12-07-2021 | | By Sam Brown
Recently, researchers from Stanford have demonstrated a new flexible artificial skin that integrates up to 40,000 transistors per square cm of material. What challenges do flexible electronics face, what have the researchers demonstrated, and could this be the final stepping stone to truly flexible electronics?
While many examples of flexible electronic circuits can be stretched and manipulated, none of these has proven to be practical. For example, flexible electronic circuits such as RFID tags commonly found on product packaging are actually flexible antenna traces connected to a ridged semiconductor. Some companies such as PragmaticIC have demonstrated flexible processors, but these are yet to enter the market, and these flexible processors are still to be demonstrated as being fully functional.
The reason that truly flexible electronics are hard to manufacture comes from the very fact that many materials don’t like to be flexed or stretched. While a transistors structure may be relatively simple, its minute size and careful use of multiple layers (such as gate and insulation) mean that they are vulnerable to damage. Silicon chips function well because their solid structure ensures that such layers do not move or change shape. Furthermore, a transistor whose dimensions change (i.e. as a result of stretching), will undoubtedly suffer from changes in characteristics such as channel resistance and gate capacitance.
Many materials used to create modern electronics also suffer from fatigue and damage when stretched and manipulated. As such, trying to use ridged materials on flexible substrates will eventually succumb to cracking which will destroy the device.
It is these reasons why researchers are trying to find flexible materials that exhibit semi-conductive properties. Of course, simply finding such materials is only half the solution; those materials must also be practical to manufacture otherwise they will remain laboratory novelties with no useful purpose.
Recently, researchers from Stanford have demonstrated an artificial skin that is flexible in all three dimensions and holds flexible electronic circuits. The skin, which has taken almost two decades to develop, allows for extreme manipulation while returning to its original size and shape with all electronic components in the skin remaining functional.
According to the researchers, the skin-like material is able to integrate more than 40,000 transistors per square cm of material. While this may pale in comparison to modern billion-count devices, 40,000 transistors are more than enough to incorporate an 8-bit CPU with a small amount of memory, sensors, and communication circuitry. Furthermore, it should be considered that this figure is 40,000 per square cm, and thus large circuits could be integrated into the material with a 10cm square piece of material having 400,000 transistors.
The researchers claim that their development increases elastic-transistor density by more than 100 times compared to what other researchers have achieved. However, it should be noted that there is a big difference between transistor density and actual transistor count. For example, the researchers may have squeezed 10 transistors into a very small space that all remain functional, but while this density would equate to 40,000 per cm, it could actually just be 10 transistors in a small space.
The new artificial skin with flexible electronics can be manufactured using standard practices found in semiconductor foundries including the use of UV lithography to pattern electronic components. However, while traditional chemicals would normally destroy such polymer substrates (i.e. the artificial skin), the researchers have developed new photochemistries that leave the polymer skin intact after layering and etching of materials.
According to the researchers, the transistors developed have the same electrical performances as those found in computer displays (i.e. TFT displays), meaning that individual transistors are practical for everyday circuitry. Furthermore, the transistors are still functional after 1,000 stretches which suggests that the transistors developed by the researchers could be close to becoming a commercial reality.
The high degree of flexibility of the skin coupled with the high transistor density may suggest that what the researchers have developed at Stanford could be the final stepping stone to creating truly flexible circuits. Not only are the transistors flexible and functional, but they are also easily manufactured using similar processes already found in semiconductor foundries around the world and thus could be easily commercialized.
However, there are some challenges that the researchers have to face. The first is proving the transistor density of the new skin. While a 40,000 per cm squared figure sounds promising, the researchers would have to demonstrate a 40,000 transistor device operating. As previously mentioned, having a 40,000 cm square density only suggests how many transistors can fit per unit area as opposed to the number of transistors that will reliably operate per unit area.
As such, for the researchers to prove the functionality of their artificial skin, it may be a good idea to integrate a computing system complete with memory and I/O. The resulting device does not need to be complex, but the demonstration of 80,000 transistors operating as expected while undergoing deformation would be a major feat of engineering.