A Field-Emission Display with an Asymmetric Electrostatic-Quadrupole Lens Structure

Transcription

1 Japanese Journal of Applied Physics Vol. 44, No. 12, 2005, pp #2005 The Japan Society of Applied Physics A Field-mission Display with an Asymmetric lectrostatic-quadrupole Lens Structure Tae Sik OH 1;2, Jeong Hee L 2, Seong ui L 2, Kyoung Won MIN 2, Sung Kee KANG 2, Ji Beom YOO 1, Chong Yun PARK 1 and Jong Min KIM 2 1 Department of Nano Science and Technology, Sungkyunkwan University, 300 Chunchun Dong, Jangan-Gu, Suwon , Korea 2 Materials and Devices Research Center, Samsung Advanced Institute of Technology, P. O. Box 111, Suwon , Korea (Received March 24, 2005; revised August 16, 2005; accepted August 28, 2005; published December 8, 2005) An asymmetric electrostatic-quadrupole lens (AQL) system for high definition field emission displays (HD-FDs) was proposed. It was applied to the double-gated structure where the emitters are a thick layer of carbon nanotube paste such as a flat surface emitter. The AQL structure was designed with two opposing planar electrodes of noncircular apertures which generate the quadrupole electric field. Utilizing a design of field emitter arrays (FAs) with AQL, an optimized beam shape with horizontal reduction and vertical elongation was obtained. According to three-dimensional (3D) simulation results, this AQL structure exhibited excellent focusing effects that satisfied the aspects of pixel size and shape in HD-FDs. [DOI: /JJAP ] KYWORDS: asymmetric electrostatic quadrupole lens, field emission display 1. Introduction The conventional field emission displays (FDs) using phosphor, that can only be excited by an electron beam (e-beam) of low energy, have a narrow vacuum gap (below 1.1 mm) between the emitters and the anode in order to minimize the driving voltage. In this case, the most important design factor is the efficiency of the phosphor which determines the brightness rather than the size of the e-beam which dominates color purity. This structure has a simple manufacturing process, but is not effective until the luminous efficiency problems of low voltage phosphor are solved. Recently, high-anode-voltage-operated FDs (HV-FDs) employing the phosphors used in cathode ray tubes (CRTs) have attracted research attention due to their improved color purity, high brightness and long lifetime. In HV-FDs, a high anode voltage is required for the purpose of accelerating electrons to high energy. However, when the anode voltage is highly increased in the narrow vacuum gap, unpredictable arcing easily occurrs. Therefore, HV-FDs require a wide vacuum gap between the emitters and the anode. In this case, the size of the e-beam, especially its horizontal size, that dominates the color purity rather than the phosphor efficiency which determines the brightness, is a more important design factor than is the case with conventional FDs. The wide vacuum gap demands a design of structures capable of focusing. Several focusing structures have been reported: double-gated type, 1 6) planar-electrode type 7,8) and self-focus cathode electrode type. 9) In a double-gated type structure, a focus gate electrode (FG) is stacked on the extraction gate electrode () with an additional insulating layer, where the additional insulating layer is required to sustain electrical breakdown between and. If a thick film is used as the insulating layer, several problems can be expected such as substrate deformation by the additional thermal process and the degradation. Nevertheless, the application of has advantages such as a blocking effect against anode voltage penetration, which may raise diode emission, as well as electron beam focusing. In a planar-electrode type structure, -mail address: ots99.oh@samsung.com 8692 the focus-electrode is located on the coplanar with the gate electrodes. This structure has a simple manufacturing process, but is restricted for high resolution. In a self-focus cathode electrode type structure, the self-focus electrode is in contact with the cathode electrode surrounding the emitter layer at the center of each gate aperture. This type is effective in the focusing scheme, but the emission current is reduced owing to the auxiliary self-focus electrode surrounding the emitter layer. In this report, we proposed an asymmetric electrostaticquadrupole