TWI853179B - Stacked oled microdisplay with low-voltage silicon backplane - Google Patents

Abstract

A microdisplay comprising a light emitting OLED stack on top of a silicon-based backplane with individually addressable pixels and control circuitry wherein the light emitting OLED stack has three or more OLED units between a top electrode and a bottom electrode; and the control circuitry of the silicon-based backplane comprises at least two transistors with their channels connected in series between an external power source V _DD, and the bottom electrode of the OLED stack. The light-emitting OLED stack preferably has a Vth of at least 7.5 V or more. The control circuit can include a protection circuit comprised of a p-n diode, preferably a bipolar junction transistor.

Description

Stacked organic light emitting diode microdisplay with low voltage silicon backplane

This application relates to an OLED microdisplay, and more specifically to a stacked OLED microdisplay with a low voltage silicon backplane.

Typically, a microdisplay is less than 2 inches diagonal (about 5 cm), with an ultra-small display size less than 0.25" diagonal. In most cases, the resolution of microdisplays is high and the pixel pitch is typically 5 to 15 microns. When first commercialized in the late 1990s, microdisplays were commonly used in rear projection TVs, head-mounted displays, and digital camera viewfinders. In recent years, devices such as smart watches have taken advantage of the high resolution and low power consumption of these displays. Microdisplays are expected to proliferate over the next few years, with the global market expected to grow at a CAGR of 20%. One of the trends driving this growth will be the increasing adoption of near-eye displays, augmented reality devices, and virtual reality devices, such as head-mounted displays (HMDs), head-up displays (HUDs), and electronic viewfinders (EVFs).

There are two main categories of microdisplays. The first is a projection microdisplay, which involves projecting a highly magnified image onto a surface. Types of projection microdisplays include rear projection TV and compact data projectors. The second is a near-eye display (NED), which consists of a highly magnified virtual image viewed through an eyepiece (e.g., a virtual reality headset or camera viewfinder). These displays are increasingly used in HMDs and HUDs, especially in the military and medical industries.

Both types of microdisplays offer significant advantages over conventional intuitive displays such as flat-panel LCDs. Advantages of microdisplays over other display types include the ability to produce a large image from a very small, lightweight source display unit, making them easily integrated into space-constrained technologies such as wearables; the ability to create large pixel capacities, allowing for high resolution and clarity; and the ability to achieve greater power efficiency. The higher the resolution and brightness and the lower the power consumption, the better the quality of the microdisplay. However, the challenges facing microdisplay manufacturers are relatively high production costs and the need for high brightness and high contrast as well as long operating lifetimes.

Microdisplays can be made using a variety of display technologies, including liquid crystal on silicon (LCoS), liquid crystal display (LCD), digital micromirror device (DMD), digital light processing (DLP), and more recently, microLEDs (light emitting diodes) and organic light emitting diodes (OLEDs).

LCDs have dominated the microdisplay market in recent years. The advantages of LCD technology are high brightness, relatively low cost, and a relatively simple manufacturing process. The use of LCDs has enabled device manufacturers to reduce the size of microdisplay components over time. LCD displays are currently used in certain HMDs, HUDs, EVFs, and thermal imaging goggles and wearables. However, LCD microdisplays require a light source or backlight together with a liquid crystal array for modulating the light to form an image. This technology has limitations, such as polarization, color space, maximum illumination limitations, LC temperature sensitivity, viewing angles, LCD transmission and extinction ratios, system size limitations, and others, which do not provide all of the desired performance characteristics.

Microdisplays based on microLED technology can have advantages over LCD microdisplays, such as self-emission, a larger color gamut, wide viewing angles, better contrast, faster refresh rates, lower power consumption (image related) and a wide operating temperature range. Currently, microLED microdisplays are based on a standard gallium nitride (GaN) wafer used for standard LEDs. This approach can provide high-illuminance display devices without lifetime issues at a relatively low price. Typically, a standard GaN wafer is patterned into a microLED array. The microLED display is then produced by integrating the microLED array with transistors. However, this approach presents several manufacturing concerns due to color and illumination variations between individual micro-LEDs, including monolithic formation of the micro-LEDs on transistors, pixel spacing, color rendering, and spatial uniformity.

OLED technology shares many of the attractive features of microLED technology for microdisplays. It is self-luminous, has excellent image quality, is very efficient compared to LCD or LCoS and has a very high color rendering and wide color space. An important advantage of self-luminous OLED devices over backlit devices such as LCDs is that each pixel produces only the intensity required for the image, while the backlight pixels produce maximum intensity and then absorb the unwanted light. In addition, since the OLED layers can be vacuum deposited or coated directly on the transistor backplane, it is easier and less expensive to form an OLED on top of the transistor than a microLED. On the other hand, OLEDs can have limited illumination and lifetime.

Resolving the motion blur problem is also important for the control circuitry in OLED microdisplays, which are sample-and-hold type displays (see “Why Do Some OLEDs Have Motion Blur?” at https://www.blurbusters.com/faq/oled-motion-blur/, December 28, 2018, and “Is Motion Resolution an Issue with OLED TVs” at https://www.soundandvision.com/content/motion-resolution-issue-oled-tvs, January 15, 2015).

The only way to reduce the motion blur caused by sample-and-hold is to shorten the amount of time a frame is displayed. This can be done by using extra refreshes (higher Hz) or via dark periods between refreshes (flicker). For OLED microdisplays, the best solution is to "turn" the displayed image "on and off" by turning off the entire active area at the same time or by a "scrolling" technique where only a portion of the displayed image is turned off at a time in sequence. The "scrolling" technique is preferred. The pixel off time is very short and well below the threshold for the human eye to avoid perceptible flicker. This is accomplished by including a switch transistor in the control circuitry that, when activated via a select line, prevents current from flowing through the OLED and switches emission by turning the OLED pixel "off" for the desired time period. In other words, a switch transistor is a switch transistor in that it simply turns the pixel "on" or "off" without regulating voltage or current. However, this solution of turning the pixel off for a portion of an image display time (commonly referred to as the frame time) simply increases the need for increased illumination of the OLED each time it is "on" because this is the average illumination perceived by the eye within the frame. Switching to reduce motion blur can be applied to any method of powering the OLED stack, such as current control or PWM.

OLED microdisplays using silicon backplanes are very attractive from a cost and manufacturability perspective. For example, see Ali et al., “Recent advances in small molecule OLED-on-Silicon microdisplays”, Proc. of SPIE Vol. 7415 74150Q-1, 2006; Jang et al., J. Information Display, 20(1), 1-8 (2019); Fujii et al., 4032ppi High-Resolution OLED Microdisplay”, SID 2018 DIGEST, p. 613; US2019/0259337; Prache , Displays, 22(2), 49 (2001); Vogel et al., 2018 ^48th European Solid-State Device Research Conference, p. 90, Sept. 2018; and Wartenberg et al., “High Frame-Rate 1" WUXGA OLED Microdisplay and Advanced Free-Form Optics for Ultra-Compact VR Headsets”, SID Proceedings, 49 (1), pp. 40–5, 514 (2018).

Microdisplays require very high illumination to be useful in all environmental conditions, such as outdoors in bright sunlight. Even in controlled environmental conditions, such as in VR headsets, very high illumination is required to create an immersive visual experience. Very high illumination from the display allows the use of less efficient optics that are smaller, lighter, and less expensive, resulting in a more competitive headset. Currently, state-of-the-art OLED microdisplays do not provide the desired illumination.

For example, a press release from one maker of tandem OLED microdisplays states that it may be possible to deliver full-color products at 2.5k nits, but acknowledges that 5k nits would be a more desirable goal (see https://www.kopin.com/kopin-to-showcase-latest-advances-in-its-lightning-oled-microdisplay-line-up-at-ces-2020/, dated Jan 7, 2020). Some manufacturers have proposed targets of 10k nits or more (see https://hdguru.com/calibration-expert-is-10000-nits-of-brightness-enough/, dated Jul 26, 2018). A recent press release on June 20, 2020 (https://www.businesswire.com/news/home/20200630005205/en/Kopin-Announces-Breakthrough-ColorMax%E2%84%A2-Technology-Unparalleled-Color) describes a tandem (2-stacked) OLED display emitting >1000 nits. It was also announced that "further improvements in brightness (>2000 nits) and color fidelity are expected by optimizing OLED deposition conditions. By including a structure for enhanced output coupling efficiency, the brightness of OLED microdisplays could be increased to >5000 nits within a few years."

One solution to increasing the total amount of light emitted by an OLED device is to stack multiple OLED cells on top of each other so that the total light emitted from the stack is the sum of the light emitted by each individual cell. However, while the total light emitted from such an OLED stack increases based on the total number of individual OLED light-emitting cells, the voltage required to drive the OLED stack also increases based on the voltage required to drive each individual OLED cell. For example, if a single light-emitting OLED cell requires 3V to produce 250 nits at a given current, then a stack of two such cells at the same current will require 6V to deliver 500 nits, a stack of 3 cells will require 9V to deliver 750 nits, and so on.

OLED stacks are well known; for example, US7273663, US9379346, US9741957, US 9281487, and US2020/0013978 all describe OLED stacks of multiple stacks of light-emitting OLED cells, each separated by an intermediate connecting layer or charge generating layer. Springer et al., Optics Express, 24 (24), 28131 (2016) report OLED stacks with 2 and 3 light-emitting cells, each of which has a different color. OLED stacks with up to six light-emitting cells have been reported (Spindler et al., "High Brightness OLED Lighting" at SID Display Week 2016, San Francisco, California, May 23-27, 2016).

However, this multi-stacked OLED approach, which would require relatively high drive voltages, is difficult to apply to microdisplay applications. One problem is that microdisplays also need to have high resolution, which requires that the size of the individual pixels must be as small as possible and that the active (light-emitting) area of the microdisplay contain as many pixels as possible. This requires that the transistors in the control circuitry of the backplane be small, but large enough to handle the required voltages and currents without causing permanent damage or current leakage.

Typically, as transistors get smaller, the voltage rating gets lower because they cannot handle higher powers due to leakage current and other failure mechanisms. Smaller, lower voltage transistors have thinner insulation at the gate, so they also have more quiescent current leakage. Therefore, OLED devices that require higher voltages typically use larger transistors that can handle the higher voltages. WO2008/057372 discusses the problems and prior art associated with the reduction in pixel circuit size in microdisplays. For example, see also O. Prache , Journal of the Society for Information Display, 10(2), 133 (2002); O. Prache , "Active Matrix Molecular OLED Microdisplays, Displays, 22, 49-56 (2001); and Howard et al., "Microdisplays based upon organic lighat emitting diodes", IBM J. of Res. & Dev., 45 (1), 15 (2001) for discussing the need for microdisplays on silicon backplanes that provide high illumination and a large contrast ratio at low voltages.

Furthermore, when using a MOSFET p-channel transistor to provide a constant current from a power supply at _VDD to an OLED with a cathode voltage of _VCATHODE , the total voltage must be large to power the transistor and turn the OLED "on" to high brightness. However, if a low voltage p-channel transistor is used to control a 12V OLED, then when trying to turn the OLED off to form a black pixel, the current leakage from the transistor will be high and the OLED will continue to emit because ( _Vanode - _Vcathode ) will still be greater than the OLED threshold voltage. In an OLED microdisplay, current leakage through the drive transistor will reduce contrast because the OLED pixel will continue to emit light when it should be dark. This effect will make pure black (no emission) appear gray and reduce the magnitude of the tonal scale between pure black and pure white. This is undesirable.

There is a need to increase the performance of OLED microdisplays on silicon backplanes by utilizing OLED stacks with high illumination and with high voltage requirements to emit light. However, the control circuitry on the silicon backplane must be able to handle the increased voltage and current requirements without a significant increase in size to maintain the resolution and pixel pitch within the active area of the OLED. Specifically, the control circuitry should maintain contrast by preventing or minimizing current leakage through the TFT. Contrast is the difference in light emission between when a pixel should be "off," "black," or not emitting (typically, image signal code value (CV) = 0) and when the pixel should be fully "on," "white," or at maximum emission (typically, image signal CV = 255).

Typically, in the semiconductor foundry industry that manufactures backplanes, analog transistors with an operating range of 5 V or less are considered standard "low voltage" (LV) transistors. It is also common that the voltage ratings usually have a 10% safety limit, allowing reliable operation without degradation of the life of "5 V transistors" or even 5.5 V transistors; this is high enough to allow for a certain degree of overvoltage in the OLED dynamic voltage range and additional load voltage for the driver circuit. Although voltage limits generally apply between any pair of contacts to a transistor (gate, source, drain, body (also called bulk or well)), they specifically apply to the maximum gate-drain voltage so that the performance of the transistor under these conditions is typically within the 43,000 hour specification. Sometimes, depending on the design of the transistor, the voltage limits of other contact pairs can be higher (for example, 7 V) but the transistor is still called an LV or 5 V transistor. 5 V analog transistors are widely available in the industry because they are compatible with traditional TTL logic voltage levels for communication between integrated circuit (IC) chips. As voltages for input-output communications continue to drop (e.g., 3.3 V and 1.8 V standards), these 5 V transistors are sometimes referred to as medium voltage (MV) transistors, making the LV designation a newer "low-voltage" analog transistor. Although relative designations such as LV and MV can change over time, in this patent application, the term LV or "low voltage" refers to a transistor rated at 5 V or less, and the term MV or "medium voltage" refers to transistors with voltage ratings greater than 5 V. Higher voltage analog transistors are also commonly used, but the exact voltage is not standardized within the IC manufacturing industry like the 5 V transistors. For example, higher voltage transistors are often required in industries such as automotive.

Currently, silicon backplanes with low voltage 5 V drive transistors are available, which use tandem (two light-emitting OLED units separated by a CGL) OLED stacks to emit light. For example, see Cho et al., Journal of Information Display, 20(4), 249-255, 2019; https://www.ravepubs.com/oled-silicon-come-new-joint-venture/, published 2018; Xiao, "Recent Developments in Tandem White Organic Light-Emitting Diodes", Molecules, 24, 151 (2019). The illumination of these examples is not enough to meet the technical requirements.

Han et al., “Advanced Technologies for Large-Sized OLED Displays,” Chapter 3, 10.5772/intechopen.74869 (2018). This review article describes the triple-stacked white OLED scheme and advances in backplane technology, including technology with two series-connected transistors, but separately rather than in combination. The reference also notes that this two-transistor backplane “is difficult to use for large, high-resolution panels due to large line loads and short charging times” and therefore uses a different type of backplane circuitry in their device.

