When Samsung unveiled its 40-in. AMOLED TV at SID 2005, it signaled the biggest salvo to date in the battle to join – and ultimately conquer – LCD and PDP TVs as the dominant technology for large-screen flat-panel displays. Here, Samsung explains how far OLED technology has come and how far it needs to go in order to enter the marketplace as a contender.

by Kyuha Chung, Joonhoo Choi, Jaehoon Jung, Beohmrock Choi, Jinkoo Chung, Baek-Woon Lee, and Changwoong Chu

IF THE BATTLE between liquid-crystal displays (LCDs) and plasma-display panels (PDPs) for domination of the large flat-panel-display (FPD) TV business is the industry's version of the Cold War, then the battle for superior image quality is equivalent to the nuclear-arms race. It is remarkable how much effort and money have gone into achieving the best color reproduction, brightness, contrast ratio, and sharpness in still and moving images, and, as of today, it seems the overall image quality of both technologies are comparable. Both sides realize, therefore, that the war could well be won by lowering manufacturing costs. LCD manufacturers have reacted by making the motherglass larger and larger, as we approach the launching of the first Gen 8 fabs. However, as the size of the motherglass inevitably marches toward Gens 9 and 10, the risk and the amount of capital investment increase exponentially.

Therefore, Samsung believes that future FPDs should be structurally simpler so that materials costs are lower from the onset of production. The answer to lowering material costs may well lie in the development of large active-matrix organic light-emitting-diode (AMOLED) displays.

An example of the cost breakdown for a 40-in. LCD TV is shown in Fig. 1. The materials cost accounts for about 75% of the total manufacturing cost, and out of that 75%, the backlight unit, polarizer, and color filter eat up 65%. These components are completely unnecessary in an AMOLED display, indicating that this technology has the potential for a much better cost structure than that of LCDs.Two types of companies have interest in AMOLEDs. The first group seems to be look-ing for an opportunity to penetrate the FPD market with an alternative technology. Previously, these firms have been unable to enter the FPD market because of the very high entry barriers created by existing LCD companies.

The other group includes the LCD manufacturers, such as Samsung Electronics, who currently produce LCDs and are preparing to manufacture OLEDs in the future because AMOLEDs can provide a solution to the inherent problem of LCD material cost while providing similar performance. These companies are aiming high. If the technology for large-sized AMOLED TV can be secured, it will most assuredly be applied to smaller-sized AMOLEDs. That is why Samsung developed a 40-in. high-definition (HD) AMOLED TV having an amorphous-silicon thin-film-transistor (a-Si TFT) backplane, a color filter, and a white electroluminescent (EL) emitter, which was unveiled at the SID 2005 International Symposium, Seminar, and Exhibition last May in Boston.

 Samsung Electronics Co.

Fig. 1: The Materials cost accounts for 75% of the total manufacturing cost for a 40-in. LCD TV.

Large-Sized AMOLED Development

In 2005, Samsung SDI announced an investment plan for a Gen 4 AMOLED line that will start production in Q4 '06. LG.Philips LCD, which demonstrated a 20-in. panel at FPD International in 2004, has secured a Gen 4 low-temperature-polysilicon (LTPS) line in Gumi, Korea, to be used for AMOLED production. Seiko-Epson is planning to develop 40-in. AMOLEDs manufactured by ink-jet printing by 2007. However, not all OLED companies are preparing for expansion into active-matrix addressing. For example, Tohoku Pioneer, one of the first companies to commercialize OLEDs, has announced that it is completely withdrawing from the AMOLED business.

In sharp contrast to the companies mentioned above, Samsung Electronics has been working on a-Si TFT AMOLEDs from the onset. We developed a 14.1-in. WXGA AMOLED in August 2004, a 21-in. WUXGA in January 2005 [Fig. 2(a)], and the world's largest (40-in.) WXGA AMOLED [Fig. 2(b)] in May of the same year. This incredible rate of development is an indication of how determined Samsung Electronics is to manufacture large-sized AMOLEDs.

