8.1: Advancements and Outlook of High Performance Active-Matrix OLED Displays

Size: px

Start display at page:

Download "8.1: Advancements and Outlook of High Performance Active-Matrix OLED Displays"

Rafe Kristopher Carpenter
6 years ago
Views:

8.1: Advancements and Outlook of High Performance Active-Matrix OLED Displays Takatoshi Tsujimura *, Wei Zhu, Seiichi Mizukoshi, Nobuyuki Mori, Koichi Miwa, Shinya Ono, Yuichi Maekawa, Kazuyoshi

1 8.1: Advancements and Outlook of High Performance Active-Matrix OLED Displays Takatoshi Tsujimura *, Wei Zhu, Seiichi Mizukoshi, Nobuyuki Mori, Koichi Miwa, Shinya Ono, Yuichi Maekawa, Kazuyoshi Kawabe, Makoto Kohno OLED Product Development, OLED Systems SPG, Kodak Japan Limited Abstract By introducing new terminology, active-matrix OLED s advanced performance can be well characterized. Luminous Flux and Assured Dynamic Range better represent the viewer s impression than previous terminology such as front luminance and/or viewing angle. Compensation circuit, low-stress driving technique, CVD condition, TFT shape, and high-efficiency OLEDs are the keys to enabling low-cost a-si AMOLED manufacturing in order to compete with other display technologies. 1. Introduction Active-matrix OLED (AMOLED) displays [1][][3] have been known as an emerging information display technology for years because of their superior features, such as wide viewing angle, fast response time, large contrast ratio, better color reproduction, wide temperature range operation, etc. Recently, though, liquidcrystal displays (LCDs) have been making steady improvement to overcome their weak points, such as viewing angle and response time. Although LCDs are composed of a larger number of parts compared to OLED displays, the maturity of LCD technology is reducing part costs, which results in a very low price for LCDs. This paper describes: (1) if AMOLED displays still have advantages over AMLCD displays, and () how low-cost OLED system architectures can be achieved to catch up with LCDs, especially in terms of backplane aspects, including both lowtemperature poly-crystalline silicon (LTPS) technologies and amorphous silicon (a-si) backplane technologies [4].. LTPS TFT Driving Most of the current AMOLED products have been manufactured with LTPS backplanes. LTPS thin film transistors (TFTs) can provide a large enough current with good stability compared to the a-si TFT method. Because OLED is a new technology competing with the incumbent LCD technology, naturally, people tend to use the same specifications and terminologies that are used for LCDs on OLEDs. However, such usage does not capture the true performance differentiation of OLEDs over LCDs. In our laboratory, LTPS-AMOLED displays are compared with commercially available AMLCDs to see what kind of advantage AMOLED has, beyond what is quoted in a standard specification. Also, a solution for Mura, or lack of uniformity, is discussed. Mura has been the biggest issue of LTPS-AMOLEDs, and it must be solved to achieve both good front-screen image with good uniformity and high yield..1 Misconceptions related to OLED Fig.1 Comparison between Kodak s.5" AM635LX AMOLED (left) and commercial AMLCD (right) Fig.1 shows the comparison between our AMOLED display [5] and a commercially available wide angle of view AMLCD display. Both AMOLED and AMLCD displays were characterized to compare their performance. The device structure of the AM635LX is described in Fig.. Red host +Red dopant Al Cathode Green host +Green dopant HT-1 KHI-1 Transparent Anode Blue host +Blue dopant Fig. Device structure of the.5" AM635LX AMOLED display The (x,y) chromaticity coordinate measurement results based on the 1931 CIE x,y chromaticity diagram are described in Fig.3. Also, the distribution of surface reflection color, which was reported by Spaulding et al. [6], is shown on the same diagram. The AM635LX s CIE x,y chromaticity gamut triangle is much larger than the AMLCD chromaticity triangle, and it is also designed to fit the space occupied by the distribution of real-world surface colors represented in Fig.3. * takatoshi.tsujimura@kodak.com Vol.38, Issue 1, pp (7).