Liu et al., J. Cent. South Univ., 19, 1276−1282 (2012) disclose a 3T1C circuit with two series-connected p-channel transistors in an OLED on a Si chip microdisplay.

Zeng et al., “A Novel Pixel Circuit with Threshold Voltage Variation Compensation in Three-Dimensional AMOLED on Silicon Microdisplays,” P-27, SID 2019 Digest, p. 1313, describes a 4T1C pixel circuit for driving an AMOLED display on a silicon backplane. It discloses the use of two p-channel low-voltage transistors connected in series between a power source and an OLED. The first transistor is used to drive the OLED and control the current supplied to the OLED. The second transistor is a switching transistor that turns off the OLED during program operation. The above circuit also contains additional transistors to determine and compensate for _Vth variations.

US9,066,379 describes a control circuit system suitable for an OLED microdisplay. It discloses the use of two p-channel low voltage transistors connected in series between a power source and an OLED. The first transistor is used to drive the OLED and control the current supplied to the OLED. The second transistor is a switching transistor used to turn the OLED on and off. The circuit also requires a third transistor, which is a medium voltage or high voltage p-channel transistor located between the switching transistor and the OLED. The purpose of this third transistor is to prevent current from flowing to the OLED whenever the voltage ( _Vss - _Vcathode ) is greater than the OLED critical voltage during the "off" cycle of the OLED.

Pashmineh et al., “High-voltage circuits for power management on 65nm CMOS”, Adv. Radio Sci., 13, 109–120, 2015, describe high-voltage circuits based on stacked low-voltage CMOS transistors connected in series (also commonly referred to as “stacked” transistors).

Dawson et al., "The Impact of the Transient Response of Organic Light Emitting Diodes on the Design of Active Matrix OLED Displays", International Electronic Devices Mtg 1998, 875-878, describes a pixel circuit having two p-channel transistors connected in series.

Kwak et al., "Organic Light-Emitting Diode-on-Silicon Pixel Circuit Using the Source Follower Structure with Active Load for Microdisplays", Japanese Journal of Applied Physics, 50, 03CC05 (2011) describes a pixel circuit having three series-connected p-channel transistors and an overvoltage protection circuit. This reference states that since the operating voltage of an OLED is higher than the operating voltage of a MOSFET, an "overvoltage protection circuit" is required to prevent the metal oxide semiconductor field effect transistor (MOSFET) from collapsing. U5760477 and US9066379 also describe the use of a protection circuit.

Vogel et al., SID 2017 DIGEST article 77-1, pages 1125 to 1128, disclose the use of a protection circuit in a low voltage OLED Si microdisplay to extend the OLED voltage operating range.

Other references for overvoltage protection of OLEDs are disclosed in US6580657, WO2009072205 and CN200488960. US9059123, US9299817, US9489886, US20080316659 and US20200202793 disclose the use of n-p junction diodes, such as bipolar junction transistors, in pixel control circuits of OLED displays.

Additional patent references disclosing two series-connected transistors include: CN109166525, US20140125717, US7196682, US6229506, US9262960, US7443367, US6930680, US10614758, WO2019019590, US6946801, US7755585, US8547372, US8786591, US8797314, US9818344, US2018/0211592, and US10269296. The following references disclose pixel circuits having three series-connected transistors: US9324266, US9384692, US10600366, US7180486, US9858863, and US20190279567.

US9059123 and US9489886 disclose the use of n-p junction diodes, such as bipolar junction transistors, in pixel control circuits of OLED displays.

A microdisplay is described that includes a light-emitting OLED stack located on top of a silicon-based backplane, the silicon-based backplane having individually addressable pixels and a control circuit system, wherein the light-emitting OLED stack has three or more OLED units located between a top electrode and a bottom electrode; and the control circuit system of the silicon-based backplane includes at least two transistors for each individually addressable pixel, and the channels of the at least two transistors are connected in series between an external power supply _VDD and the bottom electrode of the OLED stack.

In the microdisplay as above, the threshold voltage Vth of the light-emitting OLED stack _is at least 7.5 V or greater, or preferably at least 10 V or greater.

As in any of the above microdisplays, wherein the OLED stack comprises four or more OLED light-emitting units.

As in any of the above microdisplays, wherein the OLED light emitting units are each separated from each other by a charge generation layer (CGL), or wherein the bottom electrode is segmented and each segment is electrically contacted with the control circuit system in the backplane, or wherein the OLED stack is top emitting.

A microdisplay as in any of the above, wherein the OLED stack forms a microcavity in which the physical distance between the segmented bottom electrode and the top electrode is constant across all pixels.

A microdisplay as in any of the above, wherein the transistors having their equal-channel series connection are both rated at 5 V or less, or wherein the transistor closest to the power supply is a drive transistor and is rated at 5 V or less, and the transistor closest to the bottom electrode of the OLED is a switching transistor and is rated greater than 5 V.

As in any of the above microdisplays, wherein the two transistors having their equal-channel series connection are both p-channel transistors, or wherein the two transistors having their equal-channel series connection are each located in a separate well.

As in any of the above microdisplays, wherein the control circuit system further includes a protection circuit having a p-channel transistor or diode. The protection circuit may include a pn diode, and wherein the cathode of the pn junction diode is connected to the node of the bottom electrode of the OLED stack, and the anode is connected to a voltage reference V _REF or a current reference I _REF .

As described above, the protection circuit, wherein the diode is a bipolar junction transistor. The bipolar junction transistor may be an NPN transistor, wherein the base is connected to a voltage source V _PROTECT or a current source I _PROTECT , the emitter is connected to a node connected to the bottom electrode of the OLED stack, and the collector is connected to the voltage source V _DD . The base of the BJT may be isolated.

As in the above protection circuit, the bipolar junction transistor is located in a well separated from the two transistors having their equal-channel series connection, or the two transistors having their equal-channel series connection are both p-channel transistors and each is located in a separate n-well and the bipolar junction transistor is an NPN transistor located in a separate p-well.

Any of the microdisplays as described above, further comprising an RGB color filter array (CFA) located on top of the OLED stack, wherein the individual red filters, green filters, blue filters are aligned with the segmented bottom electrode to form R, G, B pixels; or it further comprises an RGBW color filter array (CFA) located on top of the OLED stack, wherein the individual red filters, green filters, blue filters and transparent or colorless filters are aligned with the segmented bottom electrode to form R, G, B and W pixels.

Microdisplays provide very high illumination and contrast with a small pixel pitch size.

Cross-reference to related applications

This application claims priority to U.S. provisional U.S. application 62/966,757, filed on January 28, 2020, with attorney docket number OLWK-0021-USP, and entitled “STACKED OLED MICRODISPLAY WITH LOW-VOLTAGE SILICON BACKPLANE” and U.S. provisional U.S. application 63/054387, filed on July 21, 2020, with attorney docket number OLWK-0021-USP2, also entitled “STACKED OLED MICRODISPLAY WITH LOW-VOLTAGE SILICON BACKPLANE”.

For the purposes of the present invention, the term "on" or "above" means that the structure in question is located above another structure, that is, on a side opposite to the substrate. "Top", "uppermost" or "upper" refers to a side or surface farthest from the substrate, while "bottom", "bottommost" or "bottom" refers to a side or surface closest to the substrate. Unless otherwise specified, "located above..." should be interpreted as two structures can be directly in contact or there can be an intermediate layer between the two structures. When referring to a "layer", it should be understood that a single layer has two sides or two surfaces (topmost and bottommost); in some examples, "layer" can mean multiple layers considered as a whole and is not limited to a single layer.

With respect to light emitting units or layers, R indicates a layer that primarily emits red light (>600 nm, desirably in the range of 620 to 660 nm), G indicates a layer that primarily emits green light (500 to 600 nm, desirably in the range of 540 to 565 nm), and B indicates a layer that primarily emits blue light (<500 nm, desirably in the range of 440 to 485 nm). It is important to note that the R, G, and B layers may produce some degree of light outside the indicated range, but the amount is always less than the primary colors. Y (yellow) indicates a layer that emits a lot of R and G light, but a much smaller amount of B light. "LEL" means light emitting layer. Unless otherwise indicated, wavelengths are vacuum values, not in situ values.

An individual OLED light emitting cell can produce a single "color" of light (i.e., R, G, B, or a combination of 2 or more primary colors such as Y, C (cyan), or W (white)). A single color of light can be produced within the OLED cell by a single layer or multiple layers of the same color or multiple emitters, each layer having the same or different emitters emitting primarily in the same color. A single OLED cell can also provide a combination of two colors (i.e., R+G, R+B, G+B) within a single OLED cell by having a layer of a single emitter emitting two colors of light, a layer with two different emitters, or a combination of multiple separate layers that each emit one but different color. A single OLED cell can also provide white light (a combination of R, G, and B) by having a layer that emits light of all three colors or a combination of individual layers that each emit a single (but different) color, the sum of which is white. Individual OLED light-emitting cells can have a single light-emitting layer or can have more than one light-emitting layer (directly adjacent to each other or separated from each other by an intermediate layer). Individual light-emitting cells can also contain various non-emissive layers, such as hole transport layers, electron transport layers, blocking layers, and other layers known in the art to provide desired effects (e.g., promoting emission and managing charge transfer across the light-emitting cell).

Because OLED light-emitting cells can include multiple layers, an individual cell is sometimes referred to as a "stack," which can be confused with an OLED device having multiple cells. In this application, a "stacked" OLED has at least two OLED light-emitting cells stacked on top of each other on a substrate so that there are multiple light sources within the device. In the stacked OLED of the present invention, individual OLED light-emitting cells are separated from each other by a charge generation layer (CGL) rather than by separate and independently controlled intermediate electrodes. In order for an OLED light-emitting cell to be considered, it must be separated from another light-generating cell by a CGL. Therefore, a light-emitting layer that is adjacent to one of the OLED light-emitting cells but is not separated from the OLED light-emitting cell by a CGL is not considered as an individual cell. Within a stack, all or some of the individual OLED light emitting units may be the same or all may be different from one another. Within an OLED stack, the individual OLED light emitting units may be placed between the top cathode and the bottom cathode in any order. Stacked OLEDs may be monochromatic (each pixel of the OLED stack emits primarily the same color of light, such as green light) or may have multi-mode emission (where each pixel emits 2 or more colors of light (e.g., yellow or white) or where different pixels emit different colors of light so that the overall emission contains 2 or more colors of light).

In some cases, the threshold voltage ( _Vth ) of the OLED stack can be estimated by straight-line extrapolation of the IV curve after significant light emission begins to return to the voltage axis. Since this method is not accurate due to the fact that the IV response curve of an OLED is not completely linear over its response range, the value calculated in this way is not accurate. One approximate range for this measure is +/-10%. More precisely, the threshold voltage can be defined as the voltage at which the current density exposing the anode layer does not exceed 0.2 mA/ ^cm2 and there is at least some reliably detectable illumination (i.e., at least 5 cd/A). This is the method used in this application.

Hereinafter, a transistor may be referred to as being "on" or "off." In an "off" transistor, it is expected that a signal is sent to the gate so that no current will pass through the terminal; in other words, the signal (usually, CV = 0) indicates that the signal requires no current to pass through the transistor so that instead of turning the transistor off, the transistor is regulated to be "off." In such cases, even though a transistor may be "off," there may still be some current leakage. Likewise, in an "on" transistor, it is expected that a signal will be sent to the gate so that at least some current will pass through the terminal; in other words, the signal (usually, CV = greater than 0 but less than 255) indicates that the image requires the pixel to emit to a certain degree so that instead of turning the transistor "on", the transistor is modulated to be "on." Similarly, the pixel or OLED can be said to be "on" or "off" depending on the requirements of the image, and the appropriate signal is sent to the pixel or OLED accordingly.

The silicon backplane is derived from a silicon wafer (also called a chip or substrate). It is a semiconductor thin sheet, such as a crystalline silicon (c-Si), used to make integrated circuits. The wafer serves as the substrate for the microelectronic devices placed in and on the wafer. It undergoes a number of microfabrication processes, such as doping of various materials, ion implantation, etching, thin film deposition, and photolithographic patterning. Finally, the individual microcircuits are separated and packaged into an integrated circuit by wafer dicing. Wafers are grown from crystals with a regular crystal structure, in which the silicon is in a rhombic cubic structure with a lattice spacing. When cut into wafers, the surfaces are aligned in one of several relative directions, which are called crystal orientations. Silicon wafers are typically not 100% pure silicon, but are initially formed with certain doping concentrations of boron, phosphorus, arsenic or antimony, which are added to the melt and define the wafer as either bulk n-type or p-type. See Chapter 7 in "Flat Panel Display Manufacturing", Souk, L., Ed., 2018 for internal reference. The silicon backplane is expected to be a single crystal Si wafer.

To provide control circuitry for the operation of the stacked OLED, thin film transistors (TFTs) and other components (e.g., capacitors, resistors, connecting wiring or bus bars, etc.) are disposed on the surface of the silicon wafer. For example, see T. Arai, "High Performance TFT Technologies for the AM-OLED Display manufacturing", Thesis, Nara Institute of Science and Technology, 2016, M.K. Han, Proc. of ASID '06, 8-12 Oct, New Delhi, US9066379, and US10163998. It should be understood that the TFT may or may not include a silicon wafer as part of the TFT structure, or may be made of a separate material deposited on the surface.

TFTs can be made using a variety of semiconductor materials. The characteristics of a silicon-based TFT depend on the crystalline state of the silicon; that is, the semiconductor layer can be amorphous silicon, microcrystalline silicon, or the semiconductor layer can be annealed into polycrystalline silicon (including low temperature polysilicon (LTPS) and laser annealing).

The fabrication of silicon backplanes with suitable control circuitry is a well-known, understood, and predictable technology. However, due to the cost and complexity of the manufacturing process and equipment, it is often impractical to build a facility to manufacture a specific backplane. Instead, a foundry model is widely adopted in the industry, in which the functional characteristics of microelectronic devices have become more standardized. This standardization allows the separation of design and manufacturing. A design that follows appropriate design rules can be easier and cheaper to manufacture by different companies with compatible manufacturing methods. Therefore, the control circuitry on the silicon backplane is usually limited to the use of standard components selected from a variety of options provided by the backplane manufacturer. For example, a silicon backplane manufacturer may offer the option of including various standard transistor designs rated for 1.8 V, 2.5 V, 3.3 V, 5 V, 8 V and 12 V in a customer design, but will not be able to offer (without significant cost) transistors that are not included in the offered design.