A Variety of Technologies

Large-sized AMOLEDs require a TFT backplane and driving technology, along with OLED-material and process-technology development. The technology available to make the TFT backplane typically includes either LTPS or a-Si. LTPS TFTs can be fabricated by using two crystallization methods: one with and the other without the use of a laser beam. There are two laser-beam techniques: excimer-laser annealing (ELA) is the most commonly used in the industry, while the sequential lateral solidification (SLS) method has recently became available. Both tend to result in the non-uniformity of the TFT threshold voltage (V_th), causing very fine mura artifacts. Various compensating pixel circuits can be utilized to overcome this problem, but it is still one of the major factors in determining yield. Another issue is the limited availability of the required laser equipment. Currently, the largest equipment available for the production of LTPS TFTs is only applicable to Gen-4-sized motherglass (730 x 920 mm), while the equipment for the production of a-Si TFTs for TFT-LCDs has already exceeded Gen 7 (1870 x 2200 mm).

Crystallization can also be achieved by using a non-laser-annealing method. The techniques include metal-induced lateral crystallization (MILC), metal-induced crystallization (MIC), and solid-phase crystallization (SPC). In these methods, the a-Si precursor film is crystallized by heating to a temperature above 650–750°C, either in the presence of a very thin metal catalyst such as Ni or without any catalyst at all. TFTs prepared by these techniques are relatively more uniform in their performance, so the displays may be mura-free. Overall, LTPS TFTs have good mobility and excellent V_th stability under long gate bias stress. The disadvantages, however, include relatively high process cost and high TFT off-currents, which causes severe vertical cross-talk on the display.

The a-Si TFT is widely used in the LCD industry. Although it has low mobility (< 1 cm²/V-sec), there are a number of good features built in, such as excellent uniformity, scalability (greater than Gen 7), and low process cost. However, poor V_th stability is a major drawback. During TFT operation, a gate bias stress on the TFT V_th tends to increase, causing the drain current passing through the TFT to decrease. As a result, the display gets darker and darker with time. It is critical to find a solution to this problem. Table 1 shows a comparison of the properties of LTPS and a-Si TFTs.

 (a)  (b) Samsung Electronics Co.

Fig. 2: Shown are photographs of Samsung's (a) 21-in. and (b) 40-in. AMOLED TVs.

Another major concern is the limits of the fine-metal-mask method that has been used in OLEDs. The fine-metal-mask mode cannot be expansible to more than Gen 4 substrate application under existing technology, thus many approaches have been proposed to overcome this issue, such as the use of laser-induced thermal imaging (LITI), ink-jet printing, a white OLED with a color filter (CF),etc. Ink-jet printing appears to be the most promising in terms of expansibility and cost, while on the material side, polymer-OLED (P-OLED) technology needs more time to catch up to the performance of small-molecule OLED (SM-OLED) technology.

When the white OLED with CF route is applied to a display, the CF is prepared by conventional lithography, as is done in the LCD industry, and then the stack of organic layers in the OLED is evaporated on it without any patterning. Since the anode – indium tin oxide (ITO) or indium zinc oxide (IZO) – is patterned to be aligned with the CF subpixel, the panel can display full color while the OLED stacks are deposited in one large piece to cover the entire display area. One drawback, however, is relatively lower light efficiency than that for the shadow-masked RGB system because only about one-third of the white light can pass through the CF. To improve the situation, a display that uses a RGBW four- primary-color system was developed. The RGBW CF system improves the light efficiency by as much as 50% versus the strict RGB CF structure. On the other hand, when applying a white OLED, it may sharply reduce the most important material cost, especially for large-sized displays because the evaporator for white-OLED deposition does not require fine metal masks for RGB patterning. One would expect the equipment to become much simpler and the material usage during evaporation to become more efficient. Comparison data between various patterning methods are illustrated in Table 2.

