Ae and with E =β(v/d) accounting for enhancement where V is the applied voltage, the total prebreakdown current I as I = JAe in logarithmic form is given by log(i/v 2 ) = (db 2 φ 3/2 /β)(i/v ) + log(aeb 1 β 2 /dφ) (2) 2.2 Field Emission Properties of CNT SWNT vs. MWNT CNTs can be manufactured in laboratory quantities by several techniques, such as arc-discharge, laser ablation and chemical vapor deposition (CVD) [20]. The structure and morphology of the CNTs fabricated by these techniques vary significantly. Especially, the defect density affects the field emission characteristics due to stable CNT heating and subsequent destruction during field emission [27]. Many experiments carried out on CNTs are valuable in providing a fundamental understanding of the emission mechanism. In the first report of electron emission from CNTs, Rinzler et al. studied field emissions from an individual multiwalled nanotube (MWNT) and reached 0.1 to 1 µa when applying a bias voltage, 80 volts, 12

26 along the field direction. They also reported that emissions could be enhanced by opening the tips of the nanotubes [28]. The CNTs can emit very large electron currents. Bonard et al. showed a single MWNT tip is capable of emitting at a stable status for more than 100 hours at 2 µa [29]. Generally, SWNTs have the capability for achieving higher current densities than MWNTs due to their smaller diameter and higher degree of structural perfection [13]. However, MWNTs exhibit a longer life span than SWNTs, according to Li et al. [30]. Furthermore, a theoretical study predicted that an open-ended SWNT has much better field-emission properties than a closed SWNT [31], due to the electronic effects that alter the bonding mode and decrease the work function. The similarities between SWNTs and MWNTs are not surprising considering the fact that the diameter of a SWNT bundle is comparable to that of a MWNT. The SWNTs, however, are more stable than the MWNTs at high current densities, which are attributed to their higher degree of structural perfection. It is important to note the exact value of the threshold field depends on how the macroscopic cathode is prepared and how measurements are performed [7]. The research groups at UNC, Bell Labs, and Duke University have demonstrated the comparable field emission properties of various CNT materials, which include SWNTs by the laser ablation, arc-discharge, and thermal CVD process; MWNTs by the CVD and arc-discharge processes; and double-walled carbon nanotubes (DWNTs). The results are summarized in Figure 2.1 which shows the emission current density versus applied electric field plotted for a variety of CNT emitters and nanostructured diamonds. The results show the emission threshold fields of these CNTs differ only by a factor of 2 3. They are all significantly lower than the reported values for other field emitters, such as the Spindt tips and diamond materials [7]. 13

27 Figure 2.1 Comparison figure of threshold voltage of various CNTs Fabrication of CNT Intuitively, vertically aligned CNTs are expected to have better emission properties than random CNTs because the electrons are emitted from the tips of the CNTs. However, the field emission properties are complicated by the effect of electrical screening. Calculations have shown that the minimum of the screening effect would be reached while the individual emitters/cnts are evenly separated so that their spacing is greater than their height. Bower et al. presented that the enhanced emission properties of most vertically aligned MWNTs fabricated by CVD process do not show as expected due to densely packed CNTs. As shown in Figure 2.1, the emission characteristic of the aligned MWNTs is significantly different from that of the random MWNT [7]. Maximal efficiency indicated for film field emitters can be reached only (a) when the emitters are well aligned and placed with their long axis perpendicular to the film substrate, and (b) when the emitters are well separated from one another to avoid screening effect from taking place [32]. Moreover, by lithographically patterning the location of the catalysts on the substrates, several groups have recently reported on aligned CNTs and carbon 14

28 nano-fibers with controlled spacing to minimize the screening effects. Nilsson et al. predicted that an intertube distance of about two times the height of CNTs optimizes the emitted current per unit [33]. For cathodes with randomly oriented CNTs, several experiments [34, 35] have shown that the CNTs can be easily bent and aligned to the electrical field direction under a moderate electrical field. Another mode to align random CNTs mixtures is taping. ITRI of Taiwan developed a handy taping process to orient the CNTs. However, the performance is limited and uncontrollable [36] Aspect ratio versus field enhancement factor Field emission measurements are often measured by attempting to quantify the behavior in terms of the applied electric field and the local electric field at the nanotip, via the field enhancement factor, β. Smith et al. showed that the field enhancement factor is fundamentally associated with electrode separation. Depending on the experimental conditions to obtain a true value for electric field, a set of alternative definitions for enhancement factor is required [37]. Cheng et al. have investigated the effect of nanotube length on field emission properties of randomly oriented SWNT films [27]. SWNT bundles synthesized by the laser ablation method were purified and chemically etched to different aspect ratios using a previously reported procedure [27, 38]. The averaged bundle length was determined by transmission electron microscopy (TEM) measurements. The experimental results showed that the emission threshold field increases with decreasing CNT bundle length. The measured threshold fields and the estimated enhancement factors are listed in Table 2.1. We attributed this to the dependence of field enhancement factor β on the average bundle length: β is reduced due to shortened CNT bundles. In summary, the larger aspect ratio, and the larger field enhancement factor indicated the lower turn-on voltage. 15

29 Table 2.1 Effect of bundle length on the field emission properties of SWNTs Emission current-voltage measurement Figure 2.2 shows the emission current voltage (I V) characteristics of the SWNT film measured using a hemispherical current collector of 1 mm diameter (anode) at torr base pressure at different anode cathode gap distances. The emission material used in this study is SWNT bundles that were produced by the laser ablation method at UNC. A uniform layer of SWNTs was coated on a flat metal disc by electrophoretic deposition. As shown in Figure 2.2 and the inset, the SWNT film shows the classic Fowler Nordheim behavior with a threshold field of 2 V/µm for 1 ma/cm 2 current density. Emission current density over 1 A/cm 2 can be readily achieved. The threshold fields are substantially lower than those reported for other electron field emissive materials [39]. 16

30 Figure 2.2 Electron emitted current versus applied voltage of a SWNT film measured at different anode cathode distances using a hemispherical anode Li et al. fabricated a thin MWNT film field emission cathode by the screen-printing method. The density of the carbon nanotubes was about 2.5 X 10 8 /cm 2, 5% CNT content. 3% and 10% CNTs were also made by the same way and examined for the field emission properties. The fabricated CNTs cathode was pretreated using increased electric field conditioning. The threshold field shown on Figure 2.3 was less than 3 V/µm and the emission current density approached 1 ma/cm 2 at 6.5 V/µm. The field emission experimental results of CNTs cathodes with different CNTs densities indicate that there is an optimal density being used as the field emission cathode. These results will enable us to produce carbon nanotube-based large area field emission displays in the future [30]. 17

31 Figure 2.3 I-V and F-N plot of diode configuration carbon nanotube film Emission stability of CNT De Heer et al. indicated the electron beam from CNT was stable for a long period of time. Figure 2.4 shows the stability is remarkable, considering the exponential dependence of the current on the applied field. Curve B was modified based on the feedback stabilization system of Curve A. During 48 hours of continuous operation at 30 µa, current drifts on the order of 5% were observed, but the average current did not degrade [15]. Figure 2.4 Emitted current as a function of time 18

32 Cheng et al. found the stability of emission performance can be readily evaluated from a 0.2 cm 2 area SWNT film by monitoring the evolution of total emission current at a fixed electric field. Figure 2.5 shows the result from a macroscopic sample measured using the parallel-plate geometry. After the initial burn-in, under a constant DC voltage without feedback, no overall decay in the emission current was observed over a 10-hour period. Recently we demonstrated stable emission at comparable current level for over 300 hours at 100% duty cycle. The standard deviation of the current fluctuation was 2 4% [27]. Figure 2.5 Stability of the emission current measured under a constant applied electric field from a 0.2 cm 2 area SWNT film 2.3 Applications: Field Emission Display and Back Lighting Since the initial report of electron field emission from CNTs, there have been growing interests on new vacuum electronic devices based on these novel emitters in both academic and industrial laboratories. One of the first CNT-based display devices reported was a cathode ray lighting element using CNT field emitters as the electron 19