For the purposes of this application, "low voltage" (LV) is defined as those analog microelectronic components that are sized, designed, and rated to operate safely and reliably at 5 V or less. "High voltage" (HV) microelectronic devices are generally considered to be in the range of 18 V to 25 V. "Medium voltage" (MV) microelectronic devices are generally considered to be microelectronic devices between LV and HV. It should be noted that these voltage ratings are set by the manufacturer and that the manufacturer does not recommend exceeding the set maximum voltage for each transistor.

Complementary metal oxide semiconductor (CMOS) technology uses p-type and n-type metal oxide semiconductor field effect transistors (MOSFETs) to realize complex integrated circuits. Depending on the manufacturing process, different voltage domains can be used (i.e., 1.8 V, 2.5 V, 3.3 V, 5 V, 12 V, etc.). In all voltage domains, the MOSFET transistor has a drain, source, gate, and bulk/well. The base of the MOSFET is the substrate; for an n-channel FET, the substrate is a p-type doped substrate or well with a low doping rate; for a p-channel FET, the substrate is an n-type doped well with a low doping rate. The source and drain regions are formed by highly doped regions of either n-type or p-type for n-channel FETs or p-channel FETs. The controlled channel is formed between the source and drain, isolated with a thin oxide and usually covered with a layer of polysilicon to serve as a gate. All four terminals of the FET (source, drain, gate, substrate/well) are connected by metal contacts to a metal interconnect layer, which is ultimately connected to the OLED.

The emission area of a microdisplay is small and to achieve the necessary pixel pitch, the space for the control circuitry of each pixel is limited. For a full-color microdisplay, the space occupied by the control circuitry of each individual pixel should not exceed 100 square microns and preferably not exceed 50 square microns. For a monochrome microdisplay where all pixels emit the same color, the space for the control circuitry can be 3 to 4 times larger because fewer pixels are required.

In a suitable low voltage 5 V transistor, the maximum voltage between any pair of terminals of all terminals must not exceed 5 V to avoid damage to the device. A typical safety margin is considered acceptable for a 10% overvoltage over a short period of time. A medium voltage transistor rated for greater than 5 V (e.g., a medium voltage transistor rated for 7.5 V) has roughly the same setup as a 5 V transistor, but with a thicker gate oxide and larger geometry (channel width and length) to withstand the higher voltage. Therefore, an MV transistor will generally be larger and take up more space than a corresponding LV transistor.

Although transistors can be made in any size range regardless of voltage rating, a low voltage MOS transistor rated for 5 V suitable for microdisplay applications has a total area of no more than 20 square microns and preferably no more than 10 square microns. A channel area (channel length x channel width) of a 5 V transistor suitable for microdisplay applications should be no more than 1 square micron and preferably no more than 0.30 square microns. The two transistor contacts should each be no more than 1 square micron and preferably no more than 0.30 square microns.

For the purposes of the application, a BJT suitable for the protection circuit is a vertical one with a common configuration for both NPN and PNP types. For an NPN BJT, the collector is formed as a low doped deep n-well with a common silicon substrate (bulk) with low p doping. The base is formed as a p-well inside the deep n-well and connected by a highly doped p-region. The emitter of the BJT is formed by a highly doped n-region inside the p-well. The typical size of the emitter region is about 500 nm x 500 nm and preferably does not exceed 0.30 square microns to make the pixel size very small. The maximum voltage between any pair of terminals of all the terminals of the BJT (bulk, base, emitter, collector) must not exceed 5 V to avoid damage to the device. A typical safety margin of 10% overvoltage is acceptable for a short period of time.

OLED microdisplays (often referred to as "AMOLEDs") consist of an active matrix of OLED pixels that generate light (luminescence) when electrically activated, deposited or integrated onto a transistor array on a silicon chip, where the array acts as a series of switches to control the current flowing to each individual pixel. Typically, this continuous current flow is controlled by at least two transistors at each pixel (to trigger luminescence), one of which starts and stops charging a storage capacitor and controls another transistor, where the other transistor provides a voltage source at the level required to create a constant current flowing to the pixel. In some embodiments, a transistor (often called a scan transistor or select transistor) that controls the current provided by another transistor (often called a drive transistor) is controlled by a select line that is common to all pixels in a row. The signal delivered by the scan transistor is supplied by a data line that is common to all pixels in a row.

This is illustrated in Figure 1, which shows the simplest form of a prior art AMOLED pixel design. The simplest AMOLED pixel with pixel memory uses two transistors and a capacitor. The current drive transistor MP2 is typically connected to the anode of the OLED from the supply voltage _VDD . One transistor (MP2) drives the current for the OLED, and the other transistor MP1 (also called a scanning transistor) is used as a switch to sample a voltage and hold it on the storage capacitor C1 shown. A data line (supply _VDATA ) controls the current _VDD through the drive transistor MP2. A select line controls MP1 and therefore controls the charging of capacitor C1. Typically, transistors have intrinsic capacitance, so based on the intrinsic capacitance of the transistor and the leakage current through the transistor, no additional capacitance may be required.

In the figures following FIG. 1 , any capacitor present is omitted from the figures for clarity. The presence of a capacitor is optional and the circuit may have no capacitors or any number of capacitors as desired. The reference voltage for the capacitor may be V _DD or some other voltage.

Note that variations in the threshold voltage ( _Vth ), carrier mobility or series resistance will directly affect the uniformity of the OLED current driving the TFT and therefore the brightness of the image display. One major factor affecting non-uniform current is variations in the threshold voltage ( _Vth ) of the OLED drive transistors. The pixel may require control circuitry to make other types of compensation, such as aging, degradation or burn-in of the OLED material over time, unevenness or non-uniformity over the active area or voltage drops in metal interconnects. In addition, the control circuitry provided by the TFT may need to control the timing of the current delivered to the pixel, such as by PWM. The design of control circuit systems for OLEDs including various types of compensation and driving schemes has become a topic of much attention and many methods have been proposed. These compensation circuits may also be present in control circuit systems for stacked OLEDs having three or more stacks.

It was unexpectedly discovered that the high voltage/current required for an OLED stack having three or more stacks can be handled by a control circuit comprising at least two transistors having their channels connected in series. This is sometimes referred to as a "stacked transistor". Desirably, one of the series-connected transistors is a low voltage drive transistor and the other is a switching transistor. This configuration allows the OLED pixel to be driven without significant current leakage from the TFT, so that high brightness is obtained without loss of contrast.

Specifically, the driving TFT and the switching transistor are connected in series so that a first terminal of the driving transistor is electrically connected to (and closer to) an external power source, a second terminal of the driving transistor is electrically connected to a first terminal of the switching transistor, and a second terminal of the switching transistor is electrically connected to (and closer to) the bottom electrode of the OLED stack. This configuration allows the OLED stack to operate at a higher voltage than individual transistors designed without significant current leakage or physical damage to the low voltage transistors.

This basic control circuit configuration is shown in FIG2 , where a p-channel driver transistor T1 is connected to V _DD (external power supply) at a first terminal (source in the case of a p-channel transistor) and to a switch transistor T2 at a second terminal (drain in the case of a p-channel transistor). The gate of the driver transistor T1 is controlled via the data line and a series select transistor T3, the gate of which is controlled via a select line SELECT1. The switch transistor T2 receives current from the second terminal of T1 at a first terminal (source) (when T1 is turned on) and is connected to the bottom electrode of the OLED at a second terminal (drain). The gate of the switch transistor T2 is directly controlled by the select line SELECT2. The driving transistor T1 and the switching transistor T2 are connected in series. When T1 and T2 are selected to be "on", current flows to the bottom electrode of the OLED stack and the OLED stack emits light. If T2 is selected to be "off", no current flows to the OLED stack regardless of whether T1 is "on" or "off". In some embodiments, the switching transistor T2 can provide an on-off function to prevent motion blur. If T1 is "off", no matter whether T2 is "on" or "off", the OLED will not emit light.

In Figure 2, the control circuitry of the backplane having at least two series connected transistors between _VDD and the bottom electrode of the OLED electrode is a driver transistor and a switch transistor. In this application, a driver transistor has the function of regulating the voltage and current flowing to the OLED pixel to an appropriate level according to the image being displayed, and a switch transistor has the purpose of turning the current flowing to the OLED pixel on and off to provide a switching function. For example, during operation of the microdisplay, the switch transistor T2 interrupts the power flowing from the driver transistor T1 to the bottom electrode of the OLED stack, and therefore does not emit light (black) during a portion of the frame time. This is different from the scanning transistor T3, which is turned on only for a very small fraction of the frame time (e.g., 1/1200 for a display with 1200 rows) to charge a storage capacitor (not shown in FIG. 2 ). Typically, a switching transistor can be smaller in size than a driving transistor at the same maximum operating voltage. However, if the switching transistor needs to be handled more like in the control circuitry of an OLED with three or more stacks, it can be larger than the driving transistor and rated for a higher voltage.

Desirably, the driver transistor (T1) is closer to the power supply _VDD than the switch transistor (T2), which is closer to the bottom electrode of the OLED. Preferably at least the driver transistor is a p-channel MOSFET transistor, and more preferably, both the driver transistor and the switch transistor are p-channel MOSFET transistors. The two series-connected transistors can be both LV, both MV, or one LV transistor and one MV transistor. Preferably they are both LV transistors to reduce oversizing of the pixel circuit. However, in some embodiments, the driver transistor is LV and the switch transistor is MV.

Although only p-channel transistors are illustrated in FIG. 2 , n-channel transistors or a mixture of n-channel transistors and one of the p-channel transistors may be used. In such cases, the circuit system will need to be appropriately reconfigured in view of the polarity difference between the n-channel transistor and the p-channel transistor. In addition, other microelectronic components such as other transistors, capacitors, and resistors may be included in this control circuit as desired.

It is known in the art that MOSFET transistors require an intrinsic body diode connection to function as desired. Due to the structure of a MOSFET transistor, a parasitic diode is inherently present and can affect the operation of the transistor. Typically, the intrinsic body diode is connected to a power supply internally or externally to apply a bias. These body connections are also called "bulk connections" or "transistor wells" and other terms. Such a connection is shown in Figure 3, where IBD1, IBD2 and IBD3 are intrinsic body diodes (for T1, T2 and T3) connected to a separate voltage source _VDD2 . These transistors can share the same well to achieve the intrinsic body diode connection. However, the same power supply V _DD used to power T1 and T2 can also be used for IBD; that is, V _DD and V _DD2 are a common source.

However, it may be desirable to float one or more of the transistors in their separate wells on the Si backplane to avoid exceeding the operating voltage range of any of the components. Specifically, when both the first drive transistor and the switch transistor are p-channel transistors, each transistor may be located in its own separate n-well. This allows a larger dynamic voltage range to achieve control of the OLED than can be achieved if both transistors are located in the same n-well. Series connection of transistors using isolated, floating or different wells is known in the art, see, for example, US9066379, US5764077, US7768299, US9728528 and JP2016200828.

This is illustrated in Figure 4 where T1 occupies its own well, as indicated by the dashed line, and is biased via IBD1 through a connection to one of _VDD2 , and T2 occupies a different well, as indicated by the dashed line, and is biased by a separate connection IBD2 to one of the transistor sources ( _VDD ). In the embodiment of Figure 4, the bias applied to each IBD is different and therefore each n-well is independent of the others.

The embodiments shown in FIGS. 2-4 have several advantages in design and operation for driving an OLED having three or more stacks. In design, the circuit can be very compact because all transistors are relatively small LV transistors, as is common in most foundries. All transistors can be p-channel transistors, and all n-wells are biased to V _DD , which eliminates the need for isolation wells or floating wells. These features can allow very high resolution and very small size microdisplay designs with very compact pixel circuit designs.

FIG. 5A shows a control circuit having three series transistors. The circuit shown in FIG. 5A is similar to the circuit shown in FIG. 2 , but in FIG. 5A a third transistor T4 is connected in series between the power supply V _DD and the source of the drive transistor T1. As shown, the gate of T4 is controlled by the data line and a column select transistor T5, the gate of which is controlled by a select line SELECT3. The lines SELECT2 and SELECT3 in FIG. 5A may be supplied with a signal at the same or different voltages, may be switched at the same time or at different times, or may be left unswitched depending on operating conditions.

Desirably, the added transistor T4 may be a second driver transistor so that between the power supply _VDD and the bottom electrode of the OLED are two driver transistors (T4 and T1) and a switch transistor (T2), all in series. In this case, the signal supplied to the gate of T4 may also be the same as the signal supplied to T1 and may be supplied via T3/SELECT1. Although FIG. 5A shows a design with one data line and two select lines for data control, a similar design with two data lines and one select line may be used.

In embodiments where there are three or more series transistors, the transistors may be a combination of LV transistors and one of MV transistors. Preferably, all of the transistors are LV transistors to reduce the size of the pixel circuit. These multiple transistors may be p-channel transistors, n-channel transistors, or a mixture. Preferably all transistors are p-channel transistors. If there is a mixture of n-channel transistors and one of p-channel transistors, preferably at least one drive transistor is a p-channel transistor and more preferably an LV transistor. The switching transistors are preferably LV transistors; however, when the operating range of the backplane circuit must be expanded while complying with the voltage limitations of the individual transistors, it may be desirable to use one or more MV switching transistors.

In the case where any of the multiple series transistors is of one type, multiple common transistors of the same type can share the same well to reduce the design size. However, to expand the operating range of the backplane circuit while complying with the voltage limitations of the individual transistors, it may be necessary to place transistors of the same type in separate wells.

FIG5B shows a schematic of a control circuit with two drive transistors similar to those shown in FIG5A, where the two drive transistors are controlled by a level shift circuit (LSC). A level shift circuit is a circuit used to convert signals from one logical level or voltage domain to another logical level or voltage domain, often used to resolve voltage incompatibilities between portions of a system. For clarity, the internal components and all connections to the LSC (e.g., _VDD , ground, and other possible input connections) are not shown. In this embodiment, LSC sets the gate voltages of driver transistors T4 and T1 based on signal SELECT1 and the signal from the data line so that the total voltage across T4 and T1 is always approximately equally split between the two driver transistors. As is known, small logic transistors can be used for this function.

OLED type microdisplays typically contain a protection circuit in the MOSFET type control circuitry of the backplane to limit the amount of electricity flowing through the transistors to prevent damage. It is desirable to include a protection circuit in the control circuitry of the backplane of an OLED microdisplay having three or more stacked cells to achieve this goal, since the power required to make these devices emit is relatively high. A protection circuit should maintain or "clamp" the voltage at the bottom electrode of the OLED so that it does not fall below a desired voltage level when the OLED is not emitting. Such protection circuits may also be referred to as "voltage holding" circuits.