Samsung's 40-in. AMOLED Technology

Samsung Electronics announced the production of a 40-in. full-color AMOLED TV last year. However, many technical obstacles must be overcome before it can enter the flat-panel TV market currently dominated by PDP and AMLCD TVs. What follows is our approach to entering the large-sized-TV market.

Full-color AMOLED displays based on a white emitter with an RGB color-filter array have been reported as an alternative technology to those displays with patterned RGB emitters due to their relatively higher cost. However, RGB displays based on a white emitter have a disadvantage in power consumption because part of the energy of the white light is absorbed by the color filters. Recently, a white-emitter-based AMOLED display with an RGBW pixel format has been demonstrated. It consumes approximately one-half the power of an analogous white-emitter-based RGB display. The RGBW pixel format developed by Samsung to achieve high luminescence and a high contrast ratio for AMLCDs has been applied to the AMOLED display for the first time. It works by using unfiltered white subpixels with a relatively high efficiency to replace the combined emission from the lower efficiency RGB subpixels.

The successful performance of the RGBW format requires an efficient and stable white emitter. In order to minimize the power consumption of RGBW displays, it is also important that the white OLED emits close to the desired white point of the display. Although most development work has been accomplished on simple test devices, the emission of the white OLED must also be optimized to the display white point using an active-matrix substrate because of additional dielectric layers for planari-zation and passivation that the emitted light must pass through. These layers affect the spectral characteristics of the emission as a result of thin-film interference effects. In addition to an optimized white point, the white OLED display must also be designed to have little or no perceptible color shift when viewed at an oblique angle, and of course, an acceptable lifetime.

To compete with AMLCDs in the flat-panel TV market, AMOLEDs must adopt a cost-effective mass-production process. To attain an inexpensive fabrication process for large-sized substrates, an a-Si TFT architecture is the desired choice, despite its intrinsic performance limitations such as poor stability and low electron mobility. There are also existing LCD-panel facilities that produce large-sized a-Si TFT backplanes. LCD-TV panels larger than 80-in. on the diagonal have already been fabricated in these facilities. Samsung has an advantage by using this infrastructure to produce large-sized AMOLED displays by simply adding OLED processing equipment to the existing a-Si TFT process line.

A five-mask back-channel etch (BCE) process with an inverted staggered a-Si TFT architecture integrated with a color filter on array (COA) has been adopted. The COA process is identical to that used in LCD panels, except for the need of an additional photolithography process for the wall layer.

The field-effect mobility of the a-Si:H TFT is about 0.75 cm²/V-sec with a threshold voltage of 2.5–3.0 V. There are some on-current variations of less than 10% in the switching TFTs over the panel. Despite this variation, it is hard to perceive any variation in brightness, because the a-Si:H TFTs have good short-range uniformity.

In order to use an a-Si TFT backplane to drive a 40-in. AMOLED, we had to develop a panel architecture that can handle the very large current required to drive the OLED pixels as well as mitigate the potentially large voltage drop across the length of V_dd. It requires the optimization of metal lines with low resistance, the pixel layout, and color-filter architecture. We have chosen a four-color, RGBW, checkerboard subpixel layout and optimized the metal line; i.e., optimization of the thickness and material for V_dd. We have also developed an optimum panel design to reduce possible panel defects, which greatly increases the yield of line-defect-free panels. In this article, some AMLCD mass-production process technologies have been examined for AMOLED application to determine the quality of the TFT backplane before continuing the EL deposition process.

 Samsung Electronics Co.

Fig. 3: The EL spectra of a three-peak white emitter, with and without a buffer layer, used in a Samsung OLED display is shown.

 Samsung Electronics Co.

Fig. 4: Lifetime expectation to 50% initial luminance for both an improved three-peak white OLED and a conventional white OLED display.

