To build a display that meets high performance expectations, panel makers require a thermally and dimensionally stable glass to improve yields while achieving the desired resolution.
The display industry is continuing to move toward mid-to-large-size, immersive displays in high-performance tablets, notebooks and 8K TVs. As these trends become industry standards, the oxide market emerges as an important opportunity for enabling the next-generation of high-performance displays. These displays feature: higher resolution and faster refresh rates; enhanced circuitry integration to achieve slim bezels; and cost savings for panel makers by improving the panel aperture ratio and enabling large gen size manufacturing.
To achieve these technical requirements, new breakthroughs are needed in thin-film-transistor (TFT) technologies. Among the display industry’s current offerings, amorphous silicon TFT (a-Si TFT) maintains a leading position among all applications, while low-temperature poly-silicon TFT (LTPS) is the predominant display technology for enabling high-performance handheld displays. The key differences between a-Si and LTPS are that an a-Si TFT has a simpler process, structure, and is easier to scale up in terms of manufacturing. However, LTPS offers better TFT performance to achieve higher resolutions and lower power consumption. The drawbacks of LTPS come in size limitations and increased manufacturing costs. For these reasons, neither a-Si or LTPS can fully meet the technical requirements for this next generation of high-performance displays.
As a result, an industry need arises for a glass substrate that is precisely engineered to enable the higher pixel density of high-performance displays that panel makers require to meet consumer demand for brighter, faster, more lifelike images. To build a display that meets these performance expectations, panel makers need a thermally and dimensionally stable glass to improve yields while achieving the desired resolution.
All of these industry requirements create new process and glass composition challenges, which present the need to develop an advanced oxide TFT glass technology.
For decades, the dominant technology for flat panel displays was an amorphous silicon (a-Si) backplane. The vast majority of displays were made using a-Si backplanes due to the simplicity in manufacturing process, good economics, and scalability to larger sizes. As demands for brighter and/or higher resolution displays grew due to the introduction and proliferation of handheld mobile devices, alternative backplane technologies, such as low temperature polysilicon (LTPS), became more prevalent. LTPS is similar to a-Si, but requires higher processing temperatures and a more complicated manufacturing process. This results in advanced properties for the backplane, such as >50X higher electronic mobility. These properties allow smaller TFTs (enabling higher resolutions and brighter displays) and faster refresh rates. While clearly a superior technology to a-Si, the higher temperatures and more complex manufacturing process make LTPS considerably more expensive than a-Si. Additionally, LTPS is not easily scaled to larger sizes to enable better panel economics.
The ideal backplane technology would combine the simplicity, economics, and scalability to larger panel sizes of a-Si with the heightened performance of LTPS. This is exactly what oxide TFT technologies offer. The most commonly implemented oxide TFT technology is based on Indium-Gallium-Zinc-Oxide or “IGZO” technologies.
Though the mobility of oxide TFT is not as high as LTPS, it is an order of magnitude better than a-Si technology and capable of driving OLED displays and 8K 120Hz + LCD TVs. Additionally, the low off-current of an oxide TFT could enable low refresh frequency without flicker effects on static images (a comparison of different TFT technologies are shown in Table 1). While, like LTPS, oxide TFT backplanes have improved electrical properties relative to a-Si backplanes, oxide TFT backplanes can scale up to Gen 10.5 at reasonable costs (unlike LTPS), thereby enabling high-end, large-size LCD and OLED TVs. It is for this “just right” compromise of a-Si and LTPS properties that oxide TFT is garnering so much attention from panel makers worldwide. It offers the ability to manufacture displays far superior to a-Si at sizes and costs unachievable by LTPS.
Oxide TFT Process
There are two major oxide TFT processes to consider: etch-stop and back channel etch (BCE). The key difference between the processes is the use of an etch-stop layer, also known as ESL, that is required to protect the IGZO channels during the etching process.