2 y Color in real world Kodak AM55L OLED Kodak AM635LX OLED Typical LCD for DSC x Fig.3 Color gamut comparison Fig.4 describes the comparison of the luminance distribution as a function of viewing angle for the AMOLED and AMLCD. The AMOLED curve is close to a Lambertian distribution. On the other hand, the AMLCD s luminance distribution is more centered around the normal. This distribution difference frequently causes a misconception to those who look at the specification table. cd/m AMOLED can look brighter than 3cd/m in many occasions. This means that Front luminance measured along the normal, which is usually used for display characterization, does not give the correct information as to which display is brighter. To estimate how bright it is, namely, how much light energy is emitted from the display, the term Luminous Flux can be used. Luminance (cd/m ) OLED (AM635LX) LCD Viewing Angle (degrees) Fig.4 Luminance distribution comparison between AMOLEDs and AMLCDs Table 1 shows data representing an approximation 1 to total luminous flux. The equation for computing total luminous flux is given by: 8 Total luminous flux= 8 L S Ω i cos Θ Where, L is the luminance, S is the area, Ω i is the solid angle associated with each equal angular increment, and Θ i is each measured angle. 1 Approximate value by angular measurement. Angular measurements were made in the tip and turn directions, and averaged for Table 1. i Table 1 shows that an OLED display at the same luminance (e.g., 3 cd/m ) is emitting approximately 1.8 times more photon energy. This can also explain why the OLED at cd/m (765 lm) looks somewhat brighter than an LCD at 3 cd/m (595 lm) in terms of approximate Luminous Flux. When the color gamut is higher, the Helmholtz-Kohlrausch effect (HK effect) will also make the viewer perceive the display as brighter. Luminous Flux and the HK effect may be the reasons why OLED display looks brighter than its specification. Also, since OLEDs can produce blacks with lower luminance than blacks from LCDs, their contrast ratio along the normal and at all viewing angles is very high. Good contrast ratio across all angles contributes to the perception of OLEDs as brighter than LCDs. cd/m 3 cd/m OLED 765 [lm] [lm] (AM635LX) LCD 1491 [lm] 595 [lm] (IPS-LCD) Table. 1 Luminous Flux difference between AMOLEDs and AMLCDs. Viewing angle Fig.5 shows the difference in viewing-angle performance between AMLCDs and our AMOLED. Although LCD specifications claim that the viewing angle is over 17, contrast is lost when they are rotated to the azimuth angle as is shown in Fig.5 left. On the other hand, the AM635LX AMOLED display contrast measurement result shows over 7 for all angles between 17 to 17. This difference causes a large difference in the perception of the degree of contrast. Let us define an Assured Dynamic Range (ADR), where, ADR the minimum contrast within a viewing angle between 17 to 17. Therefore, the ADR of the LCD in Fig.5 is about 5 when the ADR of the AMOLED is 7. This causes the difference in the impression of image quality between LCDs and OLEDs, even when the claimed viewing angle specifications are similar < > Fig.5 Viewing angle performance with EZcontrast of an AMLCD (left) and an AMOLED (right) Although the specifications of AMOLED and AMLCD displays are close, the perceived difference is larger than that the specification values indicate. Such large perceived differences can be characterized by the new terminology Luminous Flux and ADR, as introduced in this paper Vol.38, Issue 1, pp (7).

3. LTPS Driving, A-Si TFT, and Non-ELA LTPS Driving To compete with LCDs, it is very important to achieve low costs. The following are approaches at achieving such low-cost displays. 3.

ELA LTPS causes TFT current variations that lead to Mura or luminance uniformity variation, as shown in Fig.6.

. In the past, AMOLED could not achieve LCD-level brightness uniformity due to the current variation of LTPS TFTs, in spite of various compensation schemes, such as voltage programming or current

3 3. LTPS Driving, A-Si TFT, and Non-ELA LTPS Driving To compete with LCDs, it is very important to achieve low costs. The following are approaches at achieving such low-cost displays. 3.1 LTPS driving In order to achieve low costs, it is very important to achieve high yield. So far, most AMOLED products are using excimer laser annealing (ELA) LTPS technology. ELA LTPS causes TFT current variations that lead to Mura or luminance uniformity variation, as shown in Fig.6. To achieve LCD-level display quality, it is necessary to compensate for the Mura appropriately. Fig.7 shows the compensated image after Global Mura Compensation (GMC).. In the past, AMOLED could not achieve LCD-level brightness uniformity due to the current variation of LTPS TFTs, in spite of various compensation schemes, such as voltage programming or current programming. GMC method for the first time achieved LCD-level uniformity and successfully solved AMOLED s issue. The detail of GMC will be presented in a future publication. Fig. 8 Newly developed AM76L AMOLED display Display size 3. Display format 16:9 Resolution QVGA Color number 16M Peak Luminance cd/m Viewing Angle U/D: 85/85º L/R: 85/85º White point (.31,.33) Table AM76L AMOLED display specification Fig.6 LTPS AMOLED display before Mura compensation (Minolta CA15 measurement) Fig.7 LTPS AMOLED display after GMC compensation (Minolta CA15 measurement) 3. Approaches to cope with the issues of a-si TFT LTPS TFTs normally require more than 6-9 masking steps, in addition to laser annealing, which constrains the throughput. Therefore, the LTPS TFT backplane tends to be expensive. Amorphous silicon TFT backplanes can be fabricated with 4-5 masking steps, and they are generally less expensive than LTPS TFTs. However, a-si TFTs have instability problems that can cause image sticking. To solve the instability issues, it is necessary to combine several approaches to make the instability invisible to the human eye. The following is a list of the approaches to be used to solve the instability issues of a-si TFT driving and make them less visible to the human eye: (1) Vth shift suppression by means of a driving technique () a chemical vapor deposition (CVD) condition to suppress Vth shift (3) TFT shape effects (4) high-efficiency OLED devices (5) compensation circuits Also, a-si TFTs have low mobility (.4-.8 cm /Vsec), which restricts the design. This design restriction is also discussed. (1) Vth shift suppression by means of driving technique The amount of a-si TFT instability can be changed according to Vol.38, Issue 1, pp (7).