To protect the low voltage transistors present in the control circuitry and to stay within the specified operating range of the transistors set by the foundry, the protection circuitry would be expected to maintain a black level current (CV = 0 or pixel "off") at the bottom electrode of the pixel's stacked OLED below 4 µA/cm2, or more desirably at or below 2 µA/cm2 for a 3-cell stacked OLED with a Vth of about 7.5 V. Similar black level currents are expected for 4-cell stacked OLED devices, with a typical Vth of about 10 V.

FIG6 shows a schematic control circuit similar to that shown in FIG2 but with the addition of a protection circuit. In this particular embodiment, the protection circuit includes a p-channel transistor connected on one end to a node between the drain of T2 and the bottom electrode of the OLED and on the other end to a power source (in this example, a reference voltage V _REF (common to all pixels)). The gate of T6 is also connected to the node. Thus, the diode-connected transistor T6 limits the minimum drain voltage of T2 to prevent it from crashing. There may be other electronic components as part of this circuit, which are not shown. This configuration is similar to that described by Kwak et al., but one end of the p-channel transistor is connected to ground instead of V _REF .

However, the problems caused by the higher voltage requirements of using an OLED with three or more cells still exist when using a protection circuit with a MOSFET transistor example shown in Figure 6. In this embodiment, the use of a LV or MV MOSFET transistor (i.e., T6) to achieve protection for the LV drive transistor and switch transistor does not make the pixel design significantly smaller than using a MV drive transistor and switch transistor and eliminating the protection circuit.

FIG. 7 shows a protection circuit similar to the protection circuit shown in FIG. 6 , but in FIG. 7 the p-channel transistor T6 is replaced by a diode D4. Desirably, the diode D4 in the protection circuit is a pn junction diode. A pn diode is a circuit element that allows electricity to flow in one direction but not in the other (opposite) direction. A pn diode can have a forward bias in the direction where current flows easily or a reverse bias where little or no current flows. This pn diode can be formed in a CMOS design in different ways; for example, as a vertical diode with a highly doped p region in an n-well. D4 is connected on one end to a node between the drain of T2 and the bottom electrode of the OLED, and on the other end to a power source 45, which is a reference voltage _VREF or a reference current _IREF .

FIG8 shows a protection circuit similar to the protection circuit shown in FIG7 . In FIG8 , the diode D4 of the protection circuit is specifically BJT1 , a BJT (bipolar junction transistor) with the collector (C) connected to V _DD , the emitter (E) connected to a node between T2 and the bottom electrode of the OLED stack and the base (B) connected to a power source 50 , which is a voltage source V _PROTECT or a current source I _PROTECT . One benefit of using a BJT in the protection circuit is that the current is amplified from the base current to the collector current. If desired, the collector of the BJT can be connected to a separate power source different from V _DD . The power source V _PROTECT or I _PROTECT can be shared by all pixels. It should be noted that V _PROTECT or I _PROTECT may or may not be constant, but may vary depending on whether the OLED is desired to emit or not emit. This may be useful when used with a switching transistor to reduce the durability of the display by switching. In FIG. 8A , BJT1 is an “NPN” type BJT transistor (preferably), but may also be a “PNP” transistor with appropriate changes in design.

For an NPN type BJT, whenever the emitter voltage _VE at the bottom electrode of the OLED is greater than the base voltage _VB and _VB is less than the collector voltage _VC ( _VE > _VB < _VC ), the BJT will be in the off mode and no current will flow. However, whenever the voltage _VE at the bottom electrode of the OLED is less than the voltage _VB and _VB is less than _VC ( _VE < _VB < _VC ), the BJT will be in the forward acting mode. In this mode, the base-emitter junction is forward biased and the base-collector junction is reverse biased, and therefore the collector-emitter current will be approximately proportional to the base current.

Thus, in FIG8A , if _VC of BJT1 is _VDD and _VPROTECT (which is _VB ) is set to be less than _Vth of the OLED, then whenever the voltage _VE at the bottom electrode is above _Vth + _VCATHODE , BJT1 is “off”, but whenever _VE decreases below _Vth + _VCATHODE (i.e., whenever T1 or T2 is “off”), the voltage at the bottom electrode is maintained at approximately _Vth . Thus, the protection circuit is designed to provide sufficient current whenever needed to protect the drive transistor and the switching transistor from having voltages on either side of their terminals that exceed their rated values whenever the cathode voltage decreases (to a more negative voltage) below a certain value. Furthermore, by setting the base voltage B (V _PROTECT ) below the turn-on voltage of the OLED (V _th is higher than V _cathode ), power loss is minimized whenever the voltage delivered to the bottom electrode of the OLED is greater than or equal to V _th .

In certain embodiments utilizing the protection circuit shown in FIG8, the base voltage ( _VB ) of the BJT is isolated from any external power source. That is, there is no electrical connection between the power source _VPROTECT or _IPROTECT and the base of the BJT. The BJT physically has the existing collector and emitter connections shown in FIG8, but the base connection that remains is not connected to any external source. In these cases, _VB is not intentionally maintained at any particular value, nor is any voltage or current intentionally applied to _VB . _VB is allowed to "float" independently of voltages _VC and _VE , _which are part of the action control circuitry _that operates the OLED. However, there may be parasitic current paths within the backplane that bias the base internally under high impedance conditions. Note that in such embodiments, _VB may vary during operation of the OLED due to parasitic currents generated within the circuitry.

In other embodiments, _{the VB} of the BJT may be self-biased (sometimes referred to as having a "base bias"). When the BJT is self-biased, no external control input signal is applied to the base of the BJT, but rather the signal applied to the base of the BJT is set by the value of a constant supply voltage (i.e., _VDD ) and the value of any bias resistors connected to the transistor. One method of achieving self-biasing of the base of the BJT forms a "fixed base bias circuit." In this configuration, the base of the BJT is connected to a constant power supply (i.e., _VDD ) with a single current limiting resistor. This will allow the base current (I _B ) of the BJT to be kept constant at a given value V _DD , and therefore the operating point of the BJT is also kept fixed. Alternatively, a simple voltage divider network can provide the required bias voltage. Note that in such embodiments, the bias voltage and resulting parasitic current within each pixel will respond to the operation of the OLED. This response can provide effective protection for the transistors.

FIG9 is a schematic cross-sectional illustration of the transistor wells of the circuit shown in FIG8. Note that T1, T2, T3 may each be located in a separate but non-isolated or floating n-well connected to _VDD and all separated from the p-well of BJT1 (an NPN BJT). For convenience, the source (s), gate (g) and drain (d) regions are labeled for T1, T2 and T3. Similarly, the emitter (e), base (b) and collector (c) regions are labeled for BJT1. The collector (c) region of BJT1 is shown as being connected to _VDD through a deep n-well, but in some embodiments it may alternatively be connected directly to _VDD .

Figure 9 also indicates IBD5, which is an intrinsic body diode connected to _VDD of the deep n-well of the Si substrate in which all transistors are located, and IBD6, which is an intrinsic body diode connected to ground of the entire Si substrate. It should be noted that the BJT in the protection circuit can be an NPN transistor or a PNP transistor. If the BJT is an NPN transistor, the transistor is located in a p-well; if the BJT is a PNP transistor, it will be located in an n-well.

It is preferred to use a protection circuit comprising a BJT. When used as part of a control circuit comprising at least two transistors having their equal channels connected in series, the pixel size of the BJT transistor design is smaller. In addition, the inherent amplification factor of a forward active mode BJT means that a relatively small parasitic current can produce a larger protection current. Therefore, the current from the reference/protection voltage source is smaller and the voltage drop on the reference voltage interconnect is greatly reduced.

In addition to maintaining a minimum voltage at the bottom electrode below the critical voltage of the OLED, the protection circuit can also be designed to prevent other undesirable effects by including other appropriate circuit components, such as short circuit protection, transient spikes, etc. However, in some examples, when the stacked OLED of the present invention is used, it can still be used to include such known types of protection circuit systems in the pixel circuit.

For many OLED stacks, especially those including three light-emitting cells, it is desirable that there be no intervening transistor in the electrical connection between the second terminal of the switching transistor and the bottom electrode of the OLED in the control circuitry, as shown in FIGS. 2 to 5 . By intervening, it is meant that the current does not pass directly through another transistor in the electrical connection between the switching transistor and the OLED (i.e., one terminal of the transistor is connected to the switching transistor and the other terminal is connected to the OLED so that the current will pass through the intervening transistor at some point during operation of the OLED). There may be other (non-transistor) microelectronic components in series between the switching transistor and the OLED.

There may be a spur (meaning not directly in-line) connection between the switch transistor OLED connection and a first terminal of a microelectronic component (i.e., a non-interfering transistor or diode) where the OLED operating current does not pass directly through the non-interfering transistor or diode. For example, this configuration is illustrated in Figures 4 to 6 where one side of a non-interfering component is connected to the T2-OLED connection and the other side is connected to a power source.

Desirably, the drive transistor is a low voltage (LV) transistor. That is, the drive transistor is designed and sized to operate safely and efficiently at 5 V or less, regardless of or irrespective of the actual load in the circuit. This necessarily helps to minimize the pixel size to maximize the microdisplay resolution. It should be noted that an OLED stack having three or more OLED cells and driven by a drive transistor will typically require voltages in excess of 7.5 V. If desired, two or more series-connected low voltage drive transistors in separate wells (as shown in FIG. 2 c ) may be used to further mitigate the effects of the higher voltage.

The drive transistor is preferably a p-channel thin film transistor (also known as a p-channel MOSFET (metal oxide semiconductor field effect transistor)). The structure, characteristics and preparation of p-channel transistors are well known. When the drive transistor is a p-channel transistor, its source is electrically connected to an external power source and its drain is electrically connected to a first terminal of the switch transistor. The gate of the drive electrode is controlled by a data line that is separate from the power source.

When the OLED stack includes three OLED light-emitting units, it is desirable that the switching transistor is a low voltage transistor. That is, the switching transistor is designed and sized to operate safely and efficiently at 5 V or less, regardless of or without regard to the actual load in the circuit. The current required for OLED operation flows through this transistor and the drive transistor.

However, if the OLED stack includes four or more OLED light-emitting units, the operating voltage can be well over 7.5 V and the switching transistors can be medium voltage (designed for 7.5 V to 12 V, for example) or high voltage (designed for 18 V to 25 V) as desired. Alternatively, two or more series-connected switching transistors, selected together or independently, can be used to mitigate the effects of the higher voltages. The series-connected multiple switching transistors can all be LV or a mix of LV transistors and one of the MV transistors.

It is desirable that the switch transistor is a p-channel transistor. When the switch transistor is a p-channel transistor, its source is connected to the drain of the drive transistor and its drain is electrically connected to the bottom electrode of the OLED. The gate of the switch electrode is controlled by a select line that is separate from the power supply and the data line and the select line that control the drive transistor.

In some embodiments, it is desirable that the driver transistor and the switch transistor are p-channel transistors, wherein both the driver transistor and the switch transistor are low voltage transistors, or the driver transistor is a low voltage transistor and the switch transistor is a medium voltage or high voltage transistor. In some embodiments, it is also desirable that all transistors in the control circuit system are p-channel transistors.

Although the driver transistor and the switch transistor are connected electrically in series, there may be other intervening microelectronic components between the two transistors. The intervening components may be in-line, where the current between the driver transistor and the switch transistor passes directly through the in-line component. There may also be other microelectronic components electrically connected to the connection between the driver transistor and the switch transistor, so that the current flowing from the driver transistor also flows to this additional component as well as the switch transistor.

In the most basic form of operation, a voltage is supplied to the drive transistors from an external power source. The select lines are set to allow the correct data voltage to be applied from the data lines to the gate of the drive transistors, thereby allowing current from the power source to flow through the drive transistors to the switching transistors. The select lines for the switching transistors are set to allow the correct charge to be applied to the switching transistors, thereby allowing current to flow through the switching transistors to the bottom anode of the OLED stack in each individual pixel. This turns the pixel "on" and it emits light according to the current received. When the image being displayed requires an individual pixel to not emit or reduce emission, the data voltage of the drive transistor can be changed to prevent or limit current flow through the driven pixel.

Other components in the control circuit system can be used to control the current supplied by the driving transistor, controlling the current that is intended to power the OLED when the OLED is "on" or the current that leaks through the driving transistor when the OLED is intended to be "off."

To prevent or minimize motion blur by providing a switching function, a switching transistor prevents current from flowing to the OLED regardless of the operation of the drive transistor (which controls the power supplied to its pixel) (specifically, when the drive transistor is "on" (allowing current to flow)). At the appropriate time for the pixel, the switching transistor can be selected so that the OLED pixel is "off" even if the image to be displayed requires the OLED pixel to be "on". This allows all segments or rotating segments of a pixel in the display to be "off" (not emitting) to minimize the perception of motion blur in the microdisplay.

In a microdisplay, power can be delivered from an external power source in the form of a variable current or voltage to the driver transistors to drive the OLED stack and thereby deliver the desired illumination level. This is typically stored during write operations by charging a storage capacitor. The power level can be controlled at the external power source or if the power delivered is constant, the power can be set to the appropriate level by other microelectronic circuits in the backplane. This is called "current control" and is commonly used to power most OLED devices. Alternatively, the power supplied to the OLED stack can be constant and the total amount of light emitted within a set time period (frame) controlled by the time the OLED pixel is fully "on" compared to the time the OLED pixel is "off". This is called pulse width modulation or PWM control.

Because the described control circuit can handle voltage and current requirements higher than at least the design voltage or rated voltage of the drive transistor without significant leakage or damage, an OLED stack with increased emission (and higher voltage) can be used. A tandem OLED device with two light-emitting cells and a relatively low _Vth requirement does not emit as much light as an OLED stack with three or more light-emitting cells and a relatively high _Vth requirement. The described circuit can be used with a stacked light-emitting OLED having a critical voltage ( _Vth ) greater than 7.5 V; more desirably, the _Vth of the light-emitting OLED stack is at least 10 V or greater. Alternatively, the circuit may be used with a stacked OLED that provides a full-color microdisplay having a light emission of at least 2500 nits or preferably at least 5000 nits.