Etch-Stop Oxide TFT Process
Oxide TFT reliability was the major concern in early stage of oxide TFT development. The oxide TFT channel was usually damaged in subsequent processes, so an etch stop structure was designed to protect the oxide TFT channel. The etch-stop (ESL) oxide TFT manufacturing process begins with a bottom gate structure which is covered by a gate insulator and TFT islands. After the gate insulator (GI) layers and TFT patterning, a patterned SiO2 layer is deposited to cover the IGZO channel area in order to protect oxide TFT from following source/drain (S/D) etching. This enables better TFT reliability, and after the S/D etching, then followed by passivation, ITO layer as the Figure 1 shows. In the ESL process, temperatures may go up to 300-400°C for up to an hour or more. While these are higher temperatures than some a-Si processes, it is considerably lower than the typical LTPS processes that can exceed 500°C.
BCE Oxide TFT Process
The BCE oxide TFT process (Figure 2) is very similar to the ESL oxide TFT process in the first two photo etching processes (PEP) steps. However, a high temperature (400-500°C) annealing process enhances the TFT reliability that allows the removal of the ESL. The higher temperature annealing step requires a thermally stable glass that can withstand harsh manufacturing environments and processing times relative to the conventional oxide (ESL) or a-Si processes.
To panel makers, the BCE oxide TFT process is similar to the a-Si process, which has been widely used for the past two decades. Also, there is one photo-mask process reduction compared to the ES oxide TFT process, therefore, BCE oxide TFT is becoming a mainstream process of oxide TFT manufacturing.
While the oxide TFT process has clear technical benefits for the manufacture of large and high-performance TVs, it presents a unique set of challenges for the glass substrate used in the process.
When put through a typical TFT backplane process, glass substrates will change shape or size (i.e., strain) which is called a change in total pitch (TP). One of the most important glass substrate attributes is total pitch variation (TPV), which is the deviation from predictable glass movement within a glass sheet and from sheet-to-sheet. For a glass substrate to have good TPV performance, the substrate must have the required balance of physical properties to resist the various causes of strain of the substrate: elastic distortion, stress relaxation, and compaction. These sources of strain, and the corresponding glass property that resists them, are discussed below.
In TFT processes, there are several sources of stress applied to the glass substrate, such as film stresses and gate metals. In oxide TFT, the latter is particularly significant due to the substantial thickness and covered area of the gate metal. The pitch change associated with these stresses is determined by the size of the stress, the elastic modulus of the glass, and the thickness of the substrate. Since the stresses are determined by the TFT manufacturer and the industry is continually driving to thinner and thinner substrates, the only attribute within the control of the glass manufacturer is to increase the elastic modulus to increase the stiffness of the substrate. Also, because the stresses in the TFT process can vary across a sheet or sheet-to-sheet, a higher elastic modulus will reduce the strain due to variations in the applied stresses, thereby minimizing TPV from this potential cause.
The stresses from applied films and gate metal can also contribute to the overall TPV through the relaxation of those stresses during subsequent thermal treatments. As the substrate progresses through the various steps of the TFT process, the films, gate metal, and substrate itself will all undergo stress relaxation. As the stress state of the composite changes with time and temperature, the concomitant strain will accordingly change, causing a pitch change and an increase in TPV. The glass substrate resists this stress relaxation in proportion to its effective viscosity at the process temperatures. In a-Si TFT processes, the temperatures are low enough that there is a minimal amount of stress relaxation due to the glass substrate having a relatively high viscosity at these low temperatures (the viscosity of the glass increases as the temperature decreases). In oxide TFT processing, however, temperatures are higher and, therefore, the potential for stress relaxation is greater due to the lower effective viscosity of the glass. This is particularly acute for the BCE oxide TFT process, which has process steps with temperatures in excess of 400°C. Traditional glass substrates which are sufficient for the typical a-Si applications may also be sufficient for the lower temperature ESL oxide TFT processes. However, the higher temperature BCE oxide TFT process may require a substrate with a higher effective viscosity at temperatures in the range of 400°C.