33 source. The field emitted electrons were accelerated towards the phosphor screen which acted as the anode. Different colors were obtained by using different fluorescent materials. The luminance of the phosphor screens measured on the tube axis was cd/cm 2 for green light at an anode current of 200 µa, as shown in Figure 2.6, which was two times more intense than that of conventional thermionic CRT lighting elements operated under similar conditions [40]. Figure 2.6 Photos of CNT emitters illuminating three primary colors-green, red, blue-on an anode current of 200 µa and a voltage of 10kV Matrix-addressable diode-type FEDs usually consist of CNT stripes on the cathode glass plate and phosphor-coated indium-tin-oxide (ITO) stripes on the anode plate. Pixels formed at the intersection of cathode and anode stripes [8]. Later, a 4.5-inch three color display (Figure 2.7) using SWNT emitters were fabricated [8]. The fully sealed diode display had 128 addressable lines and demonstrated a brightness of 1800 cd/m 2 at 3.7 V/µm [41]. Recently, Samsung introduced a 9-inch full-color FED with lines, which is shown in Figure 2.7 (b) [42]. Compared to the commercial LCD and plasma displays, the current CNT based FEDs require further improvements in pixel uniformity and stability. 20

34 (a) (b) Figure 2.7 The color image from (a) 4.5 inch three color display (b) 9 inch full-color FED 2.4 Progress Made by Major Manufacturers Most of the major players in FPD industry are in Asia. Korea, Japan, Taiwan, and China plunged the greatest resources into display technology. Some recent milestones about CNT-FED or CNT-BLU were expounded the significance of CNT field emission technology. In August 2004, Samsung Corning Inc. began developing a surface light source based on carbon nanotubes, while Nano Pacific has teamed up with Iljin Nanotech and Samsung SDI to develop carbon nanotube-based backlight unit with a goal to commercialize it by Moreover, Samsung demonstrated the latest 38 inch CNT-FED with superior color purity in October In November 2004, the Dialight Corporation in Japan displayed a 7-inch CNT-BLU at the CEATEC Conference (Figure 2.8(a)). Korea s LG Electronics, in cooperation with Korea Advanced Institute of Science and Technology (KAIST), developed a 5-inch CNT-BLU with 10,000 nits (Figure 2.8(b)). Furthermore, LG was successful to exploit 20 inch CNT-FED and demonstrated it in IDMC in 2005, where LG declared their commercial program of CNT-FED. IRTI ERSO in Taiwan, in order to help local manufacturers to improve the backlight technology for large-area display applications, combined carbon nanotubes and field emission display technologies with low-cost thick-film printing and vacuum 21

35 device technologies to develop an environmentally friendly flat backlight source. With its high brightness, low power consumption, low surface temperature, and elimination of mercury vapor, the CNT-BLU developed by ERSO provides the industry with a new technology option for making large-area backlight modules. The technology adopts planar electronic field as the backlight emission base, resulting in even luminance on the backlight surface that is suitable for application in large-area LCDs. Using carbon nanotubes coated on the backlight plate as the electron source for field emission provides good luminescence efficiency, 3500 nits. The 20 inch CNT-BLU developed by ERSO, shown in Figure 2.8(c), offers greater improvement in both process simplification and cost reduction compared to the current technology of using CCFL backlight. The technology will be a helpful solution for the local manufacturers to enter and compete effectively in the large display market [43]. (a) (b) (c) Figure 2.8 The images of (a) 7 CNT-BLU by Dialight (b) 5 CNT-BLU by LG (c) 20 CNT-BLU & LCD by ERSO 22

36 2.5 Summary of Literature Review CNT s unique structure provides excellent emission characteristics, such as a low emission turn-on or threshold field and a high current density, which make CNT attractive as a potential cold cathode electron source. Before CNT can be utilized in practical applications, there are some issues that need to be addressed including emission uniformity, lifetime (durability), power consumption, cost, etc. Tremendous progress has been made in all these areas in the last few years. CNT-based field emission devices with enhanced performance are expected to be commercially available in the near future. 23

37 CHAPTER 3 BUCKYPAPER FILMS AND CNT PASTE FIELD EMISSION PROPERTIES TESTING METHODOLOGY AND STANDARD 3.1 Introduction Building upon de Heer s use of carbon nanotubes of field emission emitters in 1995 [15], research on different forms of carbon nanotubes such as vertically grown carbon nanotube or carbon nanotube paste, has been investigated. In this research, BuckyPaper films were proposed to be a new field emission emitter material for future advanced application. Therefore, the current-voltage (I-V) curve and durability performance were measured to characterize BuckyPaper s field emission properties. In addition, open-ended CNT paste and carbon nanosphere paste also were studied for their field emission properties in this research. In this section, an experimental apparatus, Field Emission Measurement System (FEMS), is proposed to measure the field emission properties. The CNT BuckyPaper films and CNT paste samples were made for experiments using various manufacturing processes and placed in the FEMS. Field emission properties were measured in ultra high vacuum (UHV) condition. Current density, threshold voltage and durability testing are the major criteria data that were collected in this project. The data were analyzed and compared. Figure 3.1 describes the sample categories and the criteria of this study. 24

38 Figure 3.1 Categories and criteria of BuckyPaper films and CNT pastes 3.2 Experiment Apparatus In order to measure the current and turn-on voltage of various BuckyPaper films and CNT pastes, field emission apparatus was assembled using a bell chamber, source meter, and vacuum pump as shown in Figures 3.1 and 3.2. The vacuum chamber is a transparent bell jar, which can provide vacuum level lower than 10-7 torr. It was chosen so that the samples can be placed and luminescence can be measured inside. A source meter, Kethley 2410, provided voltage to anode and cathode of the samples to form the electric field. Electrons can be forced and extracted by electric field from carbon nanotubes of BuckyPaper or CNT paste and be measured by the source meter. Another important piece of equipment, Varian Inc. turbo cart pumping system, was applied to pump the air out and create the UHV condition of chamber. Generally, the UHV condition level would maintain at 10-7 to 10-8 torr or lower at room temperature. However, 10-6 torr is the best condition that could be reached in 5 hours due to the pump limitation of the FACCT field emission apparatus. In the field 25

39 emission properties testing, the UHV condition is required to reduce noise current and get more precise current output. Figure 3.2 Schematic of field emission apparatus (a) (b) Figure 3.3 (a) Bell jar chamber; (b) Varian Inc. turbo cart pumping system 26

40 (c) Figure 3.3 (c) Kethley 2410 source meter 3.3 Experiment Criteria Based on the physics of field emission, current density, turn-on voltage, and field enhancement factor are the measurement criteria of I-V testing. Current and ramp voltage can be collected from the source meter, Keithley 2410, and drawn into an I-V curve to exhibit the field emission properties. We define the turn-on voltage as the voltage reading at the point when the current density reaches 1µA/cm 2. Moreover, the turn-on voltage will be divided into the gap distance that was defined as threshold field of emitter. Furthermore, field enhancement factor β, the ratio of barrier field and marcroscopic field, can be calculated by the equation of Fowler Nordheim (F N): I = av 2 exp( bφ3/2/βv ), where I, V, φ, β are the emission current, applied voltage, work function, and field enhancement factor, respectively. However, there are no specific criteria for repeatability, durability and stability testing of emitter. The best way to evaluate the durability testing is the slope of current output logarithm. The more horizontal slope indicated the better durability. Nevertheless, it is not the general criterion to determine the performance of emitter durability. Besides, luminescence should be one of the criteria to determine the performance of emitters. Luminescence means the radiance degree of an object or lighting source, describing the brightness of an orthogonal projection within a unit area. However, this project would not characterize the performance of emitters by their luminescence. The luminescence experiment will be carried out in the future. 27

41 3.4 Specimen Preparation BuckyPaper emitter preparation Various BuckyPaper films (magnetically aligned or randomly oriented) containing different compositions, such as SWNTs, MWNTs, and VGCNFs, were characterized in this investigation. BuckyPaper films were cut into roughly 2 cm by 2 cm pieces and placed on an indium tin oxide (ITO) coated glass substrate to act as a cathode. Several 0.15 mm thick cover glasses were placed on top of the perimeter of the BuckyPaper leaving a 1 cm by 1 cm center area of BuckyPaper uncovered for electron emission. The gap distance between anode and cathode is 300 µm for BuckyPaper samples. Next, the anode was produced by laying another ITO glass on the sample, using clips to fasten the assembly upper and lower the ITO glasses. Glass ITO Cover Glass Cover Glass Nanotube film ITO Glass Figure 3.4 Schematic of BuckyPaper field emission 28