lens (AQL) structure based on a double-gated type design. The proposed structure is designed to produce asymmetric quadrupole electric fields with off-centered two planar electrodes using a horizontally and vertically elongated aperture. The AQL structure results in an electron beam with a horizontally reduced and vertically elongated shape due to the AQL effect (as explained in the next section). According to three-dimensional (3D) simulation results, this structure demonstrates an excellent focusing effect that satisfies the aspects of a relatively small pixel size and rectangular shape in HD-FDs. In this study, we used a commercial simulator (Opera-3D) using the finite element method (FM). 2. Asymmetric lectrostatic-quadrupole Lens system The ideal electrostatic-quadrupole lens (QL) system consists of four parallel electrodes with hyperbolic crosssections as shown in Fig ) It has four planes of symmetry intersecting along the z-axis with an angle =4 between them. The lens, centered at z ¼ 0, extends in the z-direction. The aperture of the lens (2L) is defined by the diameter of the hypothetical circle tangential to the four electrodes. We assume one pair of the opposing electrodes (A and A 0 ) to be held at a positive potential (þv) with respect to ground while the other opposing pair (B and B 0 )is at a negative potential ( V). The established electric field lines are shown as dashed lines. It is assumed here that electrons move in the direction of the reader (the positive z-direction). Those electrons initially in the yz-plane will diverge toward the positive potential on A and A 0 electrodes. However, those electrons initially in the xz-plane will be repelled by the negative potentials on electrodes B and B 0 converging toward the z-axis. Therefore the incident circular

2 Jpn. J. Appl. Phys., Vol. 44, No. 12 (2005) T. S. OH et al Incident e-beam V 1 < V 2 y A(V 2 ) Incident e- beam V 2 L V 1 <V 2 B(V 1 ) -L L x B (V 1 ) V 1 -L Deformed e-beam A (V 2 ) lectric Field Line Force (F=-e) Deformed e-beam Fig. 1. Principle of an ideal QL system consisting of four hyperbolic electrodes: A, A 0, B, and B 0. e-beam will be changed to a horizontally elongated and vertically reduced shape. If a vertically elongated and horizontally reduced e-beam shape is preferred, the exchange of potentials between two electrode pairs would simply accomplish the desired shape. Moreover, a control of the aspect ratio of the e-beam is achieved by varying the opposing apertures size. In practice, however, the electrodes with hyperbolic cross sections are difficult to fabricate and a rather simple structure is demanded in FDs. Therefore, we have proposed a symmetric electrostatic-quadrupole lens (SQL) system 13) as shown in Figs. 2 and 2, which is composed of two opposing plates with elongated rectangular apertures in a 90 staggered position and with different potential applied to each plate. Applying this SQL structure to the double-gated type FDs, a positive potential (þv) is applied to the (V 2 ) and a negative potential ( V) is applied to the (V 1 ). In this structure, a horizontal slot on the first plate is aligned with the vertical slot on the second plate, respectively, to result in vertically elongated e-beam shape. It should be noted that the fields produced in this SQL system are similar to those produced by the ideal QL system, even without the complex four electrodes. The simulation results based on this structure are shown in Figs. 3 and 3, Fig. 2. QL system in FDs: structure with two planar electrodes using a horizontal slot on the first plate and a vertical slot on the second plate, and diagram of diagram of an SQL structure, and an AQL structure. presenting the horizontal and vertical sections of equipotential lines, together with schematic e-beam trajectories. As indicated in the horizontal section, the larger aperture of the extraction gate and the smaller aperture of the focus gate produce a focused beam by converging force on the emitted electrons from the emitter layer. On the contrary, in the quipotential line e-beam trajactory Deflected e-beam trajactory K K K Horizontal Verticle Verticle [SQL sytem] [AQL sytem] Fig. 3. quipotential lines and e-beam trajectories in the QL system: horizontal section of SQL, vertical section of SQL, and vertical section of AQL.

