A Short History of Backplane Technology
by Adi Abileah
Backplanes for LCDs have experienced a long and very crucial technological evolution in the last 20-plus years. Before the adoption of active matrix (AM), the direct driving of LCD pixels or segments was the standard driving mode for all LCD devices. Today, it is still used for some low-information-content displays, typically those with alphanumeric characters. However, device developers were driven to switch to AM addressing when information-content requirements went up. Along with the need for an AM came the need for a switching device located inside the pixel matrix on the backplane of the display. Such a device allowed for a full image to be displayed and almost eliminated the partial driving of unselected pixels.
In the late 1980s, amorphous silicon (a-Si) became the overwhelmingly popular choice for switching devices. This was a derivative of a-Si development for solar cells. In the early 1990s, several pioneers in a-Si TFTs for LCDs, such as OIS, IBM, NEC, Sharp, and a few others, demonstrated products based on a-Si. A contender to a-Si at that time was CdSe (cadmium selenide), promoted by the late Dr. Peter Brody. However, the a-Si process was proven to have better manufacturability. Another debate in those days was whether to use back-to back diodes or a transistor as a switching device. The industry settled on the a-Si thin-film transistor (TFT).
Beginning in the early 90s, a-Si TFT was used for all LCD manufacturing. Its process resulted in very high yields and was expandable to large glass substrates. The limitations of a-Si are relatively low mobility and some instability of the current–voltage curve. For an LCD, the instability is not critical while the device is used as a switching device.
Now, enter OLED technology for large-area displays. As for LCDs, a backplane and AM-addressing architecture are also used to achieve high information content. Unlike LCDs, the OLED pixel requires a continuous supply of electrical current to emit light, and therefore the switching devices must do much more than simply select and store a voltage between refresh cycles. Consequently, the driving circuits for OLED pixels are more complex and require several switching components in each cell as well as much higher mobility and stability. It has been widely reported that a-Si is not a suitable solution in this case. As a result, a new and more urgent focus has been placed on developing advanced structures from several different materials, including crystalline-silicon structures or polysilicon. This process is more complex and requires the annealing of the silicon layer on the glass. Sometimes this is done with localized exposure of an excimer (UV) laser. As of now, several manufacturing plants are dedicated to the polysilicon process. Almost all are using the low-temperature process with laser annealing and several processes are in commercial use today.
However, there are other areas of research, including oxide semiconductors, that have shown a great deal of promise and similar commercial adoption. In this issue, we examine this new technology, more specifically indium gallium zinc oxide (IGZO). Work on IGZO started almost two decades ago in Japan and in the labs of
Professor John Wager at Oregon State University. The technology is now becoming mature and is being integrated into manufacturing. We are fortunate to have an article by John Wager in this issue of Information Display, explaining the technology and comparing it to a-Si [hydrogenated a-Si (a-Si:H)] and polysilicon [or low-temperature polycrystalline (LTPS)] processes. John’s article is well-explained, and I hope that you will enjoy reading it. It nicely outlines the tradeoffs that are necessary as new materials and processes continue to develop and impact each other. •