42 (a) (b) Figure 3.5 BuckyPaper field emission trail samples CNT Paste emitter preparation The CNT paste samples were made by sprinkling CNT powder on silver adhesive. Aligned SWNT BuckyPaper is cut into shorter CNTs of two different lengths (400 nm and 50 nm) using a microtome. Silver adhesive was applied on the ITO glass followed by uniformly sprinkled CNT powder. After several minutes, the silver adhesive dried and cured. The CNTs randomly adhered to the silver adhesive, forming the CNT paste emitter cathode. The following procedures were the same as BuckyPaper preparation. However, the gap distance was 450 µm between anode and cathode for 400 nm open-ended SWNT paste, and 300 µm for 50 nm open-ended SWNT paste and carbon nanosphere paste. The schematic and picture are shown in Figures 3.6 and

43 Figure 3.6 Scheme of open-ended CNT paste field emission Figure 3.7 Open-ended CNT paste sample 3.5 Test Procedures Field-emission measurements were accomplished in a vacuum chamber. The vacuum level was set at 10-6 torr due to the limitation of vacuum pump. CNT emitter was located between upper and lower glasses, which were secured with alligator clips. Moreover, ITO glasses, connected to a source meter, were used as an anode and cathode. A positive voltage was applied to the anode, while a negative voltage was applied to the cathode. Emission current was measured with the Keithley 2410 source meter, and applied voltage was ramped up from 0 to various set values. The electric 30

44 field was estimated by dividing the applied voltage by the sample anode separation (V/d). Emission current density was calculated from the measured emission current and by assuming a 1 cm by 1 cm square emission area. After data collection, I-V and F-N curves were plotted to visualize the field emission properties. The schematic of field emission testing procedures is shown in Figure 3.8. The durability testing was carried out in the same way as I-V measurement. However, the relationship of current density versus time would be the basis to determine the performance of emitter performance. Moreover, there are three data collecting procedures to characterize the repeatability, durability and stability properties of CNT emitter. For the first experimental durability testing, a certain voltage was applied to the sample, and the current density was recorded for one and a half hours. The applied energy was turned on and off every ten minutes to observe the repeatability of BuckyPaper film emitters and CNT pastes. In the second testing procedure, a certain applied voltage, 400 V, was set for 20 hours, and the current output was collected from BuckyPaper emitters or CNT pastes. The degree of current decay showed the BuckyPaper films and CNT pastes durability. The better current density maintenance displayed the better durability properties. In the third durability testing, a feedback control system was applied into the computer for 20 hours. The feedback system adjusts the applied voltage to keep the current output at a constant level. The variation of applied voltage is the major index to check the performance of emitters in the third testing procedure. Analyses on current degeneration, voltage variation, and current repeatability were performed. 31

45 Connect cathode and anode to power supply Pull vacuum to 10-5 torr Place and secure the sample in vacuum chamber Apply charge to sample to measure the currents Collect data Figure 3.8 Schematic of field emission test procedures 32

46 CHAPTER 4 FIELD EMISSION AND DURABILITY PROPERTIES OF BUCKYPAPER FILMS 4.1 Introduction BuckyPaper films are thin sheets formed by nanotube suspension with controlled dispersion, using a special filtration method as shown in Figure 4.1. BuckyPaper films are made by passing a suspension of nanotubes through a porous membrane or filter. BuckyPaper films are a mesh of carbon nanotubes and freestanding mats peeled off of the filter membrane in filtration allowing ease of handling and processing. BuckyPaper films are considered to be good field emission emitters. This chapter describes the details of BuckyPaper field emission characteristics. First of all, the principles of field emission and emitter durability properties were addressed. Based on the principles, the basic explanations of experiments were organized as follows. First, we provided a preliminary testing for I-V of different types of BuckyPaper. The results will be calculated and compared with the conventional Fowler-Nordheim models for electron field emission. Based on these analyses, we selected several BuckyPaper films that exhibited high current output for future testing. Except for current-voltage properties, the durability, stability and repeatability of BuckyPaper films are also investigated in this study. Three different experiments described in chapter 3 were used for characterizing BuckyPaper s durability. The stable current output is the major index to evaluate the durability performance of 33

47 emitters. Moreover, the slopes of logarithmic current output were calculated so as to find the relationship between slopes and durability performance. 1cm 1cm Figure 4.1 Randomly dispersed BuckyPaper sample 4.2 Field Emission Theory and Durability Principle Electron emission is defined as liberation of free electron from a surface of a substance caused by the external energy transferred to the electrons. In order to emit electrons from the surface of substance, these free electrons require additional external energy, the escape work. The amount of outside energy in order to emit substance electrons is known as work function. Field emission is the process whereby electrons tunnel through a barrier in the presence of a high electric field. This phenomenon is highly dependent on both the properties of the material and the shape of the particular cathode, such that a higher aspect ratio produces a higher field emission current. The current density produced by a given electric field is governed by the Fowler-Nordheim equation, which was described in Chapter 3 [26]. Carbon nanotube s field emission current voltage characteristic and current stability are of major interest in this investigation among 34

48 others. Saito et al. [45] showed that the I V relation of single-walled carbon nanotubes (SWNTs) follows the Fowler Nordheim (F-N) law, but Bonard et al. and Collins and Zettl reported a deviation from the F-N behaviors. Carbon nanotube field emission properties can not consist with F-N behaviors when the applied voltage exceeds the knee point [46, 47]. In fact, the β factor derived from an experimental I V characteristic may be considered as a measure of the efficiency an emitter. If β is large, it means that an emission current can be obtained at a low applied field. However, Xu et al. explained another mechanism for the change of β. They pointed out that vacuum space charge is an important factor. Ionizing the gas existing in a gap and the emitted electron charge are the two reasons for space charge to affect β [48]. A failure of the F-N model in describing conventional field emitters is often accounts for by the incorporation of additional interactions. In Figure 4.2, this space charge sharply reduces the actual electric field at the emitter. The FN linearity criterion will fail when the space charge becomes significant because the local electric field is no longer proportional to the applied voltage. Saturation of the emitter current due to limited carrier concentration in a nonmetallic emitter can add to or exaggerate this space-charge effect [3]. Therefore, the lower voltage areas were used to calculate the β values in this project to contrast the I-V results. 35

49 Figure 4.2 Fowler-Nordheim Plot for a nanotube matrix sample 4.3 Field Emission Properties of BuckyPaper Comprehensive testing for different types of BuckyPaper For comprehensive current-voltage properties testing of BuckyPaper films, different compositions of BuckyPaper were first fabricated for characterizing. Figure 4.3 shows several categories of BuckyPaper films with different compositions. Randomly dispersed SWNT BuckyPaper (RB), aligned SWNT BuckyPaper (AB), VGCNF mixed SWNT in ratio 1:1 BuckyPaper (N61-1), VGCNF mixed SWNT in ratio 3:1 BuckyPaper (N61-2), field emission SWNT mixed regular SWNT BuckyPaper (58/61), e-beam irradiated SWNT BuckyPaper (EB), and MWNT mixed SWNT BuckyPaper (M61) were fabricated and tested. 36

50 Figure 4.3 BuckyPaper types and identification codes Current-Voltage (I-V) properties A. Randomly dispersed SWNT BuckyPaper Figure 4.4 shows the randomly dispersed SWNT BuckyPaper (RB) I-V curves. The turn-on voltage is defined as the voltage corresponding to the current density of 1 µa/cm 2. Therefore, the best turn-on voltage in SWNT61 BuckyPaper was V/µm when the current density reached 1 µa/cm 2. The plot shows the backside of most BuckyPaper samples exhibited better field emission performance. Current Density (ma) Figure 4.4 I-V Curves of randomly dispersed SWNT BuckyPaper 37