There are two basic approaches to making a pixelated OLED microdisplay, where the brightness of each individual pixel must be controlled by supplying power to one of the pixel electrodes via control circuitry on a backplane. The first approach involves having each pixel individually produce red, green, or blue light (R, G, B, respectively), or the same color if it is a monochrome display. In this case, the light-emitting OLED stack can be configured so that all stacked light-emitting units located above a respective bottom electrode segment emit the same color light (selected from R light, G light, or B light) to form R, G, and B pixels. In certain embodiments having this feature, each color pixel forms a microcavity in which the distance between the segment bottom electrode and the top electrode depends on the color of the emitted light. In this case, the length of the microcavity will depend on the color emitted, and will be different when the pixel is red, green, and blue.

The second method is to have all pixels of a color filter array (CFA) have a common multi-mode (white) emitting OLED layer to form individual RGB pixels. The advantage of the second method over the first method is that it is not necessary to form individual OLED pixels for different schemes and thus will reduce manufacturing costs.

The number of individual OLED light cells within a stacked OLED is limited only by the overall thickness of the OLED and the ability of the control circuitry to handle the power required to operate the OLED. As the number of OLED cells increases, the total amount of light emitted increases, but the package thickness, manufacturing process complexity, and critical voltage all increase as well. An OLED having at least three stacked light cells will increase illumination over a series (two OLED cell) OLED. However, preferably the OLED has at least four stacked OLED light cells, and more preferably the OLED has at least five stacked OLED light cells. An OLED having six or even ten or more stacked OLED light cells is conceivable.

To minimize the increase in voltage required to drive an OLED stack, charge generating layers (CGLs; sometimes also called connectors or intermediate layers) are positioned between individual OLED light emitting cells. This is because the CGL is constructed so that electrons and holes are generated when voltage is applied, and the electrons and holes are injected into the adjacent organic emissive layer. Therefore, using a CGL can convert one injected electron into multiple photons, allowing higher illumination. Specifically, it is desired that a CGL is located between each light emitting cell in the stack. However, a light generating cell does not have to have an adjacent CGL on both sides. OLED light generating cells on the top and bottom of the stack will typically have only one adjacent CGL. It is generally not necessary to use a CGL between a light emitting cell and either the top electrode or the bottom electrode, but a CGL may be used if desired.

Many different types of CGLs have been proposed and can be used in OLED stacks. For example, see US 7728517 and US 2007/0046189. To form a CGL, an np semiconductor heterojunction at the interface of the n-type layer and the p-type layer is usually required to generate charges. Therefore, the CGL will have two or more layers. For example, n-doped organic layer/transparent conductive layer, n-doped organic layer/insulating material, n-doped organic material layer/metal oxide layer and n-doped organic material layer/p-doped organic material layer have all been reported. One of the desired metal oxides for the CGL is MoO ₃ . In some instances, the n-layer and the p-layer may be separated by a thin intermediate layer. Typically, the CGL is configured so that the n-layer is closer to the anode and the p-layer is closer to the cathode.

One desirable embodiment of a CGL has three layers; an electron transport material doping with an n-dopant (e.g., Li), a thin intermediate layer of the same (but undoped) electron transport material, and a hole transport material doping with a p-dopant. Suitable electron transport and hole transport materials, as well as n- and p-dopant suitable for use in a CGL are well known and commonly used. The materials may be organic or inorganic. The choice of appropriate material is not critical and any material may be selected based on its performance. The thickness of the CGL should desirably be in the range of 200 Å to 450 Å. In many instances, the CGL will have an ETL on the anode side and a HTL on its cathode side to help improve charge transfer and to help separate charge generating dopants (if present) from the LEL in the light emitting cell.

Although using a CGL when OLED light cells are stacked on top of each other helps minimize the voltage increase, the total voltage required for the stack still increases by approximately the voltage required for each individual cell alone. It is expected that an OLED with at least three stacked OLED light cells will require voltages outside the recommended operating range of a 5 V drive transistor.

In one embodiment, all OLED light emitting units within an OLED stack located above a respective bottom electrode segment emit the same color, such as red, green, or blue. This forms a pixelated RGB microdisplay. FIG. 10 illustrates a microdisplay 100 using three different OLED sub-pixel stacks to form R pixels, G pixels, and B pixels. Each OLED sub-pixel stack contains three OLED light emitting units emitting the same color, with a CGL separating each unit vertically from another unit.

In the microdisplay 100 there is a silicon backplane 3 which includes a control circuit array such as shown in FIGS. 2 to 8 and other necessary components to supply power to the sub-pixels according to an input signal. Above the layer 3 with the transistors and control circuitry there may be an optional planarization layer 5. Above the layer 5 (if present) are individual first electrode segments 9 connected by electrical contacts 7 which extend through the optional planarization layer to form electrical contacts between the individual bottom electrode segments 9 and the control circuitry in the layer 3. The individual bottom electrode segments 9 are electrically isolated from each other laterally by a pixel defining layer 1. A non-light-emitting OLED layer 11 (e.g., an electron or hole injection (EIL or HIL) or electron or hole transport (ETL or HTL) layer) is located above the segmented bottom electrode segment 9. A first light-emitting OLED unit 13 is located above the OLED layer 11. Layer 15 is a charge generation layer, which is located between the first light-emitting layer 13 and a second light-emitting OLED unit 17 and separates the first light-emitting layer 13 from the second light-emitting OLED unit 17. There is a second charge generation layer 19 above the second light-emitting OLED unit 17, which is located between the second light-emitting OLED unit 17 and a third light-emitting OLED unit 21 and separates the second light-emitting OLED unit 17 from the third light-emitting OLED unit 21. Above the third light-emitting layer 21 are non-light-emitting OLED layers 23, such as electron or hole transport layers or electron or hole injection layers and a transparent top electrode 25 through which light can be transmitted. The OLED microcavity is protected from environmental influences by an encapsulation layer 27. In the embodiment shown, all organic layers within a single OLED stack are horizontally separated from adjacent stacks by a pixel definition layer 1, but the top electrode 25 and the encapsulation body 27 are common and extend across the entire active area. However, the top electrode 25 does not need to be continuous and can be segmented as needed. In the microdisplay 100, all OLED light-emitting units 13, 17, 21 located above the same bottom electrode segment 9 emit light of the same color, B, G or R. This particular microdisplay is not a microcavity device; however, similar pixelated RGB designs using a microcavity effect can be used.

A well-known method of increasing the luminance and color purity of OLED emission is to exploit the optical microcavity effect. This effect is based on the formation of an optical resonator between an emitting surface and a semi-reflecting surface that allows some light to pass through. Multiple reflections between the two surfaces form resident waves, depending on the optical distance between the two surfaces, which will enhance light of certain wavelengths and reduce light of other wavelengths due to constructive and destructive interference effects, which will occur depending on whether the emission occurs at the antinodes or nodes of the resident wave, respectively. The antinodes appear at different locations depending on the total space between the reflectors and depending on the wavelength being optimized. However, the light emitted from the microcavity can exhibit strong angular dependence, where color shifts and luminance losses can occur when the viewing angle deviates from the perpendicular to the observation surface. This is usually not a problem for NED applications due to the limited entry angle of the projection optics.

It would be desirable to use the microcavity effect to further increase the illumination of the OLED stack. For example, the microdisplay 100 shown in FIG. 10 can be redesigned to form a microcavity effect by using a reflective bottom electrode or a reflective layer below the bottom electrode, making the top electrode semi-transparent so that it has a certain degree of reflectivity and adjusting the distance between the uppermost surface of the reflective element (bottom electrode 9 or underlying reflective layer) and the bottommost surface of the top electrode to form a microcavity suitable for that particular color of light.

FIG. 11 illustrates a microdisplay 200 using a multi-mode (white) OLED microcavity that is shared by all pixels and a color filter array (CFA) to form R pixels, G pixels, and B pixels. The multi-mode OLED produces light of one color. Ideally, a multi-mode OLED produces a white light with approximately equal amounts of R light, G light, and B light. Typically, this will correspond to CIE _x , CIE _y values of approximately 0.33, 0.33. However, depending on the characteristics of the color filters used to form the RGB pixels, some variation in these values may still be acceptable or even desirable. The microdisplay 200 also includes a microcavity effect. In this embodiment, the multi-mode OLED stack contains three OLED light-emitting cells emitting different colors, where each cell is vertically separated from the other cell by a CGL, where the distance between a reflective surface and the top electrode is constant over the active area.

In the microdisplay 200, there is a silicon backplane 3 including a control circuit array such as shown in Figures 2 to 8 and the necessary components to supply power to the sub-pixels according to an input signal. On top of the layer 3 with the transistors and control circuitry there may be an optional planarization layer 5. On top of the layer 5 (if present) are individual first electrode segments 9 connected by electrical contacts 7 extending through the optional planarization layer to form electrical contacts between individual bottom electrode segments 9 and the control circuitry in the layer 3. In this embodiment, the bottom electrode segment 9 has two layers: a reflective layer 9b closer to the substrate 1 and an electrode layer 9a closer to the OLED layer. The individual bottom electrode segments 9 are electrically isolated from each other in the lateral direction. Above the segmented bottom electrode segments 9 is a non-light-emitting OLED layer 11, such as an electron or hole injection or electron or hole transport layer. A red OLED light generating unit 13A is located above the OLED layer 11. Layer 15 is a first charge generating layer, which is located between the red OLED light generating unit 13A and a green OLED light generating unit 17A and separates the red OLED light generating unit 13A from the green OLED light generating unit 17A. Above the green light emitting layer 17A there is a second charge generating layer 19 which is located between and separates the green OLED light generating unit 17A and a blue OLED light generating unit 21A. Above the blue OLED light generating unit 21A is a non-light emitting OLED layer 23 (e.g. an electron or hole transport layer or an electron or hole injection layer) and a semi-transparent top electrode 25. This forms an OLED microcavity 30 which extends from the uppermost surface of the reflective surface 9B to the bottommost surface of the semi-transparent top electrode 25 which is also a semi-reflective electrode. An encapsulation layer 27 protects the OLED microcavity from the environment. In this embodiment, there is a color filter array having color filters 29B, 29G and 29R, which filter the multi-mode emission generated by the OLED microcavity 30 so that B light, G light and R light are emitted according to the power supplied to the underlying electrode segment 9.

12 shows a microdisplay 300, which is a variation of microdisplay 200, having an additional blue OLED light emitting layer 22 between blue OLED light emitting unit 21A and OLED layer 23. In this embodiment, there is no CGL between blue OLED unit 21A and additional blue light emitting layer 22. As with microdisplay 200, microdisplay 300 has three (rather than four) OLED light generating units because blue unit 21A and blue layer 22 are separated by a CGL (and thus layer 22 is not considered a separate light emitting OLED unit and should be considered as part of blue light emitting unit 21A). Even if the blue cell 21A is separated from the blue LEL 22 by a non-charge generating interlayer, there will still be three OLED cells in the stack. However, if an intermediate CGL is located between the blue emitting cell 21A and the additional blue light-emitting layer 22, the OLED stack will contain four OLED cells.

FIG. 13 shows another embodiment of a multi-mode OLED microdisplay 400 with a microcavity effect, the multi-mode OLED microdisplay 400 is similar to the microdisplay 200, but the multi-mode OLED microdisplay 400 has five OLED light-emitting units. In this case, some of the OLED units emit light of the same color and other units emit light of a different color. Specifically, the multi-mode microdisplay 400 has two blue OLED light-generating units, one green OLED light-generating unit, one yellow OLED light-generating unit, and one red OLED light-generating unit, all of which are separated by CGLs.

In microdisplay 400, OLED cell 13A emits red light. OLED cell 13A is separated from yellow light-emitting OLED cell 16 by CGL 15. Yellow light-emitting OLED cell 16 is separated from green light-emitting OLED cell 17A by CGL 14. Green light-emitting OLED cell 17A is separated from a first blue light-emitting OLED cell 21A by CGL 19. A second blue light-emitting cell 32 is separated from the first blue light-emitting OLED cell 21A by CGL 24. All light-emitting OLED cells of another layer are the same as in FIG. 12.

As previously described, OLED microdisplays are constructed on a silicon backplane used as a substrate. Typically, the backplane will be flat, with a uniform thickness. Since silicon backplanes are typically opaque, the OLED stack is preferably top emitting. However, transparent backplanes are known; in such cases, the OLED stack can be top emitting or bottom emitting. The top surface of the substrate faces the OLED. The silicon backplane can have various types of alternative layers (i.e., planarization layers, light management layers, light blocking layers, etc.), which can be patterned or not and can be located on the top surface or the bottom surface.

The bottom electrode segment (9 or 9a) can be an anode or a cathode and can be transparent, reflective, opaque or translucent. If the OLED is top emitting, the bottom electrode can be made of a transparent metal oxide or a reflective metal (such as Al, Au, Ag or Mg or their alloys) and have a thickness of at least 30 nm, preferably at least 60 nm.

In microcavity applications where the first electrode is located on a reflective layer, the first electrode should be transparent. However, in other applications, the first electrode layers 9a and 9b can be collapsed into a single reflective electrode so that its uppermost reflective surface forms one side of the optical microcavity (i.e., 30 in Figure 10).

When the OLED stack is a top emitting microcavity and the bottom electrode is transparent, there should be an emitting layer below the bottom electrode defining a first side of the microcavity 30. When a transparent anode is located on a reflective surface, it is part of the optical cavity. The reflective layer 9b can be a reflective metal such as Al, Au, Ag, Mg, Cu or Rh or alloys thereof, a dielectric mirror or a highly reflective coating. Dielectric mirrors are constructed from multiple thin layers of materials (e.g., magnesium fluoride, calcium fluoride and various metal oxides) deposited on a substrate. Highly reflective coatings consist of multiple layers of two materials, one with a high refractive index (e.g., zinc sulfide (n=2.32) or titanium dioxide (n=2.4)) and one with a low refractive index (e.g., magnesium fluoride (n=1.38) or silicon dioxide (n=1.49)). The thickness of the layer is typically a quarter wave from the wavelength of the reflected light. It is desirable that the reflective layer reflect at least 80% of the incident light and optimally at least 90%. Preferred reflective layers are Al or Ag, with a thickness of 300 Å to 2000 Å, optimally 800 Å to 1500 Å.

Desirably, when the OLED stack is bottom emitting, the bottom electrode is a transparent anode and should transmit as much visible light as possible, preferably with a transmittance of at least 70% or more desirably at least 80%. Although the bottom transparent electrode can be made of any conductive material, metal oxides (such as ITO or AZO) or thin metal (such as Ag) layers are preferred. Weakly conductive materials (such as TiN) can be used given that they can be made very thin.