The evidences of our white OLED with RGBW format are discussed below. Figure 3 shows the EL spectra of a three-peak white emitter with a buffer layer included in our OLED device. By incorporating the buffer layer in the emission stack, herein referred to as the improved device, we can easily obtain the optimal EL spectra. The buffer layer produces a clear separation of three color regions, blue, green, and red, which eventually generate the white color.

Figure 4 shows a lifetime measurement of a conventional and an improved three-peak white OLED based on a change in luminescence. The projected lifetimes at an initial brightness of 1000 cd/m² are 30,000 hours for the improved three-peak white OLED and 5000 hours for the conventional white emitter. Measurements were taken from an encapsulated device under constant driving current at room temperature. Adopting a buffer layer into the OLED device markedly increases lifetime.

In addition to luminescence degradation, the voltage in the EL diode degrades with time for a constant current. By increasing the operating voltage with time, the drop in voltage in a controlled three-peak OLED device with a buffer layer is reduced (Fig. 5).

The 40-in. display has 1280 x 800 pixels, each of which has individual R, G, B, and W subpixels. We applied a conventional evaporation method to fabricate the pixel arrangement in the 40-in. display. The pixel alignment is well controlled for a pixel pitch of 0.675 mm. This display produces a wide color gamut by optimizing the color-filter structure. The R, G, B, and W luminescence yield CIE coordinates of (0.67, 0.33), (0.25, 0.67), (0.14, 0.12), and (0.31, 0.38), respectively. The color gamut is about 84.8% of the NTSC standard. Another feature of our display is high luminance, a white-peak luminance of more than 600 cd/m², which is favorable for the reliability of the panel. Power consumption during motion-picture operation is typically 80–100 W at a brightness of 300 nits. This represents a better performance than 40-in. AMLCD TVs, which consume about 200 W for the identical operating condition.

 Samsung Electronics Co.

Fig. 5: Voltage shifts vs. operating time for a conventional and a controlled white OLED device.

 Samsung Electronics Co.

Fig. 6: The pixel-to-pixel luminance variation is almost invisible on a 40-in. WXGA AMOLED display.

Figure 6 shows a typical image on an AMOLED display. The pixel-to-pixel luminance variation is almost invisible. To our knowledge, it is the first demonstration of a 40-in. white OLED display having color and image-quality characteristics in a range considered acceptable.

Outlook

In order to impact the TV market, AMOLED TVs must be competitively priced with identically sized AMLCD and PDP TVs. In addition, certain milestones must be achieved, such as 40,000 hours of lifetime, a color gamut exceeding 80% of the NTSC standard, a peak luminance higher than 400 nits, a darkroom contrast ratio higher than 5000:1, and no image burn-in after at least 10 days of continuous operation.

The issues that must be resolved include the attainment of a TFT threshold-voltage change of less than 0.2 V for the lifetime of the display, the elimination of fine mura, and a brightness uniformity of more than 80%. From the viewpoint of power consumption, a combination of energy-efficient EL material and a TFT with a high current output is required in order to yield a power efficiency of more than 5 lm/W. Because an AMOLED is a hold-type display, as is an AMLCD, impulsive or double-frame-rate driving is necessary to reduce motion blur. A metal with low resistivity, such as aluminum or copper, is essential in overcoming the non-uniformity problem as a result of resistive voltage drops in the driving matrix. To be cost-competitive, AMOLEDs must maintain a production yield comparable to that of AMLCDs, and the TFT backplane should be produced by using a process consisting of four or five photo-masking steps.

The above-mentioned difficulties have to be resolved in order for AMOLEDs to be competitive with LCDs and PDPs in the large-sized TV market. Technological breakthroughs, a large amount of investment capital, and the creation of a supply-chain infrastructure are necessary. With the current pace of innovation and technological development, we anticipate that large-sized AMOLED TVs will enter the FPD TV market in about 2–3 years and dominate the TV market in the very near future. •