51 B. Aligned SWNT BuckyPaper Aligned SWNT BuckyPaper (AB) I-V curves are shown in Figure 4.5. The best turn-on voltage was V/µm, when the current density reached 1µA/cm 2. From the I-V curve plot, the frontside of BuckyPaper displayed better field emission, a result that goes against our original hypothesis, which is backside has good performance than frontside. Moreover, field emission properties of aligned SWNT BuckyPaper were not as good as other types of BuckyPapers. Current Density (ua) Figure 4.5 I-V Curves of aligned SWNT BuckyPaper C. VGCNF Mixed SWNT in ratio 1:1 BuckyPaper The VGCNF/SWNT in ratio 1:1 BuckyPaper (N61-1) I-V curve is shown in Figure 4.6. The best turn-on voltage was V/µm when the current density reached 1 µa/cm 2. The backside of the BuckyPaper sample performed better than frontside of the same sample. The results are consistent with the hypothesis and randomly dispersed SWNT BuckyPaper results. Most frontside samples of N61-1 BuckyPaper films failed to reach 1 µa/cm 2, even when the voltage was as high as 400 V. 38

52 Current Density (ma) Figure 4.6 I-V Curves of VGCNF/SWNT in ratio 1:1 BuckyPaper D. VGCNF Mixed SWNT in ratio 3:1 BuckyPaper The VGCNF/SWNT in ratio 3:1 BuckyPaper (N61-2) I-V curves are shown in Figure 4.7. The best turn-on voltage was 0.8 V/µm when the current density reached 1 µa/cm 2. The backside of the BuckyPaper sample performed better than frontside of the same sample. However, N61-2 BuckyPaper films displayed an unstable I-V curve after 400 V. Current Density (ma) 0.8 Figure 4.7 I-V Curves of VGCNF/SWNT in ratio 3:1 BuckyPaper 39

53 E. Field Emission SWNT Mixed Regular SWNT BuckyPaper The field emission SWNT (58) mixed regular SWNT (61) in ratio 1:1, BuckyPaper (58/61) I-V curve is shown in Figure 4.8. The best turn-on voltage of 58/61 BuckyPaper was V/µm when the current density reached 1µA/cm 2. Both sides of 58/61 BuckyPaper performed equally well. The result was not consistent with the hypothesis. Current Density (ma) Figure 4.8 I-V Curves of 58/61 in ratio 1:1 BuckyPaper F. E-beam Irradiated SWNT BuckyPaper The I-V curves of e-beam irradiated SWNT BuckyPapers (EB) are shown in Figure 4.9. The best turn-on voltage was V/µm when the current density reached 1uA/cm 2. From the I-V curve plot, the backside of BuckyPaper film displayed better field emission properties than the frontside and was consistent with the hypothesis. However, e-beam irradiated BuckyPaper did not perform as well as other types of BuckyPaper films. 40

54 0.937 Figure 4.9 I-V Curves of e-beam irradiated SWNT BuckyPaper G. MWNT Mixed SWNT in Ratio 1:1 BuckyPaper The MWNT mixed SWNT in ratio 1:1 BuckyPaper (M61) I-V curve is shown in Figure The best turn-on voltage was V/µm when the current density reached 1µA/cm 2. From the I-V curve plot, the backside of BuckyPaper failed to reach 1µA/cm 2. Moreover, the M61 sample did not perform as well as SWNT BuckyPaper films. Current Density (ua) Figure 4.10 I-V Curves of e-beam irradiated SWNT BuckyPaper 41

55 H. Compendium Table 4.1 shows the lowest turn-on voltage was VGCNF/SWNT in ratio 1:1 BuckyPaper, V/µm. The ranking of turn-on voltage from lowest to highest are 1) aligned SWNT, 2) e-beam irradiated SWNT, 3) MWNT/SWNT in mixed ratio 1:1, 4) VGCNF/SWNT in mixed ratio 3:1, 5) randomly dispersed SWNT, 6) SWNT/SWNT in mixed ratio 1:1, and 7) VGCNF/SWNT in mixed ratio 1:1. Figure 4.10 shows the I-V comparison plot for all types of BuckyPaper films. The next section explains their Fowler-Nordheim behaviors. Table 4.1 Turn-on voltage comparison of all BuckyPaper categories Sample Code Sample Description Lowest Turn-on Voltage (V/µm) SWNT 61 Randomly dispersed SWNT 61 BuckyPaper AB Aligned SWNT 61 BuckyPaper N61-1 VGCNF 19/SWNT 61 BuckyPaper: mix ratio 1: N61-2 VGCNF19/SWNT 61 BuckyPaper: mix ratio 3: /61 SWNT 58/SWNT 61 BuckyPaper: mix ratio 1: EB E-beam irradiated SWNT BuckyPaper M61 MWNT 17/SWNT 61 BuckyPaper: mix ratio 1: Figuire 4.11 The I-V curves of all BuckyPaper types 42

56 Fowler-Nordheim (F-N) properties The theory of electron field emission was formally published in 1928 by Fowler and Nordheim. The Fowler-Nordheim model for field emission describes the electron current density emitting from a surface into vacuum as a function of applied field. A material is said to demonstrate the field emission phenomenon when there is a straight line in its F-N (or log (I/V 2 )-1/V) plot. The complete Fowler-Nordheim equation is shown below: log( I / V ) = log 6 2 Aeβ 10 2 φd 4.52φ β 9 dφ V The slope of log (I/V 2 ) and 1/V indicates the geometric enhancement factor, β, and determines the field emission performance of each sample. A higher β indicates better field emission properties. The model parameters are current density I, applies voltage V, the area for emission Ae, gap distance d, and work functionφ, respectively. Some BuckyPapers in comprehensive testing were selected to perform their F-N plots. As previously stated, the slope of the lower voltage area in F-N plot was calculated for the β value. Therefore, the slope of the higher area than knee point voltage was eliminated. Moreover, the emitted current density is already very high, so that the emitted electron cloud near the emitters naturally forms a space charge that can limit the emission process. Consequently, the high current density area was eliminated. The F-N plots that followed the principle of knee point and space charge theory were used to calculate the β values. A. Randomly Dispersed SWNT BuckyPaper The current density and applied voltage of randomly dispersed SWNT BuckyPaper were calculated to plot the F-N curve slope of , shown in Figure

57 ln(i/v2) y= x /V Figure 4.12 F-N curve of randomly dispersed SWNT BuckyPaper B. VGCNF/SWNT in Ratio 1:1 BuckyPaper The current density and applied voltage of N61-1 BuckyPaper were calculated to plot the F-N curve with a slope of , shown in Figure The absolute value of N61-1 slope was lower than randomly dispersed SWNT BuckyPaper film ln(i/v2) y= x /E Figure 4.13 F-N curve of N61-1 BuckyPaper 44

58 C. VGCNF/SWNT in Ratio 3:1 BuckyPaper The current density and applied voltage of N61-2 BuckyPaper were calculated to plot the F-N curve with a slope of , shown in Figure The absolute value of N61-2 slope was higher than randomly dispersed SWNT and N61-1 BuckyPaper films ln(i/v2) y= x /V Figure 4.14 F-N curve of N61-2 BuckyPaper D. MWNT/SWNT in Ratio 1:1 BuckyPaper The current density and applied voltage of M61 BuckyPaper were calculated to plot the F-N curve with a slope of , shown in Figure The absolute value of M61 slope was higher than the absolute value of randomly dispersed SWNT and N61-1 BuckyPaper films. 45

59 Y Axis Title y= x X Axis Title Figure 4.15 F-N curve of M61 BuckyPaper E. Compendium From these F-N curves, Table 4.2 shows the β values could be calculated from their slopes. If the β values are sorted in order from the highest to the lowest, it would be: 1) randomly dispersed SWNT, 2) M61, 3) N61-1, and 4) N61-2. Common values for geometric enhancement factor of SWNT was 100< β < 1000 [1]. However, the reasons of various β values of BuckyPaper might include, but are not limited to, the structure of BuckyPaper, the composition of BuckyPaper, and the surface condition of BuckyPaper. Table 4.2 The β values of four selected BuckyPaper types Sample Code Lowest Turn-on The slope of F-N β value Voltage (V/µm) plot SWNT N N M