Electron transport and hole transport materials suitable for use in non-emissive layers (i.e., 11 and 23 in FIG. 10 ) such as hole injection layers, hole transport layers or electron injection layers or electron transport layers are well known and commonly used. These layers may be mixtures of these materials and may contain dopants to modify their properties. Since they are non-luminescent, they do not contain emissive materials and are transparent. The choice of appropriate materials is not critical and any material may be selected based on its performance.

In embodiments utilizing the microcavity effect, since the spacing between various OLED units within the microcavity and the size of the microcavity are important for maximizing efficiency, the thickness of various non-luminescent layers is generally selected to provide the desired spacing. It is desirable to adjust the spacing between OLED units and the size of the microcavity by using the appropriate thickness of the organic non-emissive layer (e.g., the hole transport layer).

The light-emitting layer typically has a host material (or a mixture of host materials) and a light-emitting compound, which is the main component of the layer. It is desirable that the light-emitting compound is phosphorescent because the light-emitting compound has a higher efficiency. However, in some examples, certain LELs may use fluorescent or TADF (thermally activated delayed fluorescence) compounds as light-emitting materials, while other LELs use phosphorescent materials. Specifically, blue light OLED layers may use fluorescent or TADF compounds or combinations thereof, while non-blue light-emitting layers may use green, yellow, orange or red phosphorescent compounds or combinations thereof. Light-emitting layers may use combinations of light-emitting materials. The selection of appropriate materials for LELs is well known and unimportant, and any material may be selected based on its performance and emission characteristics. When using phosphorescent emitters, it is sometimes desirable to confine the excitons generated by the phosphorescent emitter to the layer. Therefore, exciton blocking layers on either or both sides of the phosphorescent LEL may be used as desired. Such materials and their applications are well known. In addition, it may be desirable to add HBL (hole blocking layer) and EBL layers (electron blocking layer) around the luminescent layer (especially the blue luminescent layer) to improve lifetime and illumination efficiency.

The top electrode (i.e., 25 in FIG. 10 ) should be transparent if the OLED stack is top emitting, reflective if the OLED stack is bottom emitting, and in the case where the OLED stack is a microcavity, semi-transparent and semi-reflective: that is, it reflects a portion of the light and transmits the rest. In a microcavity, the bottommost interior surface of the top electrode defines a second side of the microcavity 30. Desirably, the semi-transparent top electrode reflects at least 5% and more desirably at least 10% of the light emitted by the LEL to establish a microcavity effect. The thickness of the semi-transparent second electrode is important because it controls the amount of reflected and transmitted light. However, it cannot be too thin because it cannot efficiently transfer charge into the OLED or withstand pinned holes or other defects. A thickness of the upper electrode layer is desirably 100 Å to 200 Å, and more desirably 125 Å to 175 Å.

The top electrode is desirably a thin metal layer or a thin metal alloy layer. Suitable metals include Ag, Mg, Al and Ca or their alloys. Of course, Ag is preferred because of its relatively low blue absorptivity. To aid electron transport and stabilization, there may be an adjacent transparent metal oxide layer (e.g., ITO, InZnO or MoO ₃ ) on the electrode surface. Alternatively, metal halides (e.g., LiCl), organometallic oxides (e.g., lithium quinolate) or other organic materials may be used.

There may be a protective layer or spacer layer (not shown in FIGS. 10-13 ) over the upper electrode to prevent damage during encapsulation.

An encapsulation 27 is deposited or placed over the top electrode 25 and any optional protective layer (if present). At a minimum, the encapsulation should completely cover the light emitting areas on the top and sides and directly contact the substrate. The encapsulation should be impermeable to air and water. It can be transparent or opaque. It should not be electrically conductive. It can be formed in situ or added as a separate pre-formed sheet and provide sealing of the side edges. An example of in situ formation would be thin film encapsulation. Thin film encapsulation involves depositing multiple layers with alternating layers of inorganic materials and polymeric layers until the desired degree of protection is achieved. Schemes and methods for forming thin film encapsulation are well known and any one can be used as desired. Alternatively, the encapsulation can be provided using a pre-formed sheet or cover sheet attached to at least the sealing area and the enclosure area. The preformed sheet may be rigid or flexible. It may be made of glass (including flexible glass), metal, or organic/inorganic barrier layers. It should have a thermal expansion coefficient close to that of the substrate to achieve a more stable connection. It may be necessary to attach the preformed encapsulation sheet to the sealing area using an air-proof and waterproof adhesive (e.g., silicone or epoxy adhesive) or by thermal means (e.g., ultrasonic welding or glass frit welding), which may require additional sealants such as solder or glass frit. The side and bottom edges of the cover sheet may be specially designed to better fit the sealing area or promote a better seal. The cover sheet and the sealing area may be designed together so that they fit or are partially locked in place before the seal is formed. Additionally, the cover sheet may be pre-treated to promote better bonding to the sealing area.

Although this application describes the use of OLEDs as light emitting elements in microdisplays, the same control circuit can be used for any self-emitting display technology that would require a relatively high voltage to emit light. The invention is not limited to OLEDs, but can be used with any other display technology that would require greater than 5 V, preferably greater than 7.5 V, or even greater than 10 V to provide at least 1000 nits, or preferably at least 5000 nits of light emission.

Crosstalk in a display is when the emitted illumination provided by one pixel is unexpectedly affected by another pixel. This is undesirable because the affected pixel no longer provides accurate illumination based on the image signal and therefore the quality of the image can be degraded. Depending on the amount and nature of the crosstalk, important factors in the display such as color reproduction, contrast (the difference between maximum and minimum illumination), grayscale, resolution, and "ghosting" can all be negatively affected. Typically, crosstalk in microdisplays is a particularly large problem where the pixel pitch is small. It is desirable that the amount of crosstalk between pixels is 15% or less, preferably 5% or less, and optimally 3% or less.

The stacked microcavity OLED device described must be thick, with defined spacing between internal light emitting cells. This is necessary to provide a microcavity effect to enhance R, G, and B emission within a single microcavity. This can cause crosstalk to increase due to optical and chemical/electrical mechanisms. Some optical processes that can increase the amount of crosstalk include light scattering and waveguiding within the OLED device. Some chemical/electrical processes that can increase crosstalk include lateral carrier migration from an active pixel region to a neighboring inactive pixel region in the same layer.

It is believed that multiple modes contribute to crosstalk. Short range (0.2 µm to 0.7 µm) appears to be a combination of lateral charge carrier and optical mechanisms. Mid-range (3 µm to 7 µm) interactions appear to be primarily due to lateral charge carrier migration but may be partially due to optical mechanisms. Long range (50 µm to 200 µm) interactions appear to be primarily due to light scattering from an active pixel region to an inactive region. It is also believed that, depending on the pixel spacing, there is a longer range optical contribution to crosstalk based on the waveguide.

Some useful methods to minimize crosstalk problems due to carrier migration in OLED devices include: - Reducing lateral charge carrier migration by changing the layer thickness and composition of layers with high carrier mobility (such as HIL, HTL, CGL, ETL and EIL) (to increase the "sheet resistance"). Specifically, charge carriers (holes or electrons) are generated in an active area and can move laterally across the gap between the illuminated area and the non-illuminated area. This problem seems to occur mainly in layers that are next to or close to one of the electrodes. It is believed that the common HIL layer and HTL layer located above the anode may be the biggest factor causing the problem. It seems that once holes are generated in an illuminated area of the HIL on one anode pad, they can migrate to a non-illuminated area of an adjacent anode pad, and the resulting voltage due to the holes can exceed the _Vth of the OLED and thus the (nominally non-illuminated) pixel will emit light anyway. Alternatively, holes can follow electrons into a conductive anode pad and flow laterally through the anode, which has very little lateral resistance. Far side of the non-illuminated anode pad, current can return to the HIL (following the holes) to jump to the next non-illuminated anode pad. Therefore, carrier migration problems cannot be limited to a shorter distance between adjacent anode pads, but can also have a longer distance component. Therefore, careful attention should be paid to the thickness and composition of both electrodes and specifically the anode. Thinner organic layers with smaller carrier mobility can help to minimize these undesired carrier migration processes. - Material selection of organic layers with high carrier mobility. Specifically, the material can be selected to minimize its contribution to crosstalk. In this regard, the type and level of dopants (e.g., F4-TCNQ, F6-TCNNQ, or HAT-CN) added to the HIL and the choice of HTM (e.g., aromatic amine compounds such as NPB or spiro-TTB) in the HIL or HTL may be important. Only p-doped or non-doped HILs may also be effective. In some cases, a non-doped HIL and a p-doped HTL may be used. Inorganic HIL materials such as MoO ₃ (which may be mixed with organic materials) may also have advantages.

Some useful approaches to minimize crosstalk problems caused by optical processes in OLED devices include: - Optimizing color filters to reduce light waveguides between the air/glass interface and the reflective anode, including using optical filters designed to absorb light traveling at high angles from the substrate normal. - Reducing light scattering by reducing scattering points. Specifically, the amount of small particle debris on or near the anode should be minimized. Scattering can also occur due to the roughness of the cathode, which can depend on the composition and process used for deposition. (For example, see Shen et al., “Efficient Upper-Excited State Fluorescence in an Organic Hyperbolic Metamaterial,” Nano Lett. 2018, 18, 3, 1693–1698). -The overall anode surface should be as flat and smooth as possible over the active pixel area and between pixels. Specifically, it is known that a PDL (pixel definition layer) is formed between pixels and protrusions, bumps or other structures above the anode surface extending into the pixel area can be used to scatter light back into the pixel area and prevent light from entering a neighboring (non-illuminated) pixel. However, this approach does not work when there is a thicker OLED layer overlying the structure. Light trapped in the thicker layer is more likely to be internally reflected within the layer so that it can travel across the structure to the other side. If the anode and overlying OLED layer are uniformly flat, light waveguided within the layers of the display is more likely to continue uninterrupted until it is absorbed or reaches the edge of the display. -Use absorbers for waveguided light. - Absorb light through the dielectric of the backplane. - Design the HIL and anode to form a barrier for charges to enter the anode from the HIL.

To control crosstalk, it may be useful to add an additional separate pixel control circuit to a node between the second switching transistor and the anode of the OLED to control the voltage at the anode when the OLED is driven so that it is nominally "off" (not powered and therefore not emitting). Specifically, the pixel control circuit should reduce the voltage below the _Vth of the OLED only when the drive circuitry requires the pixel to not emit or to emit at least weakly. In this way, any voltage developed by lateral carrier migration can be minimized. Experimental Results

In the following examples, the number preceding each material (e.g., 200 HIL) is the physical layer thickness in angstroms unless otherwise specified. All % are by weight. All devices were encapsulated using the same process after cathode deposition. All materials are commercially available.

The backplane used for the following experimental setup is a single crystal silicon based backplane and is used as is with the schematic control circuitry shown in Figures 6a and 6b. The backplane contains: a drive transistor (T1) and a switch transistor (T2) which are low voltage (rated to 5 V) p-channel transistors connected in series; and a protection circuit having an NPN bipolar junction transistor (BJT1) for each pixel. In all experimental cases, the base of BJT1 is isolated; that is, the base is not connected to an external power supply so that the base voltage of the BJT is not intentionally controlled or set to any specific voltage. The backplane includes a reflective metal layer as the bottom electrode/anode.

A comparative 2-cell stacked OLED device A-1 was prepared as follows: Layer 1 (reflective bottom electrode/anode): as delivered Layer 2 (hole injection layer HIL): 200 hole transport material HTM1 + 6% p dopant PD1 Layer 3 (hole transport layer HTL1): 410 hole transport material HTM1 Layer 4 (HTL2): 50 hole transport material HTM2 Layer 5 (green LEL): 100 green host GH1 and 3% green phosphorescent dopant GPD1 Layer 6 (yellow LEL): 100 GH1 + 10% GPD1 + 3% red phosphorescent dopant RPD1 Layer 7 (Electron transport layer ETL1): 200 electron transport material ETM1 Layer 8 to layer 10 (charge generation layer CGL1): 270 three layers Layer 11A (HTL3): 300 HTM1 Layer 11B (HTL4): 280 HTM1 Layer 12 (blue LEL2): 200 blue main BH1 + 4% fluorescent blue dopant BFD1 Layer 13 (ETL2): 150 ETM2 Layer 14 (electron injection layer EIL): 150 ETM1 + 2% Li Layer 15 (semi-transparent cathode): 105 75% Ag / 25% Mg

A second comparative OLED device A-2 was prepared in a similar manner, but the thickness of layers 11A and 11B were increased to 1090 and 4090, respectively. Both A-1 and A-2 are two-cell stacked OLEDs, where layers 5 (G LEL) and 6 (red LEL) together represent one cell (a G-Y stack) and layer 11 (blue LEL) represents a second cell (a B stack) separated from the first cell by a CGL. Layers 5 and 6 are a single cell because they are adjacent LELs that are not separated from each other by a CGL (layers 8 and 9). Both A-1 and A-2 are microcavity devices, where the microcavity thickness of A-1 is 2320 Å and the microcavity thickness of A-2 is 6920 Å. These dual-cell stacked OLED devices represent the current state of the art.

An inventive 3-unit stacked OLED B-1 is prepared in the same manner as A-1, but the following additional layers are inserted between layers 11A (whose thickness is adjusted to 470) and 11B (which is 4090): Layer 16 (BLEL1): 200 BH1 + 4% BFD1 Layer 17 (ETL3): 150 ETM2 Layers 18-20 (CGL2): 270 Three layers (same scheme as CGL1) OLED device B-1 is a 3-unit stacked OLED, in which layers 5 and 6 are the first GY unit (same as in A-1), layer 16 is an additional second B unit and layer 12 (same as in A-1) is a third B unit. The second cell (layer 16) is separated from the first cell (layers 5 and 6) and the third cell (layer 12) by a CGL. That is, B-1 contains an additional B cell compared to A-1 and will therefore have higher illumination but increased voltage requirements. OLED B-1 is a microcavity device with the same microcavity thickness (6920 Å) as device A-2. Figure 14 shows the characteristic IV curves (current density in mA/ ^cm2 versus voltage) for OLEDs A-1, A-2 (comparative two-cell device) and B-1 (inventive three-cell device), which are also listed in Table 1. Table 1 - IV Data for OLEDs Listed Device I* V* L* A-1 Threshold 5.5 1 6.0 30.4 5 6.7 40.0 10 7.1 41.5 100 10.8 35.5 A-2 Threshold 5.3 1 6.3 29.0 5 7.3 30.3 10 8.0 30.3 100 13.5 24.5 B-1 Threshold 7.6 1 9.0 28.5 5 9.9 32.8 10 10.7 33.7 100 16.0 28.6 * I = current density on the anode pad in mA/ ^cm2 V = voltage L = luminance efficiency in cd/A In Figure 14 and Table 1, OLEDs A-1 and A-2 have approximately the same _Vth . However, OLED B-1, which has an additional B cell, has a significantly higher _Vth and requires a higher voltage (approximately 2.5 V higher) to produce the same current density as A-2.

