Flexible Electronics: Potential And Reality

July 11, 2012, FlexTech Alliance, San Francisco—David Allee from the Flexible display Center at Arizona State University described the past and current efforts in flexible display research.

As TFTs move from amorphous silicon to metal oxide as the base material, the challenges of higher levels of integration become clear. The InGaZnO materials are mostly settling in on a 8-mask process. One of the problems is that the materials are depletion–mode and have little ability to make act as loads.

To increase the level of integration, the industry needs CMOS. P-type devices in organic materials are under research at the University of Texas-Dallas. Some potential applications for a CMOS TFT process include OLED displays on PEN, FPGAs, memories, and source drivers.

N-MOS only designs on flexible substrates like plastic and steel that don’t need an active load would continue to be available. One way to move towards higher integration is to mix technologies and use amorphous silicon for A/D and D/A converters. Adapting CAD tools to use multiple materials is one roadblock. Another is that a-Si has significant wear-out characteristics.

The most desirable circuit functions are op-amps to perform the signal conditioning needed for the sensors associated with flexible displays. This type of circuit requires CMOS or enhancement and depletion devices to meet the performance requirements. A typical op-amp might have a 1 MHz unity gain bandwidth, 100 dB gain, 1 mW total power consumption, and operate on +/- 20 V supplies and provide a +/- 12 V output swing.

The metal oxide TFTs are much more stable than a-Si, but currently cannot meet any of the performance requirements. They cannot perform high-sped computation nor can they meet the needs of even medium rate data communications. One reason for the disparity in performance between silicon and the MO devices is size. The feature size of a TFT is is over 200 times the size of a Si device, over 5 microns versus 0.28 nm. Higher speeds and density come at the price of much greater manufacturing complexity.

They can, however, do well in functions requiring a billion transistors per square inch, and are well suited for displays and sensor arrays. Any function needing large area coverage is a good target application for MO TFTs. Large area displays and sensors are the obvious targets, and e-ink functions follow closely behind. Advances in technologies may enable e-ink displays to also function as sensors. At a macro level, flexible arrays could be used as focal plane correction tools. A concave surface eliminates spherical aberrations, while a concave surface can act as a multi-mode wide view imager.

To get to these functions, the technology has to advance to enable active pixel amplifiers. The real-time corrections will need optimizing algorithms to correct for non-linearities and optical aberrations. Sensor arrays could merge electronics and chemical sensing to act as high-sensitivity electronic noses for process and environmental monitoring as well as bio-threat analysis.

Another area where flexible substrates with some active devices provide benefits is in patch antennas. An actively tuned flexible antenna could improve mobile device operation by retuning the antenna as the local loading changes. For example, rotating the phone from vertical to horizontal would change the antenna tuning.

For really flexible substrates, one area of interest is in electronic fabrics. Any circuits on these materials will need an architecture that does not require x-y connections. Washing the fabric could be a challenge. Fabric-based electronics could host sensors and displays for applications as frivolous as a t-shirt that lights up when you like someone, to monitoring emergency workers for stress and hydration while in the field. Combining the MOTFTs with MEMS could result in an ultrasonic transducer or transmitter on PEN.

The applications for MOTFTs are now limited by the device performance. If the TFTs come close to matching Moore’s law, the future applications will be limited mostly by imagination.