60 Summary Based on the comprehensive BuckyPaper field emission testing, the best turn-on voltage of various BuckyPaper samples was VGCNF/SWNT in ratio 1:1, 0.623V/µm. In most cases, the back sides performed better than the front, which was consistent with our hypothesis. Field emission performance of e-beam treated and aligned BuckyPaper was not as good as others. Due to their unstable performance, durability testing was discontinued. The best four BuckyPaper samples appeared to follow the field emission behavior described in the Fowler-Nordheim theory. VGCNF/SWNT in ratio 1:1 BuckyPaper displayed the lowest turn-on voltage, but randomly dispersed SWNT BuckyPaper showed the highest β value of the four selected BuckyPaper films. The BuckyPaper data were consistent with field emission behavior Selected testing for high current output BuckyPaper This study also focused on the current amount for emission area. Three samples, randomly dispersed SWNT BuckyPaper, VGCNF/SWNT 1:1 BuckyPpaer, and SWNT58/61 1:1 BuckyPaper were selected to test for their higher current density by applying more power to observe high current output to broaden the applications Current-Voltage (I-V) properties A. Randomly dispersed SWNT BuckyPaper The applied voltage was increased from 400 V to 600 V or 700 V to observe the higher current output. For randomly dispersed SWNT BuckyPaper, the turn-on voltage for backside was 0.63 V/µm, and the frontside was V/µm. When applied voltage reached 700 V, the highest current output of backside was ma/cm 2. When the applied voltage reached 600V, the highest current output of frontside was ma/cm 2. Moreover, current output of the frontside dropped after applied voltage larger than 600 V. 47

61 Current Density (ma) Figure 4.16 High current output I-V curve of randomly dispersed BuckyPaper B. VGCNF/SWNT in Ratio 1:1 BuckyPaper For N61-1 BuckyPaper in Figure 4.17, the turn-on voltage for the backside was 0.5 V/µm, and the frontside was 0.71 V/µm. The highest current output of backside was ma/cm 2, and the frontside was ma/cm 2. Current output of N61-1 BuckyPaper was much lower than randomly dispersed SWNT BuckyPaper. Carbon Nanofiber is assumed to play a significant role in current degeneration. Current Density (ma) Figure 4.17 High current output I-V curve of N61-1 BuckyPaper 48

62 C. SWNT58/SWNT61 in Ratio 1:1 BuckyPaper For 58/61 BuckyPaper in Figure 4.18, the turn-on voltage for the backside was 1.79 V/µm, and the frontside was 0.65 V/µm. The highest current output of the backside was 0.026mA/cm 2, and the frontside was ma/cm 2. Current output of 58/61 BuckyPaper was much lower than the previous BuckyPaper films. Moreover, the current output proved unstable when applied voltage was increased. Field emission SWNT did not perform as well as we expected. Current Density (ma) Figure 4.18 High current output I-V curve of 58/61 BuckyPaper Summary The experiment was only performed on SWNT-61, N61-1 and SWNT 58/61 based on turn-on voltage and stability performance in the basic I-V experiments. The highest current performance was SWNT-61 BuckyPaper (61-B: ma). N61-1 failed to reach currents as high as SWNT-61 (N61-1-F: ma). SWNT58/61 showed unstable I-V measurements when applied voltage exceeded 500 V. Unfortunately, the current output did not perform as well as expected. Many issues of BuckyPaper films must be studied in the future to improve the low current output of BuckyPaper, such as tube aligned direction, surfactant effects, and screening effects. 49

63 4.4 Durability and Repeatability Properties of BuckyPaper The stability, durability and repeatability properties of BuckyPaper could be tested in several ways. The repeatability of BuckyPaper can be appraised by observing the current performance when applied power is frequently turned on and off. The stability and durability of emission performance can be readily evaluated by monitoring the variation of total emission current at a fixed electric field and monitoring the variation of applied voltage at a fixed current output by using a feedback control system. Therefore, three experiments were designed to test repeatability, stability, and durability Repeatability of BuckyPaper for 1.5 hours This experiment was designed to examine the repeatability of BuckyPaper. For one and a half hours, the constant applied power was turned on and off every 10 minutes to observe the current output performance. A. Randomly Dispersed SWNT BuckyPaper A fixed DC voltage, 450 V, was applied to observe the current output of the randomly dispersed SWNT BuckyPaper sample. The power was turned off and on every 10 minutes to test the repeatability of BuckyPaper. The performance of randomly dispersed SWNT BuckyPaper was stably repeatable even when intermittent power was provided, as shown in Figure Current Density (ua) Figure 4.19 Repeatability of randomly dispersed SWNT BuckyPaper 50

64 B. VGCNF/SWNT in Ratio 1:1 BuckyPaper A fixed DC voltage, 450 V, was applied to observe the current output of N61-1 BuckyPaper sample. The repeatability of N61-1 BuckyPaper apparently performed well on both sides even when intermittent power was provided. However, Figure 4.20 shows the backside of N61-1 BuckyPaper produced a larger current than the frontside. Current Density (ua) Figure 4.20 Repeatability of N61-1 BuckyPaper C. SWNT58/SWNT61 in Ratio 1:1 BuckyPaper A fixed DC voltage, 400 V, was applied to observe the current output of 58/61 BuckyPaper sample. The repeatability of N61-1 BuckyPaper apparently performed well on both sides even when intermittent power was provided. However, Figure 4.21 shows the backside of N61-1 BuckyPaper dropped significantly during the first 10 minutes. The backside of BuckyPaper still displayed better current output than the frontside. 51

65 Current Density (ua) Figure 4.21 Repeatability of 58/61 BuckyPaper D. Aligned SWNT BuckyPaper A fixed DC voltage, 450 V, was applied to observe the current output of the aligned SWNT BuckyPaper sample. The repeatability of aligned SWNT BuckyPaper apparently did not perform as well as other samples on the backside when intermittent power was provided as shown in Figure Moreover, the frontside of aligned SWNT BuckyPaper showed very low current density output. The field emission and durability performance of aligned SWNT BuckyPaper were not as good as other BuckyPaper types. Current Density (ua) Figure 4.22 Repeatability of aligned SWNT BuckyPaper 52

66 E. MWNT/SWNT in Ratio 1:1 BuckyPaper Two fixed DC voltages, 450 V and 500 V, were applied to the back side and front side of M61 BuckyPaper to observe the current output. The front side of M61 BuckyPaper showed unstable current output compared to aligned BuckyPaper. Moreover, the backside of M61 BuckyPaper decreased quickly during the first 20 minutes. Due to those unstable and low current output curves, we concluded that M61 would not perform as well as other BuckyPaper films. Current Density (ua) (a) Current Density (ua) (b) Figure 4.23 Repeatability of M61 BuckyPaper (a) Backside; (b) Frontside F. Summary The purpose of this experiment was to test current durability and repeatability of various BuckyPapers. Stability properties performance after frequent turning on and off was observed. SWNT 61, SWNT58/61 and VGCNF/SWNT (1:1) BuckyPaper displayed better performance stability and durability in this experiment. The backsides of most BuckyPaper samples showed better current durability and larger current 53

67 density output than the frontsides. However, e-beam irradiated BuckyPaper samples were not included in this test due to unstable performance and lack of samples as shown in previous experiments Durability of BuckyPaper for 20 hours The stability and durability of emission performance can be readily evaluated by monitoring the variation of emission current at a fixed electric field. This experiment was designed to examine the durability of BuckyPaper over 20 hours. The stability of emission current is an important indicator to propose the appropriate application for BuckyPaper. A. VGCNF/SWNT in Ratio 1:1 BuckyPaper In durability experiments, 400 V was applied to N61-1 BuckyPaper for 20 hours. Figure 4.24 shows the backside of the N61-1 demonstrated better current density than frontside after 20 hours. Current performance of frontside of N61 dropped to a very low level after 20 hours. After taking logarithm of current density and plot with time, the slope of the frontside of BuckyPaper, 1.415, was larger than the backside of BuckyPaper, (a) (b) Figure 4.24 Durability of N61-1 BuckyPaper: (a) Back side of sample; (b) Front side of sample 54