4 the driving method. Tsujimura et al. [7] reports these approaches: (a) The saturation region operation suppresses the mobility degradation, and it can also cause a very small Vth shift. (b) Pulsed stress gives smaller Vth shifts than DC. (c) Rising and falling times affect the degradation (d) Alternate TFT stresses along the source and drain operations give poorer results (e) Negative bias releases the trapped charge and a healing effect can occur. By choosing the appropriate driving scheme, the instability can be minimized. () CVD condition to suppress Vth shift The CVD condition also affects the amount of instability. Super amorphous silicon [4,8], layer-by-layer method and microcrystalline give reduced Vth shift. These approaches normally give better stability with impact to the throughput of the CVD. (3) TFT Shape Effect As reported by [9-11], TFT shape can affect the amount of instability. Double-channeled TFTs and circle TFTs give reduced instability. It is important to remove the current concentration, which causes the accelerated degradation of a-si TFT. (4) High-efficiency OLED devices It is quite important to achieve a high efficiency OLED device [15] in order to make the design of an a-si TFT easier. The highefficiency OLED device would give, (a) TFT lifetime improvement due to low current, (b) TFT lifetime improvement due to lower temperature rise (Arrhenius rule), (c) OLED lifetime improvement due to lower temperature rise (Arrhenius rule), (d) yield improvement due to wider spacing/wiring design rule, (e) voltage drop reduction due to lower current, (f) cost reduction due to driver voltage reduction, and (g) power reduction Also, the maximum display size is also affected by the efficiency value. LMAX 9a I = η where, L MAX is the display luminance, I is the OLED maximum current,η, is the OLED current efficiency, and a is the pitch. In saturation region operation, the maximum current of the driver TFT can be expressed as, W I = µ COX ( VGS VTH ) L where, W is the channel width of driver TFT, L is the channel length of the driver TFT, µ C is TFT mobility, OX VGS is the capacitance of the TFT, is the gate voltage of the TFT, and VTH is the threshold voltage of the TFT. Therefore, 18LMAX a L ηµ = WC ( V V ) OX GS Fig.9 and Fig.1 describe the display size limitation as a function of OLED efficiency and TFT mobility. To achieve large-area AMOLED TV with a-si, it is necessary to increase the OLED efficiency or the TFT mobility. (ηµ)μιν [(cd/a)(cm/vsec)] TH Diagonal size of display [inch] Fig. 9 Display-size dependence of (efficiency TFT mobility) value. ημιν[cd/a] Diagonal size of display [inch] Fig. 1 Display-size dependence of minimum OLED efficiency. (5) Compensation circuit An a-si TFT requires larger TFT than LTPS. Naturally, the circuit tends to have large a parasitic capacitance. Such capacitance causes compensation errors after -level programming. The circuit reported by Hasumi et al. [1] uses a variable capacitor to cancel the compensation in order to have a very accurate compensation result. By reducing the Vth shift with approaches (1)~(5) and compensating the reduced instability error, invisible instability can be achieved [1-14, 16, 17]. 3.3 Non-ELA LTPS driving To solve the laser-related Mura problem, non-laser crystallization approaches such as MIC (Metal Induced Crystallization), MILC (Metal Induced Lateral Crystallization), MIUC (Metal Induced Vol.38, Issue 1, pp (7).