The advantage of having a microcavity 3-cell stacked OLED over a 2-cell stacked OLED can be seen in Figure 15, which compares (intensity vs. wavelength in nm) the spectral output without the color filters of A-1, A-2, and B-1. OLED A-1, which has a thinner microcavity, has relatively less B and R emission compared to G emission. OLED A-2, which has one microcavity suitable for all 3 colors, has more B and R emission compared to G, but the B emission is still limited. OLED B-2, which has an extra B cell, has a greatly increased blue emission compared to A-2.

The above results refer to a white OLED device before the addition of color filters that will form the R, G, and B pixels for a microdisplay. It is well known in the art that a desired R, G, and B balance in a display (usually referred to as white luminance (e.g., a D65 white point)) can be achieved by driving individual color pixels to the appropriate luminance for that color, regardless of any inherent imbalance in the relative amounts of R, G, and B light produced by the W OLED. For example, if a W OLED produces insufficient blue light and too much red light compared to green light, the B pixels can be driven at a higher voltage to produce more blue light, while the R pixels can be driven at a lower voltage so that less red light is produced.

FIG. 16 shows a graph of the peak current density required to produce a D65 white point for each color of OLED devices A-1, A-2, and B-1 using color filters with equal areas of R, G, and B subpixels. The current density required in B for the two-cell devices (A-1 and A-2) is much greater than that for the three-cell device B-1. The current density in R and G is about the same as for all three devices. The three-cell device B-1 requires less current density and therefore requires less power to achieve the same target illumination at the same white point. The lifetime of a display is determined by the rate at which the efficiency of an OLED in continuous operation decreases, which is primarily a function of the primary current density and temperature. Therefore, the lifetime of display B-1 will be more than twice that of A-1 or A-2.

In microdisplays that need to produce relatively high illumination given their small size, it is desirable that the total current (and current density) for each color be approximately the same (or as close as possible) when displaying white at maximum display illumination. This will achieve several of the advantages of the invention below: neutral fading (where the illumination efficiency of each color of a subpixel degrades at approximately the same rate over the lifetime of the device), high illumination (each pixel will be driven at its maximum brightness resulting in increased efficiency), high contrast (the full dynamic range of each pixel is available), and extended lifetime (each color of the pixel shares the current load equally).

However, the advantage of using an OLED with three or more cells to increase illumination can be offset if the required voltage increases. In a display, the power for each pixel is supplied and controlled by control circuits in the backplane. However, in a microdisplay that is decreasing in area but requires a high number of pixels, the physical space available for transistors and other circuit components is limited. Generally speaking, the higher voltage and current requirements must be addressed by larger, thicker or more robust electronic circuit systems. Currently, the size of the OLED microdisplay dictates that the electronic circuit system be rated to handle 5 V or less to achieve the desired pixel size and density (pixel pitch).

Electronic circuit systems such as transistors can fail catastrophically if the power supplied to the transistor's gate is too great for a long period of time. However, not all levels or durations of overvoltage will cause a transistor to fail permanently; in many cases, the transistor will fail due to leakage. Increased leakage is a common failure mode caused by non-destructive overstressing of a semiconductor device, when the junction or gate oxide suffers insufficient permanent damage to cause a catastrophic failure. Overstressing the gate can result in stress-induced current leakage.

Solving motion blur in microdisplays used for virtual reality and similar products requires that the OLED provide very high illumination for a very short period of time and then be turned off for another period of time (the total on/off cycle is less than the human perceptible range to avoid flicker). This means that the control circuitry in the backplane must supply a high voltage/current to produce high illumination for one cycle and then cut power to the OLED for another cycle. If the power applied to the transistors in the control circuitry is too high during the cycle when the OLED is on, current can leak through the transistors when they should be "off" and the OLED will not shut down completely. The difference between the illumination of the OLED when it is on and the illumination when it should be off is often called the "contrast". It is desirable for a microdisplay to have a high contrast because it makes the image appear sharp and bright; low contrast makes the image appear dark and flat.

Figure 17 shows several sets of (illuminance versus cathode voltage) curves for a 2-cell comparative device and an inventive 3-cell device. The upper set of curves is when the device is driven at full illumination (white = CV255) and the drive transistor is "on". The lower set of curves is when the device is driven at zero illumination (black = CV0) and the drive transistor is nominally "off". However, current leakage through the drive transistor causes some small amount of illumination to occur and the voltage at the cathode is non-zero. The difference between the two sets of curves represents the "contrast" of the device and it is desirable to make the "contrast" as high as possible by minimizing the illumination brought about by current leakage through the drive transistor. As can be seen in Figure 17, the contrast of all three devices is approximately the same (slightly less than 10 ⁵ ) and is relatively constant over a wide range of cathode voltages.

Note that the inventive 3-cell device maintains approximately the same overall contrast as the comparative 2-cell device, but requires higher operating voltages that exceed the design limits of the drive transistors. Although B-1 requires less current density to achieve full white than A-1 or A-2 shown in Figure 16, B-1 still requires a higher voltage (about 2.5 V, see Figure 14). There is no evidence of increased leakage for the higher voltage OLEDs with the backplane used in the present invention. In addition, there is no evidence of any catastrophic transistor failures in B-1 due to the higher voltage requirement (see Table 1), nor does the device lifetime appear to be affected from the control circuitry. These results are unexpected because the transistors in the backplanes of A-1, A-2 and B-1 are designed to be suitable for driving 1-cell OLED or 2-cell OLED, which are acceptable and commonly used low-voltage transistors.

Another inventive 3-cell OLED device B-2 using the same low voltage backplane as A-1, A-2 and B-1 was prepared as follows: Layer 1 (bottom electrode/anode): Layer 2 (HIL): 200 HTM1 + 6% PD1 Layer 3 (HTL1): 410 HTM1 Layer 4 (HTL2): 50 HTM2 Layer 5 (green LEL): 100 GH1 and 3% GPD1 Layer 6 (yellow LEL): 100 GH1 + 10% GPD1 + 3% RPD1 Layer 7 (ETL1): 200 ETM1 Layers 8 to 10 (CGL1): 270 Trilayer Layer 11 (HTL3): 570 HTM1 Layer 12 (Blue LEL2): 250 BH1 + 4% BFD1 Layer 13 (ETL2): 150 ETM2 Layers 14 to 16 (CGL2): 270 triple layer (same scheme as CGL1) Layer 17 (HTL4): 3840 HTM1 Layer 18 (BLEL1): 250 BH1 + 4% BFD1 Layer 19 (ETL3): 150 ETM2 Layer 20 (EIL): 100 ETM2 + 2% Li Layer 21 (semi-transparent cathode): 105 75% Ag/25% Mg

OLED B-2 is a 3-cell device similar to B-1 in that it contains a G-Y cell (layers 5 and 6) separated by a CGL (layers 8 to 10) from a second blue cell (layer 12), which in turn is separated from a first blue cell (layer 18) by a second CGL (layers 14 to 16), in the order Y-G/B/B (starting from the backplane). Similar to B-1, B-2 is a microcavity white light generating OLED with a microcavity of 6910 Å.

Similarly, an inventive 4-unit OLED device C-1 was prepared as follows: Layers 1 to 11: Same as layers 1-11 in B-2 Layer 12 (Blue LEL2): 200 BH1 + 4% BFD1 (50 Å less) Layer 13 (ETL2): 100 ETM2 (50 Å less) Layers 14 to 16: Same as layers 14 to 16 in B-2 Layer 17 (HTL4): 420 HTM1 (3420 Å less) Layer 18 (BLEL1): 200 BH1 + 4% BFD1 (50 Å less) Layer 19 (ETL3): 100 ETM2 (50 Å less) Layers 20 to 22 (CGL3): 270 three layers (same solution as CGL1 and CGL2) Layer 23 (HTL5): 2620 HTM1 Layer 24 (HTL6): 50 HTM2 Layer 25 (yellow LEL): 200 GH1 + 10% GPD1 + 3% RPD1 Layer 26 (ETL4): 480 ETM1 Layers 27 to 28: Same as layers 20 to 21 in B-2

OLED C-1 is a 4-cell device because it contains an extra Y cell (layer 25) as well as the same G-Y cell and two B cells as in B-2, in the order Y-G/B/B/Y (starting from the backplane). In C-1, the extra Y cell is separated from the first B cell (layer 18) by CGL3 (layers 20-22). C-1 is a microcavity white light generating OLED with a microcavity of 6910 Å.

Similarly, an inventive 5-unit OLED device D-1 was prepared as follows: Layers 1 to 10: Same as layers 1 to 11 in C-1 Layer 11 (HTL3): 370 HTM1 (200 Å less) Layers 12 to 16: Same as layers 12 to 17 in C-1 Layer 18 (ETL2): 100 ETM2 (50 Å less) Layers 14 to 16: Same as layers 14 to 16 in C-1 Layer 17 (HTL4): 620 HTM1 (200 Å more) Layers 18 to 22: Same as layers 18 to 22 in C-1 Layer 23 (HTL5): 610 HTM1 (2010 Å less) Layer 24 (BLEL3): 200 BH1 + 4% BFD1 Layer 25 (ETL3): 100 ETM2 Layers 26 to 28 (CGL4): 270 three layers (same solution as CGL1, CGL2 and CGL3) Layer 29 (HTL6): 1440 HTM1 Layers 30 to 34: Same as layers 24 to 28 in C-1

OLED D-1 is a 5-cell device because it contains an additional B cell (layer 24) and the same Y cell, G-Y cell, and two B cells as in C-1. It has a total of 3 B cells, one Y cell, and one G-Y cell, all separated by a CGL, in the order Y-G/B/B/B/Y (starting from the backplane). C-1 is a microcavity white light generating OLED with a microcavity of 6910 Å.

Figure 18 shows the characteristic IV curves of OLED B-2, C-1 and D-1, which are also listed in Table 2. Table 2 - IV data of OLEDs Device I* V* L* B-2 Threshold 8.2 - 1 9.1 30.9 5 10.2 35.4 10 10.9 35.8 100 15.8 30.2 C-1 Threshold 10.5 - 1 11.7 52.6 5 13.0 53.6 10 13.9 52.7 100 19.0 42.2 D-1 Threshold 12.8 - 1 14.5 50.5 5 16.0 51.9 10 17.1 51.6 100 22.1 42.1 * I = current in mA/ ^cm2 V = voltage L = illuminance in cd/A

As shown in FIG. 18 and Table 2, adding each cell in the OLED provides greater illumination but also increases _Vth and the operating voltage by approximately 2.5 V.

The advantage of adding extra cells to a microcavity 3-cell OLED can be seen in Figure 19, which compares the spectral output (without color filters) of B-2 (3 cells), C-1 (4 cells), and D-1 (5 cells). OLED B-2 has a G/Y cell and 2 B cells. Adding another Y cell, like in C-1, increases G and R emission but not B emission much relative to B-2. Adding a third B cell, like in D-1, further increases blue emission while maintaining high G and R emission.

Similar to the results shown in FIG. 16 , FIG. 20 shows the relative amounts of peak current density per color pixel required to produce a balanced white illuminance of 1500 cd/m ² (using a color filter) for OLEDs B-2, C-1, and D-1. Adding an additional Y cell in C-1 reduces the current required to produce the necessary amounts of R and G illuminance compared to B-2. Adding an additional B cell in D-1 further reduces the voltage required to produce the necessary amount of B light compared to C-1. Because these OLEDs with three or more cells require less current density to achieve the same illuminance, their operating lifetimes will be significantly greater than OLEDs with fewer cells. The LT70 (time for emission output to degrade by 70%) lifetime of these devices is expected to be at least 18,000 hours of average video content.

Figure 21 (similar to Figure 17) shows several sets of (illuminance vs. cathode voltage) curves for OLEDs B-2, C-1, and D-1. In all three of these examples, the contrast is about the same (slightly less than 10 ⁵ ) and relatively constant over a wide range of cathode voltages. Notably, the contrast remains constant even when the devices are operated at voltages that are 2 or 3 times the design limit of the drive transistors.

Without being bound by any particular theory or speculation, the reason why a stacked OLED needs to maintain high contrast with a relatively high _Vth without burning or damage when using a backplane with low voltage transistors may be related to the use of series connected transistors. Current leakage in transistors is a well known problem and generally speaking, the higher the voltages and currents involved, the greater the leakage. As previously discussed, this leakage can cause the OLED to emit light at greater than Vth. Since the total leakage through the series connected transistors will be (leakage of the first transistor) x (leakage of the second transistor), it is possible that this multiplication effect is sufficient to significantly reduce the leakage at the bottom electrode of the OLED enough so that the OLED does not emit due to current leakage and thus maintain contrast.

In operation, the control circuitry in the backplane of the experimental microdisplay, including at least two transistors with their equal channels connected in series, is an example of a compact circuit that protects the drive transistor T1 from operating under conditions outside the specified range, even when driving an OLED having three or more stacks with a switching voltage range greater than the specified operating range of the LV transistor. If the switching voltage range of the OLED exceeds the LV operating range, only the switching transistor is exposed to these conditions. This is less detrimental to image quality than another pixel circuit design in which the drive transistor is bound to operate outside the specified operating range.

During operation, MOSFET devices age, which causes some of the characteristics such as threshold voltage, subthreshold slope, and transconductance at saturation to slowly change. These are all symptoms of negative bias temperature instability (NBTI) and hot carrier injection (HCI) found in p-channel transistors. By limiting the operation of a transistor to the specified voltage range of the transistor, it is ensured that the transistor characteristics will remain within a narrow specified range over a long period of time (e.g., 5 years). Operating the transistor outside the specified voltage range increases the rate at which the transistor characteristics change, thereby shortening the period during which the characteristics are within the specified range.

It is important that the characteristics of the drive transistors remain within specified limits, otherwise image degradation (commonly referred to as unbalance) may result. The operation of the switching transistors is more robust even with changes in their performance characteristics, and can be driven with a sufficiently wide gate voltage to ensure satisfactory operation even when the transistor characteristics have shifted outside of the specified range. Thus, the circuits shown in FIGS. 2 to 6 , including the circuits used in the backplane of the inventive experimental stacked OLED microdisplay, are examples of compact circuits designed to handle the higher switching voltages associated with three or more OLED cells without causing the expected degradation of the display quality due to exceeding the specified operating range of the transistors.