68 (c) (d) Figure 4.24 Durability of N61-1 BuckyPaper: (c) Logarithm of frontside sample; (d) Logarithm of backside sample B. VGCNF/SWNT in Ratio 3:1 BuckyPaper Again, 400 V was applied to N61-2 BuckyPaper for 20 hours. Figure 4.25 shows both sides of N61-2 performed equally well after 20 hours. Current performance of both sides of N61-2 dropped to a very low level in 10 hours. After calculating the logarithm of current density and plotting with time, the slope of the frontside of BuckyPaper, , was larger than the backside of BuckyPaper, (a) (b) (c) (d) Figure 4.25 Durability of N61-2 BuckyPaper: (a) Backside of sample; (b) Frontside of sample; (c) Logarithm of frontside sample; (d) Logarithm of backside sample 55

69 C. SWNT58/SWNT61 BuckyPaper As in previous experiments, 400 V was applied to 58/61 BuckyPaper for 20 hours. Figure 4.26 shows the backside of 58/61 BuckyPaper dropped 35 µa to 5 µa after 20 hours. The frontside of sample maintained low current density output around 1.5 µa for the entire duration. After calculating logarithm of current density and plot with time, the slope of the backside, , was larger than the frontside, (a) (b) (c) (d) Figure 4.26 Durability of 58/61BuckyPaper: (a) Backside of sample; (b) Frontside of sample; (c) Logarithm of frontside sample; (d) Logarithm of backside sample D. MWNT/SWNT in Ratio 1:1 BuckyPaper Again, 400 V was applied into M61 BuckyPaper for 20 hours. Figure 4.27 shows the current densities on both sides of M61 BuckyPaper dropped significantly after 20 hours. After calculating logarithm of current density and plot with time, the slope of the backside, , was larger than the frontside,

70 Figure 4.27 Durability of M61 BuckyPaper: (a) Backside of sample; (b) Frontside of sample; (c) Logarithm of frontside sample; (d) Logarithm of backside sample E. Aligned SWNT BuckyPaper 400 V was applied to aligned SWNT BuckyPaper for 20 hours. Figure 4.28 shows only the backside of aligned SWNT BuckyPaper. The current dropped significantly in the first hour. The logarithm of current density plotted against time showed a slope of for the backside of the BuckyPaper. (a) (b) Figure 4.28 Durability of aligned SWNT BuckyPaper: (a) Backside; (b) Frontside 57

71 F. Summary A fixed voltage, 400 V, was applied to various BuckyPaper samples to test current durability. Most BuckyPaper current measurements dropped quickly in the first few hours in the durability testing and stayed low for the remainder of the test. The slopes of the logarithm of current densities may indicate the BuckyPaper durability performance. From these data, the slopes that were horizontal showed better durability Durability of BuckyPaper for 20 hours with feedback control The durability of emission performance was also observed by adjusting the electric the field to maintain a fixed emission current. The fixable current densities were set by different BuckyPaper films I-V properties. Therefore, randomly dispersed SWNT BuckyPaper was set for 10 µa, VGCNF/SWNT BuckyPaper was set for 100 µa, and MWNT/SWNT BuckyPaper was set for 10 µa. The variation of applied voltage is an important indicator of BuckyPaper field emission properties. A. Randomly dispersed SWNT BuckyPaper Feedback controlled voltage was applied to maintain a current density of 10 µa. Figure 4.29 shows the increased applied voltage was observed from 420 V to 550 V after 20 hours. 58

72 Figure 4.29 Durability of randomly dispersed SWNT BuckyPaper with feedback control B. VGCNF/SWNT in Ratio 1:1 BuckyPaper Feedback regulated voltage was applied to maintain a current density of 100 µa, higher than randomly dispersed SWNT BuckyPaper. Figure 4.30 shows the applied voltage decreased from 516 V to 346 V after 20 hours. The result showed better durability of N61-1 BuckyPaper than randomly dispersed SWNT BuckyPaper. Figure 4.30 Durability of N61-1 BuckyPaper with feedback control 59

73 C. MWNT/SWNT in Ratio 1:1 BuckyPaper Feedback control adjusted voltage was applied to maintain a current density of 10 µa. Figure 4.31 shows the voltage decreased from 607 V to 410V after 20 hours. Figure 4.31 Durability of M61 BuckyPaper with feedback control D. Summary The feedback system was developed to keep a constant current constant by adjusting applied voltage. Testing duration was set at 20 hours. VGCNF/SWNT BuckyPaper reached a higher average current and lower power consumption with feedback control. This experiment showed the performance of BuckyPaper in practical applications. 4.5 Summary of Observations From the field emission and durability experiments, we found that VGCNF/SWNT in ratio 1:1 BuckyPaper demonstrated the lowest turn-on voltage, V/um, but randomly dispersed SWNT BuckyPaper showed the largest β value, Certain BuckyPaper samples, such as randomly dispersed SWNT BuckyPaper, VGCNF/SWNT in ratio 1:1 BuckyPaper, SWNT58/61 in ratio 1:1 BuckyPaper, 60

74 exhibited good current repeatability after frequent turning on and off. In the 20 hour durability test, most current measurements dropped in the first four hours. VGCNF/SWNT in ratio 1:1 BuckyPaper gave the best durability performance in high current with low power consumption with feedback control. E-beam irradiated SWNT BuckyPaper and aligned BuckyPaper did not show as good I-V performance as other samples. 61

75 CHAPTER 5 FIELD EMISSION AND DURABILITY PROPERTIES OF CNT PASTE 5.1 Introduction Emitters described in this chapter are different from the previous chapter, as they were made from open-ended SWNTs. The SWNTs were chopped with a microtome and made into CNT paste. Research by Choi et al. demonstrated a prototype field emission display using screen printed CNT, which speeded up the development of CNT electron field emitters [41]. Yuan et al. investigated the field emission properties of open-ended SWNT. It is found that the array emitters exhibited excellent field emission properties with a threshold electric field of 2.8 V/ µm and an emission current density of 0.08 ma/cm 2 at 3.6 V/ µm [49]. A special mechanical cutting method using a microtome was developed by FACCT researchers to cut CNTs with controlled lengths. Two different lengths of open-ended SWNT, 400 nm and 50 nm, were used in field emission and durability tests. The open-ended SWNTs were sprinkled on a conventional silver epoxy to make SWNT paste emitters. In addition to chopped CNTs, this section includes the use of carbon nanosphere in making paste. The measurement procedures were as the same as those in BuckyPaper testing. 5.2 Field Emission Properties of Nanoparticle Paste Current-Voltage (I-V) properties testing 400 nm and 50 nm open-ended SWNTs and carbon nanosphere were investigated in current-voltage testing. Figure 5.1 shows several types of CNT pastes and carbon 62

76 nanosphere paste. Open-ended SWNTs were chopped using a microtome. An aligned SWNT BuckyPaper film was made and chopped to open-ended SWNT powder in two different lengths, 400 nm and 50 nm. Next, silver adhesive was applied on the ITO glass and sprinkled with chopped tubes on top before the adhesive cured. The tubes adhered to the silver adhesive and their pinnacles randomly appeared throughout the adhesive and contributed to emission. The gap distances were set to 450 µm for 400 nm open-ended SWNT paste, and 300 µm for 50 nm open-ended SWNTs and carbon nanosphere paste. Figure 5.1 Nanoparticle paste types and identification codes A. 400 nm open-ended SWNTs paste Figure 5.2 shows a 400 nm open-ended SWNT paste I-V curve. The turn-on voltage is defined as the voltage reading when the current density reached 1 µa/cm 2, the same as in the field emission testing of BuckyPaper films. In this case, the best turn-on voltage of a 400 nm open-ended SWNT Paste was V/µm when current density reached 1 µa/cm 2. 63

77 0.313 Figure 5.2 I-V Curve of 400 nm open-ended SWNT paste B. 50 nm open-ended SWNTs paste Figure 5.3 shows a 50 nm open-ended SWNT paste I-V curve. The lowest turn-on voltage of the 50 nm open-ended SWNT paste was 0.47 V/µm when the current density reached 1 µa/cm nm open-ended SWNT demonstrated better I-V performance than 50 nm Figure 5.3 I-V Curve of 50 nm open-ended SWNT paste 64