5 Unilateral Crystallization), magnetic-field crystallization, and metal-oxide TFT [18], microcrystalline silicon TFT [7] have been reported. Each technology has each problem, e.g. MIC s Ni contamination problem. Even with a non-ela approach, luminance variation due to the processing appears and compensation circuit may be necessary. Either an a-si, microcrystalline Si or a non-ela LTPS approach will open the door to achieving extended mother-glass-sized manufacturing, which is possible for the low-cost manufacturing of AMOLEDs. 4. Conclusions When characterizing displays, conventional terminology sometimes leads to misconceptions. That is why some AMOLEDs with identical specifications as other displays technologies look better. Luminous Flux, which represents an approximation to the amount of photon energy emitted out of the display, can explain the phenomena that AMOLEDs with the same peak luminance along the normal usually look brighter than LCDs. The term, Assured Dynamic Range expresses the AMOLED s richness in contrast from the oblique angle better than the terminology viewing angle. By appropriate terminology, AMOLED s superiority can be well described. To achieve low-cost manufacturing of AMOLED, the choice of the backplane technology is important. Compensation circuit, low-stress driving technique, CVD condition, TFT shape, and high-efficiency OLEDs are the keys to achieving a-si driving. An a-si TFT, microcrystalline Si TFT and a non-ela LTPS technologies are competing to achieve low-cost AMOLED manufacturing. 6. Acknowledgements The authors thank Mary Jane Hellyar, Andrew Sculley, Gopalan Rajeswaran, Kiyoshi Yoneda, Kazunobu Mameno and Megumi Yamaguchi for their support and advice. The authors also thank Paula Alessi and Takumi Shibata for front-screen measurements. Also the authors thanks to LG Philips LCD team for the collaboration to fabricate the 3. display. 7. References [1] T. Sasaoka et al., A 13.-inch AM-OLED Display with Top Emitting, Society for Information Display 1 Proceedings, p. 384 (1). [] G. Rajeswaran et al., Active Matrix Low Temperature Poly-Si TFT / OLED Full Color Displays: Development Status, Society for Information Display, Proceedings, p. 974 (). [3] J. Sanford, E.S. Shlig Direct View Active Matrix VGA OLED-on-Crystalline-Silicon Display, Society for Information Display 1 Proceedings, p. 376 (1). [4] T. Tsujimura et al., A -inch OLED display driven by super-amorphous-silicon technology, Society for Information Display 3 Proceeding, p.6 (3). [5] T. Tsujimura, W. Zhu, S. Mizukoshi, N. Mori, K. Kawabe, M. Kohno, K. Onomura, Design for the OLED as the best display technology, OLEDs Asia 6 Proceedings (6). [6] K. E. Spaulding et al., Reference input/output medium metric RGB color encodings (RIMM/ROMM RGB), ICS Conference, March, Portland, OR, pp (). [7] T. Tsujimura, Amorphous/Microcrystalline silicon thin film transistor characteristics for large size OLED television driving, Japanese Journal of Applied Physics, vol. 43, no.8a, pp (4). [8] T. Tsujimura, F. Libsch, P. Andry, Amorphous silicon thin film transistor for large size OLED television driving, Journal of the SID, vol. 13 (), p. 161 (5). [9] T. Tsujimura, Large-size OLED display, Materia Japan by Japan Institute of Metals, vol. 43 (9), p.75 (4). [1] T. Tsujimura, Amorphous silicon technology and top emission structure, Organic Electroluminescence Materials and Technologies published by CMC books, p. 9 (4). [11] T. Tsujimura, Amorphous silicon driven Active-matrix display, Organic Electroluminescence Handbook published by REALIZE Science & Engineering, p. 45 (4). [1] T. Hasumi, S. Takasugi, K. Kanoh, Y. Kobayashi, New OLED Pixel Circuit and Driving Method to Suppress Threshold Voltage Shift of a-si:h TFT, Society for Information Display 6 Proceedings, p (6). [13] S. Ono, Y. Kobayashi, An Accelerative Current- Programming Method for AM-OLED, IEICE Transactions on Electronics, vol. E88-C, no., p. 64 (5). [14] S. Ono, Y. Kobayashi, K. Miwa, T. Tsujimura, Pixel Circuit for a-si AM-OLED, International Display Workshop 3 Proceedings, p. 55 (3). [15] T.K. Hatwar, J.P. Spindler, S.A. Van Slyke, High Performance Tandem OLEDs for Large Area Full Color AM Displays and Lighting Applications, IMID/IDMC 6 Digest, p (6). [16] Y. Maekawa, K. Miwa, S. Ono, T. Tsujimura, A New Feedback, Constant Current, and Constant Drain Bias test for Amorphous-Silicon Backplane of Active-Matrix Organic Light- Emitting Diode Displays, The 9th Asian Symposium on Information Display, 6 [17] S. Ono, K. Miwa, Y. Maekawa, T. Tsujimura, Shared Pixel Compensation Circuit for OLED Displays, The 9th Asian Symposium on Information Display, 6. [18] S.K. Park, C. Hwang, J. Lee, S.M. Chung, Y.S. Yang, L. Do, H.Y. Chu, Transparent ZnO Thin Film Transistor Array for the Application of Transparent AM-OLED Display, Society for Information Display 6 Proceedings, p. 5 (6) Vol.38, Issue 1, pp (7).

Overview of All Pixel Circuits for Active Matrix Organic Light Emitting Diode (AMOLED)

Chapter 2 Overview of All Pixel Circuits for Active Matrix Organic Light Emitting Diode (AMOLED) ---------------------------------------------------------------------------------------------------------------