3 8694 Jpn. J. Appl. Phys., Vol. 44, No. 12 (2005) T. S. OH et al. vertical section, the smaller aperture of the extraction gate and the larger aperture of produce an elongated e-beam by diverging force on the emitted electrons from emitter layer. Figure 2 shows the structure of the AQL system which consists of an aperture and a vertically offcentered aperture. Figure 3 shows a vertical section of the equipotential lines together with the schematic e-beam trajectories from the AQL effect. The AQL structure can be accomplished by moving the center of the aperture slightly closer to the pixel axis. In this AQL structure, the bending of the e-beam trajectories occurs at the upper portion of the e-beam shape. This is a useful method to reduce the size of the vertical e-beam without changing the number of emitters or the structure of the field emitter arrays (FAs). FGI GI 3. xperiment and Simulation Recently, we fabricated a double-gated type test vehicle having a robust focus gate structure comprised of a thick SiO x focus gate insulator (FGI) and a Cr focus electrode for HV-FDs. 5) The scanning electron microscope (SM) image of our double-gated type structure is shown in Fig. 4. The diameter of the emitter is 10 mm, and that of the concentric circular shaped extraction gate hole and focus gate hole is 16 and 60 mm, respectively. The thickness of the extraction gate insulator (GI) and FGI is 3 and 6 mm, respectively. These layers were deposited by radio frequency plasma enhanced chemical vapor deposition (rf-pcvd), with optimization of quality and thickness based on SiH 4 / N 2 O gas ratio. The FGI layer is composed of sparse SiO x, which gives it a high etching rate of more than 0.6 mm/min and a low film stress of less than 50 MPa. The stoichiometric expression of the FGI is SiO x, where x is less than 2. After processing of the FGI and formation, single walled CNT emitter layers are formed by screen printing the paste followed by backside ultraviolet exposure through circular a-si holes having diameters of 10 mm. The e-beam focusing capability of our double-gated structure was measured in a vacuum chamber (vacuum pressure < Torr) with a 1.1 mm gap between and the anode. Figure 5 is a series of e-beam spot images of our test vehicle using a green phosphor-screened anode plate, where the voltage of the extraction gate (V g ), cathode Fig. 4. SM image and computer modeling of a conventional double-gated type structure with circular extraction gate electrode and focus gate electrode. (V k ) and anode (V a ) was fixed at 60 V, 0 V, and 2 kv, respectively, and the focus gate voltage (V f ) was applied from 0 to 40 V due to the arcing problem. According to the images, the e-beam size was greatly affected by the focusing voltage. As the negative focusing voltage was increased, the size of the e-beam was decreased, but a halo phenomenon appeared around the main beams. The halo phenomenon was caused by over focused electrons. By modeling of the test vehicle structure as shown in Fig. 4, the halo phenomenon was analyzed by 3D simulator. In these simulations, electron emission was assumed on a flat surface of CNT emitters with a work function of 5 ev, and the field Halo Main beam Fig. 5. lectron beam spot image of experimentally obtained double-gated type test vehicle: the voltage of the focus gate electrode was varied: 0, 20, and 40 V.

4 Jpn. J. Appl. Phys., Vol. 44, No. 12 (2005) T. S. OH et al (d) Fig. 6. Simulation results by 3D simulator: the size and shape of e-bem at the anode plate when the voltage of the focus gate electrode was varied 0, 20, 30, and (d) 40 V. enhancement factor and an emission constant were 1800 and 1: , respectively. 14) We set the initial energy of electrons at about ev (where T ¼ 300 K) and the emitting angle at a perpendicular direction from the emitter surface. Figure 6 shows the simulation results of the e-beam shape at the anode. The simulated images of the e-beams matched well with the experimentally obtained images. The over-focused e-beams, as shown in Figs. 5 and 6(d), caused a serious problem in the color purity and contrast in FDs. In this report, we proposed a new structural design featuring the basic AQL system presented in Fig. 7. The structure has a rectangular shaped extraction gate aperture (h60 mm v20 mm) and focus gate aperture (h40 mm v50 mm), in which the two apertures are intercrossed and the focus gate aperture is off-set by 10 mm as shown in Fig. 2. The thickness of the extraction gate insulator and focus gate insulator is 5 and 6 mm, respectively. The size and thickness of the emitter are h10 mm v10 mm and 2 mm, respectively. In order to compare the e-beam focusing capability of this structure, a conventional circular gate structure and a SQL structure are shown in Figs. 7 and 7. In the circular type, the diameters of the extraction gate