Don't Count Out Field-Emission Devices for Large-Area Displays

FotoVitor050128v1_tif

by Victor Pellegrini Mammana

It is now clear that the efforts to develop field-emission displays (FEDs) over the last decade did not succeed in demonstrating a consistent or successful solution for commercial fabrication. Hasty decisions to produce FEDs based on metal-tip arrays drew investments to the wrong technology too early. So, despite the large amount of money invested in metal tips during the '90s, FEDs never made it to market. Nonetheless, a great deal of excellent research was done by the determined and talented researchers involved in those efforts, and the new knowledge that was attained has become the background for an entirely new cycle of development.

The limited success of the early development efforts is not evidence of something fundamentally wrong with the concept of commercial FEDs, although it is an indication of fundamental flaws in the way high-tech investment decisions were made in those times of abundance. Other high-tech blunders of the '90s – from satellite constellations to Internet start-ups – became obvious after the technology bubble burst, and they helped dismantle the myth that the business-planning methodology of the time was a substantial guarantee of success.

The New Wave

New advances in nanotechnology have the potential to turn FEDs into a profitable business, particularly for the production of large-area displays. But what makes the new approaches more promising than the old ones?

Because the many failed FED start-ups and technology investments left a fragmented and scattered technical record, it is hard to draw a clear picture of what prevented these companies from bringing FEDs to market. Nonetheless, we know that some of the main technical issues were lifetime (which is related to contamination and insufficient getter pumping efficiency), spacers, luminance uniformity, and blue-phosphor efficiency. Any new effort to establish the commercial feasibility of FEDs must address such issues.

Some groups say that lack of robustness is an intrinsic characteristic of metal tips, while others continue to claim that very good lifetimes are possible. No matter which claims are correct, the cost-effectiveness of metal-tip fabrication was and is a high barrier. Small, high-quality LCDs are a commodity, resulting in a hegemonic technology difficult to beat. Consequently, it would be very difficult for metal-tip arrays to compete with small LCDs even if their lithography were not based on costly steppers or unconventional solutions such as ion tracking.

The difficulty of competing with small, inexpensive LCDs actually provides a glimmer of light. Large LCDs are still expensive, so why not apply metal-tip-based FEDs to large-area displays, where the technology might be better able to compete. Indeed, FEDs do not demand an active matrix, which represents a cost advantage with respect to large-area AMLCDs. But this fact does not give much help to metal-tip technology because metal tips are very difficult to fabricate over a large area as a result of intrinsic limitations on uniformity.

It is now widely accepted that a competitive FED technology requires a process that can form addressable pixels over large areas without demanding high-resolution patterning techniques. Of course, this is true not only of FEDs but of any display technology that would compete with large-area AMLCDs.

The lithographic formation of metal tips for FEDs was an early application of nanotechnology using a tap-down approach. Bottom-up approaches, such as the synthesis of carbon nanotubes, promise a less-demanding FED lithography that is easier to implement over large areas.

One such application is the use of carbon nanotubes (CNTs). As demonstrated by Motorola and Samsung, the use of CNTs as emitters allows a distance between the cathodeemitters and grid of about 10 times larger thanin the case of metal tips (for the same operating voltage). Larger cathode–grid distances implya much smaller lithography process, with fea-tures in the 10-μm range instead of submicron. This makes FEDs easier to fabricate in larger areas with the possibility of substantially lower costs than those of PDPs or LCDs.

Several groups around the world are implementing nanostructured materials as emitters in FEDs. This includes our own effort, which relies on a novel approach for the formation of the pixels and a new geometry that permits much more efficient vacuum pumping, therefore providing a cleaner environment inside the emitting chamber. This work is supported by a major international company which is also collaborating with us.

We form our emitting sites by punching small holes in a multilayer substrate and applying a CNT or conductive-oxide coating to the raised edges for enhanced emission. Not only does this produce surprisingly efficient emitters, but it does so by fabrication techniques that are easy to integrate into industrial processes. However, the original motivation for investigating this approach was to improve the efficiency of getter pumping along a short path through the holes and perpendicular to the substrate instead of along a long path in a narrow layer parallel to the substrate (see, for example, the award-winning paper by H. Buikema et al., "Calculation of Deflection for Porous Polyimide Membranes in FEDs," Asia Display/IDW '01).

In this configuration, as well as others, certain conductive oxides are especially attractive as substitutes for CNTs because of their intrinsic stability when interacting with residual gases containing oxygen. Furthermore, the synthesis of these oxide nanostructures is thermally compatible with other materials used to form the devices and does not demand binders or plasma processing. This increases cleanliness and simplifies fabrication. We are focusing on applying this configuration to displays measuring 50 in. on the diagonal and only a few centimeters thick.

Is this another FED pipedream? We do not think so. FEDs reached a low point of interest about a year ago, but interest is once again increasing. Some examples are described below.

• Canon and Toshiba have formed a joint venture, SED, Inc., to produce surface-conduction electron-emitter displays (SEDs) "with production scheduled to begin in 2005."

• A major developer of high-voltage display drivers reports an increasing number of inquiries from participants in several FED-development projects.

• The HOPFED technology recently transferred from Philips Research to Southeast University in Nanjing is receiving support from at least one Asian company.

• At the 11th International Display Workshops held in Niigata, Japan, in December 2004, an active FED Workshop included reports of work from ERSO/ITRI (Taiwan), Mitsubishi Electric Co. (Japan), KIST (Korea), Iljin Nanotech (Korea), Sony EMCS (Japan), Futaba (Japan), NGK Insulators (Japan), Tottori University (Japan), Kyung Hee University (Korea), Ecole Polytechnique (France), TECO Nanotech (Taiwan), ULVAC (Japan), National Institute of Advanced Industrial Science and Technology (Japan), University of Bucharest (Romania), Applied Nanotech (U.S.A.), Alps Engineering (Japan), and Dialight (Japan), among others.

A general intensification of investment in nanostructure-based FEDs is an indication that FEDs may soon contribute to a radical change in large-area-display manufacturing. •


Victor Pellegrini Mammana is with the FED Project, CenPRA/FAPESP, Rodovia D. Pedro I, km 143,6, 13082-120 Campinas, SP, Brazil; telephone +55-(19)-3746-6045, fax +55-(10)-3746-6051, e-mail: victor.mammana@cenpra.gov.br.