78 C. Carbon nanosphere paste Figure 5.4 shows a carbon nanosphere paste I-V curve. The best turn-on voltage of carbon nanosphere Paste was 1.37 V/µm when the current density reached 1 µa/cm 2, much higher than those of open-ended SWNT pastes. Furthermore, the current density was significantly lower than open-ended SWNT pastes Figure 5.4 I-V Curve of carbon nanosphere paste D. Compendium As shown in Table 5.1, the lowest turn-on voltage for 400 nm open-ended SWNT paste is V/µm. The ranking of turn-on voltage from the lowest to the highest is 400 nm open-ended of SWNT paste, 50 nm open-ended of SWNT paste, and carbon nanosphere paste. In the next section, their Fowler-Nordheim behaviors will be investigated and compared to I-V properties. 65

79 Sample Code Table 5.1 Turn-on voltage comparison of three pastes Sample Description (Length) Lowest Turn-on Voltage (V/µm) 400 nm 400 nm SWNT of Paste nm 50 nm SWNT of Paste 0.47 Sphere Carbon Nanosphere Paste Fowler-Nordheim (F-N) Properties A. 400 nm open-ended SWNTs paste Figure 5.5 shows the current density and applied voltage for the 400 nm open-ended SWNT paste. They were used to calculate and plot the F-N curve; the slope was determined to be ln(i/v2) -6-8 y= x /V Figure 5.5 F-N curve of 400 nm open-ended SWNT paste 66

80 B. 50 nm open-ended SWNT paste The current density and applied voltage for 50 nm open-ended SWNT paste were used in calculating and plotting the F-N curve. The slope is , as shown in Figure Y Axis Title X Axis Title Figure 5.6 F-N curve of 50 nm open-ended SWNT paste C. Carbon nanosphere paste Figure 5.7 shows the current density and applied voltage of carbon nanosphere paste. The slope is found to be

81 -4-6 ln(i/v2) y= x /V Figure 5.7 F-N curve of carbon nanosphere paste D. Compendium From Table 5.2, the β value can be calculated by using their slopes in the previous figures. If the β values were listed from the highest to the lowest, it would be 400 nm open-ended SWNT paste, 50 nm open-ended SWNT paste, and carbon nanosphere paste, respectively. The β values of most paste emitters are larger than those of BuckyPaper emitters. The results are consistent with the summary of paste emitters which had lower turn-on voltage than BuckyPaper emitters. Sample Code Table 5.2 The β values of various CNT paste categories Lowest Turn-on Voltage (V/um) The slopee of F-N plot β value 400 nm nm Sphere

82 5.2.3 Summary The experiments showed that open-ended SWNT paste to have better field emission properties than BuckyPaper emitters and open-ended SWNT paste emitters had a lower turn on voltage and higher current density. Open-ended SWNT tubes were randomly dispersed on silver adhesive and displayed in random directions. Therefore, some tubes had the ability to contribute additional current since there were more vertically directed tubes randomly dispersed on the paste. Moreover, open-ended SWNT tubes had more defects on the tube structure. These defects usually have electrons congregate around them, which might be a contributing factor in the lower turn-on voltage of open-ended SWNTs. The 400 nm open-ended SWNT paste seemed to be more consistent with Fowler-Nordheim behavior than the 50 nm open-ended SWNT paste and carbon nanosphere paste, based on its F-N curve approaching a straight line. Moreover, β values of open-ended SWNT paste were larger than those of BuckyPaper emitters, but the β values of carbon nanosphere paste did not prove to be as large as open-ended SWNT pastes or BuckyPaper emitters. 5.3 Durability and Repeatability Properties of Nanoparticle Paste The stability, durability and repeatability properties of CNT paste were tested in the same way as in BuckyPaper testing. The experiment followed the procedures of monitoring the variation of total emission current by turning on and off every ten minutes for 1.5 hours, and also by fixing the electric field with a feedback control system. Therefore, three experiments were designed to test the repeatability, stability, and durability of these pastes Repeatability of Paste Emitters for 1.5 hours The purpose of this experiment was to examine the repeatability of paste emitter. Duration is one and a half hours, during which a constant applied power was turned 69

83 on and off every ten minutes in order to observe the current output performance. A. 400 nm open-ended CNT paste A fixed DC voltage of 400 V was applied to observe the current output of open-ended SWNT paste. The repeatability of open-ended 400 nm SWNT paste had a stable performance even when we provided intermittence power as shown in Figure 5.8. Figure 5.8 Repeatability for 400 nm open-ended SWNT paste B. 50 nm open-ended CNT in paste A fixed DC voltage of 600 V was applied to observe the current output of 50 nm open-ended SWNT paste. The repeatability of open-ended 50 nm SWNT paste had an unstable performance when we applied intermittent power as shown in Figure

84 Figure 5.9 Repeatability for 50 nm open-ended SWNT paste C. Carbon nanosphere paste A fixed DC voltage of 600 V was applied to observe the current output of nanosphere paste. The repeatability of carbon nanosphere paste was not good when intermittent power was applied. The current reading dropped to 0.1 µa after one and half hours, as shown in Figure Figure 5.10 Repeatability for carbon nanosphere paste 71

85 5.3.2 Durability of paste emitters for 20 hours Stability and durability of emission performance can be readily evaluated by monitoring the evolution of emission current at a fixed electric field. The purpose of this experiment was to examine the durability of open-ended SWNT paste in 20 hours. Furthermore, the current output was taken as a logarithm function and the relationship was determined between slope and stability. A. 400 nm open-ended CNT paste In the durability experiment, 400 V was applied to an open-ended SWNT paste for 20 hours. From Figure 5.11, the current performance of 400 nm open-ended SWNT paste dropped to a very low level after 20 hours. After taking a log of current density and plotting it against time, the slope of 400 nm open-ended SWNT paste is (a) (b) Figure 5.11 Durability for 400 nm open-ended SWNT paste (a) Durability of sample; (b) Logarithm current of sample 72

86 B. 50 nm open-ended CNT in paste In this durability experiment, 400 V was applied to open-ended SWNT paste for 20 hours. Current performance of 50 nm open-ended SWNT paste dropped in first few hours as showed in Figure After taking a log of current density and plotting against time, the slope of 50 nm open-ended SWNT paste is determined to be The portion of continued decreasing current of 50 nm open-ended SWNT paste was not as large as 400 nm paste, but the initial current density of 50 nm paste was much lower than that of 400 nm paste. (a) (b) Figure 5.12 Durability for 50 nm open-ended SWNT paste (a) Durability of sample; (b) Logarithm current of sample 73

87 C. Carbon nanosphere paste In the durability experiment, 600 V was applied to carbon nanosphere paste for 20 hours. From Figure 5.13, the current performance of carbon nanosphere paste dropped in first few hours. After taking a log of current density and plotting against time, the slope of carbon nanosphere paste is found to be The portion of continued decreasing current of carbon nanosphere paste was not as large as open-ended SWNT paste, but the initial current density of carbon nanosphere paste was much lower than those of open-ended SWNT pastes. (a) (b) Figure 5.13 Durability for Carbon Nanosphere paste (a) Durability of sample; (b) Logarithm current of sample 74

88 5.3.3 Durability of paste emitters for 20 hours with feedback control A. 400 nm open-ended CNT paste Feedback controlled voltage was applied to maintain the current density at 100 µa. Figure 5.14 shows the applied voltage increased from 274 V to 1002 V after 13 hours. Because the maximum applied voltage of 1000 V was set to protect the sample and equipment, the source meter stopped automatically when applied voltage exceeded The result showed the durability properties of open-ended SWNT paste did not perform as well as BuckyPaper films. Figure 5.14 Durability for 400 nm of open-ended SWNT paste with feedback control B. 50 nm open-ended CNT paste Feedback control adjusted voltage was applied to maintain the current density at 10 µa because of the 50 nm of open-ended SWNT paste I-V performance. Figure 5.15 displays the applied voltage increased from 191 V to 1021 V after 16 hours. Compared to 400 nm paste, the fixed current density of 50 nm paste was lower and the duration of durability was longer. However, the duration of 50 nm paste still could not reach 20 hours as the durability experiment of BuckyPaper films. 75

89 Figure 5.15 Durability for 50 nm open-ended SWNT paste with feedback control C. Carbon nanosphere paste Feedback control adjusted voltage was applied to maintain the current density at 10 µa because of the carbon nanosphere paste I-V performance. Figure 5.16 shows the applied voltage increased from 350 V to 1000 V after 10 hours. The fixed current density of carbon nanosphere paste is the same as 50 nm open-ended SWNT paste, but the time duration of durability is shorter than that of 50 nm paste. Figure 5.16 Durability of carbon nanosphere paste with feedback control 76