hole and focus gate hole are 20 and 60 mm, respectively. In the SQL structures, the dimensions of the extraction gate aperture and focus gate aperture are h60 mm v20 mm and h40 mm v60 mm, respectively. The thickness of the GI and FGI is the same as that of AQL structure. The vacuum gap between the anode and the focus electrode is set to 1.5 and 2.0 mm, respectively, to prevent arcing from occurring under high anode voltage condition above 5 kv. Figure 7(d) shows the simulation results of the e-beam size and shape for the three structures. In the case of the AQL and SQL structures, the e-beam shape become horizontally reduced and vertically elongated, which is ideally suited for adapting a pixel shape rather than a circular gate structure. The vertical e-beam size (BS v ) was and mm in the AQL structure, and and mm in the SQL structures, when an anode voltage of 7.5 and 10 kv, respectively, was applied and the Structure type 10 µm 20 µm AQL 20 µ m 20 µ m SQL Circular type (d) Vacuum gap/anode voltage 1.5 mm/7.5 kv 2.0 mm/10.0 kv Upper Lower h:137.3 µm v:357.6[164.2] h:156.6 µm v:410.9[188.6] h:119.2 µm h:135.6 µm v:373.0[186.5] v:428.7[214.3] h:195.8 µm v:195.3 µm h:221.1 µm v:220.7µm Remark Vg: 40V Vf: 0V Vk:-38V -38V [ ]:Upper Size Vg: 40V Vf: 0V Vk:-36V -36V [ ]:Upper Size Vg: 40V Vf: -60V Vk:-14V -21V Fig. 7. Simulation results of e-beam size and shape according to each electron lens structure: modeling of conventional, circular doublegated structure modeling of the basic SQL structure modeling of the AQL structure, and (d) the size and shape of electron beam at the anode. anode current (I a ) was adjusted at 15 na. specially, as is shown in Fig. 7, the upper portion of the e-beam shape (Upper) by the AQL was reduced more than that by SQL. The upper e-beam size of the AQL structure was and mm, compared to 186.5, and mm of the SQL structures. These results confirmed that the AQL structure operated well, in line with the expected trajectories shown in Fig. 3.

5 8696 Jpn. J. Appl. Phys., Vol. 44, No. 12 (2005) T. S. OH et al. 4. Results and Discussion Generally, a wide (16 : 9) display needs more than 1280 dots 768 lines to achieve high definition (HD) resolution. In the case of 38-inch wide FDs, the size of one sub-pixel (R, G, B respectively) is about h233 mm v632 mm, and about 8 14 emitter layers, each sized mm 2 and 1 2 mm thick, were involved as shown in Fig. 4. Figure 8 shows the SQL structure. The dedicated focus gate aperture sized h60 mm v60 mm has an extraction gate aperture of h60 mm v20 mm and five emitter layers of one sub-pixel in order to adjust the vertical e-beam size. Figure 9 shows the AQL structure. The first (f, f 0 ) and second (s, s 0 ) apertures from the extraction gate aperture Fig. 8. Simulation results by 3D simulator: modeling of the SQL- FAs structure, trajectories of e-beams, and the size and shape of electron beam at the anode plate where the anode voltage was set to 7.5 kv. are offset by 5 and 10 mm, and are sized h60 mm v55 mm, and h60 mm v50 mm, respectively. The reduced anode current due to the small number of emitter layer is compensated by increasing the size of each emitter layer from h10 mm v10 mm to h20 mm v10 mm. Figures 8 and 9 show the trajectories of the electrons emitted from the emitters, and Figs. 8 and 9 present the impact points of the electrons on the anode plate as determined by 3D simulator. When an anode voltage of 7.5 kv was applied for the vacuum gap of 1500 mm, the horizontal beam size were and mm, and the vertical beam size was and mm, respectively. This result adequately satisfies the pixel size and shape for the above-mentioned HD-FDs. Although the horizontal dimension is a larger 14.2 mm for the AQL case, the vertical dimension is smaller at 74.1 mm than the SQL case. Figures 9 and 10 show the effect of applying the AQL system in the pair of the aperture outer portion (f, f 0 and s, s 0 ) from Fig. 9. In this case, when the offset amount was increased, the vertical e-beam size was gradually decreases but the horizontal e-beam size underwent little changes. Therefore, the AQL system presented here is a useful method to reduce the size of the vertical e-beam without changing the number of emitters or the structure of FAs. 5. Conclusions We have proposed an AQL system in the double-gated type FDs where the emitter could be a thick layer of carbon nanotube fabricated by printing method. The structure to produce an asymmetric quadrupole electric field is composed of two, off-centered, planar electrodes with vertically s f f s s0/f 0 s5/f 0 s10/f 0 s15/f 0 s20/f 0 s20/f 10 Non tilt s: 10µm tilted s: 20µm tilted f : 10µm tilted Fig. 9. Simulation results by 3D simulator: modeling of AQL-FAs structure the trajectories of electron beams, and the size and shape of electron beam at the anode plate when the amount of offset is varied.