The effective viscosity of the glass substrate also plays a role in the amount of viscous relaxation the glass substrate undergoes in the TFT process due to structural relaxation of the glass itself. This is commonly referred to as “compaction” or “shrinkage” in the glass industry. Compaction is due to the evolution of the glass structure from a non-equilibrium state toward a structure closer to equilibrium with the customer process. The amount of this viscous relaxation that occurs is proportional to the degree to which the glass is out of equilibrium, and inversely proportional to the effective viscosity of the glass at the TFT process temperatures. Consequently, a higher viscosity glass is beneficial for minimizing TPV, just like in stress relaxation. In glass property terms, a higher viscosity glass is a glass with a higher “annealing point” therefore glass manufacturers will often tout the high annealing point of their glass compositions.
Glass sag typically occurs when a large sheet of glass is supported horizontally by its edges and allowed to naturally bend due to its own weight. Sag increases the process challenges on larger glass handling and uniformity in OLED evaporation process. The amount of this sag is proportional to the glass density and inversely proportional to the elastic modulus. The elastic modulus represents the glass’ capability to resist deformation in the manufacturing process, while low density allows for a light weight sheet of glass. This ratio of elastic modulus and density determines the amount of sag in the glass, with a higher ratio (higher modulus and/or lower density) leading to less sag and better performance.
Total Thickness Variation
Measured in microns, total thickness variation, or TTV is variation of the glass thickness over a defined area of the glass sheet. Compared to glass produced on a float platform, Corning’s proprietary fusion process creates glass with some of the industry’s lowest TTV levels.
By improving TTV, panel makers have the benefit of uniform layer thickness during deposition on the glass substrate and precision patterning in photolithography process. This is especially important from the exposure process perspective, because control of the field of focus is crucial. If the TTV of the glass is outside the field of focus in a moving window range (MWR), a crisp pattern cannot be obtained (Figure 3). Lower glass substrate TTV therefore provides a significant advantage in the precise photo-lithography steps needed for high-resolution displays.
Enabling Large Gen Sizes
Screen-sizes are continuing to grow and increase, creating new challenges for panel manufacturers to increase yields, maximize throughput, and reduce material costs. This makes glass utilization increasingly important to panel makers. Therefore, the glass substrate must enable efficient manufacturing and scale up to larger gen sixes (Gen 8.5 and above).
Corning’s proprietary fusion process manufactures glass panels at Gen 10.5 sizes (2940 x 3370mm), enabling higher glass utilization for larger-screen sizes. For example, one sheet of Gen 10.5 glass could create eight 65” display panels, or six 75” display panels. This enhanced glass utilization greatly reduces cost for panel makers and is key for enabling the oxide TFT market'
Balancing Fast Etching and Sludge Generation
For oxide TFT to be used in IT or handheld products, one of the key features is a thin and light form factor. To achieve this, the display panel usually needs to be thinned down to roughly 0.15mm / 0.15mm (for the two pieces of glass in the display) using the chemical slimming process. A faster etch rate is clearly desired to enable higher throughput and lower costs but this often comes at the cost of the generation of “sludge.” Sludge can create problems in the etch vendors’ processes and end up causing more cost than the fast etch rate reduced. By using a glass that balances maximizing etch rate while minimizing sludge generation, panel makers optimize their throughput and costs.
The technology challenges and technical requirements outlined fuel an industry need for a new glass substrate with the right balance of physical properties for oxide TFT technology. For displays applications, this includes low total pitch variation, low total thickness variation, and low sag. This package of glass attributes, alongside the ability to scale-up manufacturing to large-gen sizes, will help enable the next-generation of mid-to-large-size, immersive displays in 8K TVs.
In the case of high-performance notebooks, tablets and other handhelds, fast etching and minimal sludge generation become increasingly important glass attributes for better picture quality and response times.
These applications require a shift toward oxide technology, versus the current a-Si and LTPS TFT technologies. As the push for oxide increases, new process and technical challenges emerge for panel makers. To build a display that meets these performance expectations, panel makers require a thermally and dimensionally stable glass to improve yields while achieving the desired resolution.
— TJ Kiczenski, Senior Research Associate and Business Technology Manager, High Performance Displays and PH Su, Senior Project Manager, High Performance Displays, Corning Glass Technologies.