90 5.4 Summary of Observations From the field emission and durability experiments, we found 400 nm paste demonstrated the lowest turn-on voltage, V/um. 50 nm paste field emission properties did not present as well as 400 nm paste. In addition, chopped SWNT pastes showed better field emission performance than carbon nanosphere paste. However, durability of nanoparticle pastes did not perform as well as BuckyPaper samples. Most repeatability and durability experiments showed unstable current output. Moreover, the durability testing with feedback system for CNT paste and carbon nanosphere paste exceeded 1000V before 20 hours. The stability, repeatability and durability performance of BuckyPaper films are considered better than nanoparticle pastes. 77

91 CHAPTER 6 DATA ANALYSIS AND COMPARISON Field emission properties of carbon nanotube emitters can be fabricated in different methods. One example is the patterned nanotube emitters in which CNTs are grown vertically on a substrate. Another method is screen printing CNT paste on a substrate. The paste is a mixture of CNT, conductive powers and chemicals. Our method involves the use of a structured CNT mesh, or BuckyPaper film. In this chapter, the field emission properties of various materials made in these three methods will be compared and characterized. 6.1 Vertically Grown CNT on a Substrate Patterned nanotube emitters are grown by chemical vapor deposition (CVD). Zhu et al. were able to obtain an emission current density from 5 ma/cm 2 to 100 ma/cm 2 at the electric fields range of 2-8 V/µm [50]. Chhowalla et al. investigated the electron emissions from vertically aligned CNT of varying lengths, diameters and area densities. It was found that less densely populated short and stubby nanotubes with diameters of 400 nm and height of 0.7 µm showed the best emission characteristics with a threshold voltage of 2 V/µm and saturation emission current density of 10 ma/cm 2, as shown in Figure 6.1 [51]. 78

92 (a) (b) Figure 6.1 (a) I-V curves of vertically grown CNT emitters; (b) SEM image of the CNT emitters Choi et al. investigated crystallization and defects found in CNTs grown on an NH 3 pre-treated substrate. Figure 6.2 shows the I-V curves of CNTs grown in different conditions. The lowest turn-on voltage is 4.6 V/µm when the cuurent reached 10 µa/cm 2 [54]. CNTs grown at low temperature on a sodalime glass was achieved by a thermal CVD of acetylene gas. In Figure 6.3, the turn-on voltage was 1.3 V/µm at 1 µa/cm 2. Electron emission sufficient for flat panel displays was observed in 1 ma/cm 2 at 5 V/µm. (a) (b) Figure 6.2 (a) I-V curves of CNTs vertically grown in different conditions; (b) SEM image of the CNT emitters 79

93 (a) (b) Figure 6.3 (a) I-V curves of vertically grown CNTs using thermal CVD: (b) SEM image of the CNT emitters Xintek Inc. recently fabricated another type of commercial CNT emitters, as shown in Figure 6.4. Research by Yue et al. produced and purified SWNT emitters by a laser ablation method. A uniform layer of SWNTs was coated on a flat metal substrate and their field emission properties were measured. A total emission current of 28 ma was obtained from 0.2 cm 2 area of a CNT cathode. The threshold field for 1 ma/cm 2 is 1.3 V/µm [54]. In addition, Taiwan ITRI ERSO grew CNT emitters via CVD with the turn-on voltage being 3.8V/µm and current density 4.5 ma/cm 2 at 5.1 V/µm. Comparison of CVD grown CNT emitters is shown in Table 6.1. (a) (b) Figure 6.4 (a) I-V curves of vertically grown CNTs using laser ablation; (b) SEM image of the CNT emitters 80

94 Table 6.1 I-V comparison of different CNT growth methods CNT Emitter Growth method Current Density Turn-on Voltage Catalyzed CVD 10 ma/cm 2 2 V/µm NH 3 pre-treated substrate CVD 10 µa/cm V/µm Thermal CVD of Acetylene Gas 1 µa/cm V/µm Laser Ablation (Xintek Inc.) 1 ma/cm V/µm ITRI ERSO CVD 4.5 ma/cm 2 5.1V/µm 6.2 CNT Pastes Since the technology for growing CNTs on a substrate is not scaled up at this time, many research groups turned to other CNT emitter fabrication methods. Screen-printing of CNT paste is considered to be a scalable and cost-effective process. ITRI ERSO CNT-FED emitter is a significant investigation in Asia [4]. CNT powders were mixed with alpha-terinol, carbiol, ethylcellulose, ethanol, and silver epoxy to fabricate the CNT paste. The CNT emitters turned on at a field of 2.3 V/µm and the current density was 5.5 ma/cm 2 at 3.5 V/µm (10 ma/cm 2 at 4 V/µm). They used CNT from different vendors. Figure 6.5 shows the comparison of I-V curves of different CNT sources. Figure 6.5 shows the Xintek CNT paste made by ERSO when the turn on voltage is around 2.33 V/µm. Tsai et al. proposed an electron field emitter based on low temperature cofirable ceramics (LTCC). The LTCC materials were a mixture of glass and Al 2 O 3. CNT (10 wt.%) was mixed with conventional Ag paste and then screen printed on a glass substrate. The packaging technique using LTCC was adopted for fabricating screen-printed electron field emitters. As seen in Figure 6.6, the CNT paste turn on voltage is 0.9 V/µm, and retains 40 ma/cm 2 at 2.2 V/µm[4]. In our research, the open-ended CNT pastes in various lengths of CNT were mixed with conventional Ag 81

95 epoxy to act as the emitter. The turn-on voltage was V/µm, as shown in the I-V curve in Figure 6.7. Moreover, the comparison of I-V curves of different CNT paste emitters is displayed in Table 6.2. CNT (a) Voltage (V) Figure 6.5 (a) I-V curves of various CNT pastes; (b) SEM image of the CNT paste emitter (b) cathode (a) (b) Figure 6.6 (a) I-V curves of LTCC CNT paste; (b) SEM image of the CNT paste emitter 82

96 0.313 Figure 6.7 I-V curve of sprinkled open-ended SWNT paste Table 6.2 Comparison of I-V performance of different CNT mixture emitters CNT mixture method Current Density Turn-on Voltage ERSO CNT Paste (Xintek Inc.) 5.5 ma/cm 2 at 3.5 V/µm 2.3 V/µm LTCC CNT Past 40 ma/cm 2 at 2.2 V/µm 0.9 V/µm FACCT CNT Paste (400 nm) 1.9 ma/cm 2 at 0.9 V/µm V/µm 6.3 CNT BuckyPaper Films As in previous sections, CNT BuckyPaper films were investigated to be field emitters in our research. According to the characteristics discussed in Chapter 4, the best turn-on voltage is VGCNF mixed SWNT in ratio 1:1 BuckyPaper film, V/µm, when current density is 1 µa/cm 2. The comparison of I-V plot of BuckyPaper films and SEM image are shown in Figure 6.8. The SEM image is the backside of the VGCNF/SWNT BuckyPaper film. 83

97 VGCNF (a) (b) CNT Figure 6.8 (a) I-V curves of various BuckyPaper films; (b) SEM image of the BuckyPaper emitter 6.4 Performance and Property Comparison of Emitters Made in Various Methods In the three different methods of CNT emitters, the best field emission properties of each will be selected and compare with each other. Comparison results are shown in Table 6.3, which listed the field at the current density, 1 ma/cm 2. From the lowest to the highest are 400 nm of open-ended SWNT paste (0.796 V/µm), the purified SWNT by laser ablation method (1.3 V/µm), and VGCNF/ SWNT BuckyPaper film (2 V/µm). Table 6.3 Comparison of different CNT emitters Emitter Prepared The lowest field at the Processing Method Producer Method current density 1 ma/cm 2 Vertically grown CNT 1.3 V/µm Laser Ablation (SWNT) Xintek. Inc CNT Paste V/µm 400 nm Open-ended FACCT SWNT BuckyPaper 2 V/µm VGCNF/SWNT FACCT 84