6 Jpn. J. Appl. Phys., Vol. 44, No. 12 (2005) T. S. OH et al e - beam size [um] 800 Va=7.5kV Ia Va=10kV Ia Ver. Ver. Hor. Hor amount of tilt [um] Fig. 10. ffect of AQL system (at f, f 0 ¼ 0): horizontal e-beam size, anode current, and vertical e-beam size. and horizontally elongated apertures. An electron beam with horizontally reduced and vertically elongated shape originating from the AQL effect was obtained by 3D-simulation. We have designed a new FA structures which could achieved an excellent focusing effect of e-beams at high voltages. According to 3D-simulation results, despite a focus gate voltage of zero, the AQL structure exhibited excellent focusing effects suitable in terms of pixel size and shape for application in 38-in. wide HD-FDs anode current [ *10-9 A] 1) Y. Lee, S. Kang and K. Chun: J. Micromech. Microeng. 7 (1997) ) C. Py, M. Gao, S. R. Das, P. Grant, P. Marshall and L. LeBrun: J. Vac. Sci. Technol. A 18 (2000) ) L. Dvorson and A. I. Akinwande: J. Vac. Sci. Technol. B 20 (2002) 53. 4) L. Dvorson, I. Kymissis and A. I. Akinwande: J. Vac. Sci. Technol. B 21 (2003) ) J. H. Choi, A. R. Zoulkarneev, Y. W. Jin, Y. J. Park, D. S. Chung, B. K. Song, I. T. Han, H. W. Lee, S. H. Park, H. S. Kang, H. J. Kim, J. W. Kim, J.. Jung, H. G. Baek, S. G. Yu and J. M. Kim: Appl. Phys. Lett. 84 (2004) ) J. H. Choi, A. R. Zoulkarneev, Y. W. Jin, Y. J. Park, D. S. Chung, B. K. Song, I. T. Han, H. J. Kim, H. W. Lee, S. H. Park, H. S. Kang, M. J. Shin, H. J. Kim, K. W. Min, J. W. Kim, J.. Jung and J. M. Kim: IVNC2004, p ) C.-M. Tang, T. A. Swyden and A. C. Tang: J. Vac. Sci. Technol. B 13 (1995) ) W. D. Kesling and C.. Hunt: I Trans. lectron Devices 42 (1995) ) Y. S. Choi, Y. S. Cho, J. H. Kang, Y. J. Kim, I. H. Kim, S. H. Park, H. W. Lee, S. Y. Hwang, S. J. Lee, C. G. Lee, T. S. Oh, J. S. Choi, S. K. Kang and J. M. Kim: Appl. Phys. Lett. 82 (2003) ) P. W. Hawkes and. Kasper: Principles of lectron Optics (Academic Press, New York, 1989) Vols. 1 and 2. 11) P. W. Hawkes: Advances in Imaging and lectron Physics (Academic Press, New York, 1999). 12) R. R. A. Syms, T. J. Tate, M. M. Ahmad and S. Taylor: I Trans. lectron Devices 45 (1998) ) T. S. Oh: patent pending. 14) T. S. Oh, S.. Lee, J. H. Lee, J. B. Yoo, C. Y. Park and J. M. Kim: J. Appl. Phys. 98 (2005)