It is believed that the circuits in Figures 2-6 achieve this by selecting the SELECT2 voltage to turn transmission "on" and "off". For "off" transmission, an ideal choice for the SELECT2 voltage is _VDD , which will place T2 in the sub-threshold region of operation, effectively stopping the current. For "on" transmission, the SELECT2 voltage can be selected to be as low as 5 V, which is lower than _VDD .

If the OLED's switching voltage swing is less than about 4V, when the pixel displays "black", the black level current will be set by driving the transistors in a normal manner and the drain of T1 will be higher than the SELECT2 "on" voltage and the switching transistor T2 will still be on. However, when the OLED's switching voltage swing is greater than about 4V, as in the case of a multi-cell stacked OLED with more than 4 cells, when the black level current is formed, the drain voltage of T1 (also the source voltage of T2) is close to the SELECT2 "on" voltage, and the switching transistor T2 is turned off when the low drain voltage of T1 reduces the overvoltage of T2 (Vgs-Vth) to zero or slightly negative. This drives T2 into the sub-threshold range, thereby forming a black level. Under these conditions formed by a multi-cell stacked OLED, the driver transistor T1 controls T2 to form a black level current.

Therefore, it is believed that the simple series-connected transistor design shown in Figures 2 to 6 is capable of driving a multi-cell stacked OLED with a switching voltage range exceeding the voltage range of low-voltage transistors with reduced aging-related imbalance compared to a single transistor driving circuit shown in Figure 1, for example.

The protection circuit is designed to maintain the voltage at the anode of the OLED below a certain level whenever the OLED is intended to be "off" or not emitting and can help maintain a high contrast ratio. When the cathode voltage decreases (more negative), the protection circuit is designed to provide additional current to protect the drive transistors and the switching transistors so that the voltage level does not violate the maximum rating of the device.

However, the protection effect was still observed in experimental examples where the base of the BJT was isolated and not connected to an external power source, so there was no need to specifically control the base voltage. It was observed that the exponential slope of the black current versus voltage (0.75 decibels/volt) of the OLED display in the device without the stacked OLED was similar to the example with the OLED present. This suggests that the protection circuitry still provides some current and voltage control at the anode of the OLED when the base of the BJT is isolated. Without being bound by any particular theory or speculation, it is possible that a neighboring n-well (e.g., the n-well of the switching transistor T2) may be a source of holes that migrate to the p-well (base) of the BJT BJT1. Once these holes are in the base, they will diffuse across the depletion region to the n-type emitter contact (OLED anode pad), thereby traveling through the base emitter diode in the forward direction. This will be replenished by the diffusion of thermally excited electrons from the emitter into the base and their transport through the base and depletion region to the collector, the diffusion and transport being facilitated by the large electric field potential between the base and collector. In this case, holes from the neighboring n-well (of the drive transistor or the switching transistor) provide charge (holes) to the base which will usually come from the external base connection. When the OLED anode voltage drops to a very low level (e.g., to display black with three or more OLED cells), the potential difference between a neighboring n-well of one of the driver circuit transistors (which will be at V _DD ) and the BJT base (at the OLED anode voltage) is very large, increasing the hole flow from the driver transistor well into the BJT base. This increase in base current increases the emitter current due to amplification of the BJT.

However, the protection effect provided by the protection circuit is different for each frame and must be reset appropriately for each new frame of the image. This is not a problem when the base of the BJT is connected to an external power supply and actively controlled for each frame. For embodiments where the base of the BJT is isolated and not intentionally connected, since the switching transistor provides the switching effect, a reset for each frame can be provided when the switching transistor turns off the pixel. Therefore, this aspect of the protection effect control movement depends on having a drive transistor and a switching transistor connected in series, because they must turn off the transmission during data loading (or some part of the frame time) to reset.

OLED schemes with three or more cells can be designed so that the voltage range from black level (below Vth; e.g. 2uA/cm2) to white level (20mA/cm2) is relatively constant and less than about 6V. As shown in Figures 17 and 21, this can result in a contrast ratio of about 10,000:1 and can be slightly less, since the current efficiency drops at the high end of the current density. This voltage range is approximately within the permissible operating range of the LV transistors only when the protection circuitry comes into play at the bottom of the current range when the drive transistors and/or the switch transistors stop the current. Thus, at the low current end of the range, the protection circuitry additionally prevents the current density through the OLED from dropping below about 2 uA/ ^cm2 . Although this effect is provided with additional protection by having at least two series connected transistors as discussed above, it limits the ability to achieve higher contrast. In exchange for this slightly higher black level and reduced contrast, the protection circuit allows the pixel driver circuit to achieve higher peak brightness or compensate for efficiency loss due to OLED aging by reducing the cathode voltage to ensure that the LV transistor operates within its specified voltage range.

The above description describes several different embodiments, which may involve different combinations of different individual features. The individual features of any embodiment may be combined in any order or degree as desired without restriction, except when incompatible.

In the above description, reference is made to the accompanying drawings, which form a part thereof and in which specific embodiments that may be practiced are shown by way of illustration. These embodiments are described in detail to enable one skilled in the art to practice the invention, and it is understood that other embodiments may be utilized and structural, logical, and electrical changes may be made without departing from the scope of the invention. Therefore, the description of any exemplary embodiment should not be understood in a limiting sense. Although the invention has been described for illustrative purposes, it is understood that such details are for illustrative purposes only and that variations may be made by one skilled in the art without departing from the spirit and scope of the invention.

1: Pixel definition layer 3: Silicon backplane 5: Optional planarization layer 7: Electrical contact 9: First electrode segment 9A: First electrode layer 9B: Reflection layer 11: Non-luminescent organic light-emitting diode layer 13: First light-emitting organic light-emitting diode unit 13A: Red light-emitting organic light-emitting diode unit 15: Charge generation layer 16: Yellow light-emitting organic light-emitting diode unit 17: Second light-emitting organic light-emitting diode unit 17A: Green light-emitting organic light-emitting diode unit 19: Charge generation layer 21: Third light-emitting organic light-emitting diode unit 21A: Blue light-emitting organic light-emitting diode unit 22: Blue light-emitting layer 23: Non-luminescent organic light-emitting diode unit LED layer 24: charge generation layer 25: top electrode 27: capsule layer 29: color filter array 29B: blue filter 29G: green filter 29R: red filter 30: microcavity 32: second blue light emitting unit 45: power supply 50: power supply 100: RGB microdisplay/microdisplay 200: multi-mode microcavity organic light emitting diode stacked microdisplay/microdisplay 300: multi-mode microcavity organic light emitting diode stacked microdisplay/microdisplay 400: multi-mode microcavity organic light emitting diode stacked microdisplay/multi-mode organic light emitting diode microdisplay/microdisplay A- 1: 2-unit stacked organic light-emitting diode device/organic light-emitting diode A-2: second comparative organic light-emitting diode device/device/organic light-emitting diode B-1: inventive 3-unit stacked organic light-emitting diode/organic light-emitting diode device/display/organic light-emitting diode B-2: inventive 3-unit organic light-emitting diode device/organic light-emitting diode BJT1: bipolar junction transistor C1: storage capacitor C-1: inventive 4-unit organic light-emitting diode device/organic light-emitting diode D-1: inventive 5-unit organic light-emitting diode device/organic light-emitting diode D4: diode GND : Ground IBD1: Intrinsic body diode IBD2: Intrinsic body diode IBD3: Intrinsic body diode IBD5: Intrinsic body diode IBD6: Intrinsic body diode LSC: Level shift circuit MP1: Transistor MP2: Current drive transistor/transistor SELECT1: Select line SELECT2: Select line SELECT3: Select line T1: p-channel drive transistor/drive transistor T2: Switching transistor T3: DC select transistor/scan transistor T4: Third transistor/drive transistor T5: DC select transistor T6: Protection circuit p-channel transistor V _B : Base voltage V _C : Collector voltage V _CATHODE : Cathode voltage V _DD : External power supply/voltage source/constant supply voltage V _DD2 : Voltage source _VE : Emitter voltage/voltage V _REF : Voltage reference/reference voltage

FIG. 1 shows a simple prior art control circuit for an OLED. FIG. 2 shows a basic control circuit having two transistors with their equal channel series connection, the basic control circuit being suitable for an OLED stack having at least three OLED cells. FIG. 3 shows an embodiment of the intrinsic body diode connection of FIG. 2. FIG. 4 shows another embodiment of the intrinsic body diode connection of FIG. 2. FIG. 5A shows a basic control circuit having three transistors with their equal channel series connection, the basic control circuit being suitable for an OLED stack having at least three OLED cells. FIG. 5B is similar to FIG. 5A, except that the drive transistors are controlled via a level shift circuit. FIG. 6 shows a basic control circuit suitable for an OLED stack having at least three OLED cells and an additional protection circuit comprising a diode connected p-channel MOSFET transistor. FIG. 7 shows a similar protection circuit comprising a diode. FIG. 8 shows a basic control circuit suitable for an OLED stack having at least three OLED cells and a protection circuit having a bipolar junction transistor (BJT). FIG. 9 shows a schematic cross-section through a Si backplane illustrating connection locations, connections and transistor wells for the circuit shown in FIG. 8. FIG. 10 shows a cross-section of an RGB microdisplay 100 having three adjacent single color RGB OLED stacks, each individual stack having three OLED light emitting cells. FIG. 11 shows a cross-sectional view of a microdisplay 200 having a multi-mode microcavity OLED stack with three OLED light-emitting cells and an RGB color filter array. FIG. 12 shows a cross-sectional view of a microdisplay 300 having a multi-mode microcavity OLED stack with three OLED light-emitting cells including an additional blue light-emitting layer and an RGB color filter array. FIG. 13 shows a cross-sectional view of a microdisplay 400 having a multi-mode microcavity OLED stack with five OLED light-emitting cells and an RGB color filter array. FIG. 14 shows a graph of current density versus voltage for certain comparative OLED devices and inventive OLED devices. FIG. 15 shows the spectra of certain comparative OLED devices and inventive OLED devices (plotted as intensity versus wavelength). FIG. 16 shows a graph of the peak current density required to produce a D65 white point for each color of the comparative OLED devices and inventive OLED devices with color filters. FIG. 17 shows a graph of illumination versus voltage for the comparative OLED devices and inventive OLED devices when the comparative OLED devices and inventive OLED devices are driven to produce black and at peak white illumination. FIG. 18 shows a graph of current density versus voltage for certain inventive OLED devices. FIG. 19 shows the spectra of certain inventive OLED devices (plotted as intensity versus wavelength). FIG. 20 shows a graph of the peak current density required for each color to produce a D65 white point for certain inventive OLED devices with color filters. FIG. 21 shows a graph of illumination versus voltage for certain inventive OLED devices when they are driven to produce black and at peak white illumination.

SELECT1: Select line

SELECT2: Select line

T1: p-channel driver transistor/driver transistor

T2: Switching transistor

T3: In-line selection transistor/scanning transistor

V _CATHODE : cathode voltage

V _DD : External power supply/voltage source/constant supply voltage

Claims (19)

A microdisplay includes a light-emitting OLED stack located on top of a silicon-based backplane, the silicon-based backplane having individually addressable pixels and control circuitry, wherein: the light-emitting OLED stack has three or more OLED units located between a top electrode and a bottom electrode; and the control circuitry of the silicon-based backplane includes at least two transistors for each individually addressable pixel, the channels of the at least two transistors are connected in series between an external power supply ( _VDD ) and the bottom electrode of the light-emitting OLED stack, wherein the at least two transistors with their equal channel series connection are rated at 5V or less. A microdisplay as in claim 1, wherein a threshold voltage (V _th ) of the light-emitting OLED stack is at least 7.5V or greater than 7.5V. A microdisplay as in claim 1, wherein a threshold voltage (V _th ) of the light-emitting OLED stack is at least 10V or greater than 10V. A microdisplay as claimed in any one of claims 1 to 3, wherein the light-emitting OLED stack comprises four or more OLED units. A microdisplay as claimed in any one of claims 1 to 3, wherein the OLED units are each separated from each other by a charge generation layer (CGL). A microdisplay as claimed in claim 5, wherein the bottom electrode is segmented and each segment is electrically connected to the control circuit system in the silicon-based backplane. A microdisplay as claimed in claim 6, wherein the light-emitting OLED stack is top-emitting. A microdisplay as claimed in claim 7, wherein the light-emitting OLED stack forms a microcavity in which the physical distance between the segmented bottom electrode and the top electrode is constant across all pixels. A microdisplay as claimed in any one of claims 1 to 3, wherein one of the at least two transistors closest to the external power source is a drive transistor and is rated at 5V or less, and another of the at least two transistors closest to the bottom electrode of the light-emitting OLED stack is a switch transistor and is rated at greater than 5V. A microdisplay as claimed in any one of claims 1 to 3, wherein the at least two transistors having their equal-channel series connection are both p-channel transistors. A microdisplay as claimed in claim 10, wherein the at least two transistors having their equal-channel series connections are each located in a separate well. A microdisplay as claimed in any one of claims 1 to 3, wherein the control circuit system further comprises a protection circuit comprising a p-channel transistor. A microdisplay as claimed in any one of claims 1 to 3, wherein the control circuit system further comprises a protection circuit, the protection circuit comprising a p-n diode. A microdisplay as in claim 13, wherein the cathode of the pn diode is connected to the node of the bottom electrode of the light emitting OLED stack, and the anode is connected to a voltage reference (V _REF ) or a current reference. A microdisplay as claimed in any one of claims 1 to 3, wherein the control circuit system further comprises a protection circuit, the protection circuit comprising a bipolar junction transistor. A microdisplay as in claim 15, wherein the bipolar junction transistor is an NPN transistor, in which the base is connected to a voltage source (V _PROTECT ) or a current source, the emitter is connected to a node connected to the bottom electrode of the light-emitting OLED stack, and the collector is connected to the external power supply (V _DD ). A microdisplay as in claim 15, wherein the base of the bipolar junction transistor is isolated from an external power source, the emitter is connected to a node connected to the bottom electrode of the light-emitting OLED stack and the collector is connected to the external power source (V _DD ). A microdisplay as claimed in claim 15, wherein the bipolar junction transistor is located in a well separated from the at least two transistors having equal channel series connection therewith. A microdisplay as claimed in claim 18, wherein the at least two transistors having their equal-channel series connection are both p-channel transistors and are each located in a separate n-well, and the bipolar junction transistor is an NPN transistor located in a separate p-well.