The Growing Promise Of Printed Electronics

Printing electronics using conductive ink rather than lithography is starting to move out of the research phase, with chipmakers now looking at how to commercialize this technology across a broad range of sensor applications.

Unlike traditional semiconductors, which use tiny wires as circuits, printed electronics rely on conductive inks and often flexible films, although they can be printed on almost anything. That allows them to be flowed into places using conformal films, or to be taped onto something like an industrial valve where multiple types of sensors can be used rather than trying to attach a single, discrete sensor on a hard substrate. Moreover, these sensors can be added to existing facilities, rather than replacing existing devices with new parts that contain built-in electronics.

This kind of capability—conductive ink with built-in RF or connected circuitry—has captured the imagination of chipmakers on multiple continents and in multiple industry segments. Most experts believe they have only scratched the surface for applications, too.

“There are categories of things where printed electronics do very well,” said Dan Brewer, executive director at Brewer Science. “Anytime you need a flexible form factor that can be adjusted to fit a particular shape, this works great. It also allows an operation to benefit from lots of big data analysis because you can generate data from everywhere.”

The promise of printed electronics is hardly a new one. In fact, it’s been the subject of serious semiconductor research since the beginning of the millennium. There has been talk about conductive threads inside of clothing, or even circuitry inside of building materials. But all of this has proven more difficult to develop than initially thought.

“This has been on our agenda for the past 10 to 15 years,” said Janos Veres, manager at PARC (formerly Xerox PARC, or Palo Alto Research Center), whose focus is to bridge long-term research normally done in universities or research houses with commercial applications of that research. “There are three things we envision. One is new form factors, including 2D devices. The second is integration of sensors, including optical and mechanical components. So instead of building these devices in tiny dimensions, you can spread them out. The third is on-demand, local, distributed manufacturing. So rather than building up inventory, you can guarantee to keep them fresh in the supply chain.”

Fig. 1: Xerox printed micro-circuits on a label. Source: Xerox

Applications
It’s the possibility of adding these sensors across a variety of existing and new applications that is really driving this technology approach.

“If you think about a three-dimensional sensor today, you may have a humidity sensor or a temperature sensor,” said Gert Jorgensen, vice president of sales and marketing at DELTA Microelectronics. “But with printed electronics you can have both of those and much more. So you can develop an ink where it could be applied to a protein or a non-organic film, and you can use that to sense things that are not detected today by a diode.”

In fact, the Danish government has a small research program underway to track waste water and clean water using printed sensors.

“This could be used to determine food quality, too,” said Jorgensen. “You can measure carbon monoxide, carbon dioxide, different chemicals and metals. You can glue this on a microcontroller and use it to detect bacteria. We take air samples today. This could be cheaper and faster. And today with diseases, you need to grow a culture in a lab. That could take a week. And there are still diseases you cannot see that way, such as influenza or bacteria that cause food poisoning. You may be able to detect that with ink. We’re at the research stage today. There are prototypes of inks, which are not being made yet in high volume, that can be made sensitive to light, bacteria, and even corrosion on screens in industrial applications. And if you can get inks to change color, you can read them easily with image recognition technology.”

Industrial applications always have been an obvious application. Using silicon or any other hard substrate makes it difficult to design into these devices. For one thing, many of these operations are unique, so building a part with a built-in sensor requires a customized chip. That significantly adds to the cost of the part. In addition, these chips need to be able to operate in harsh environments in which chemicals, heat, vibration and extreme cold are common. It’s much easier to attach circuitry printed on a piece of tape or in a conformal film that is designed to withstand these conditions, and it can provide significantly more data than a discrete sensor.

“If you think about a butterfly valve, this allows you to determine what’s happening inside of the valve,” said Brewer. “You can have 30,000 data points per second for each sensor, and the opening/closing time may be 2 or 3 seconds, so you can tell exactly what’s going on. That allows you to do predictive maintenance, which was where this all started. But now you can do it for every valve and pump in an operation.”

This is a key reason why this approach is attracting so much attention in the automotive world, as well.

“Today, you’ve got sensors that are basically glued together, and you’ve got miles of wires in a car,” said PARC’s Veres. “In automotive, they’re looking at integrating structures into a laminate to create new designs. So you may have panels overhead, and you can print 2D circuits and add a conformal piece onto or into those panels. It’s the same for wearables. You can have bio sensors that can sense more than 100 different biomarkers using foil-like structures.”

Veres noted that sensors also can be integrated into warehouses and product packaging for supply chain monitoring, which would dramatically reduce the cost over current solutions.

Another application is to integrate printed electronics with 3D printing, particularly in areas such antennas. “3D printing did not take off as we had hoped,” said Stefanie Harvey, R&D program manager for flexible technology at SEMI. “But we are starting to see new materials being used here for things like antennas. What we’re also finding is that additive manufacturing doesn’t have to be a nozzle or ABS (acrylonitrile butadiene styrene, a thermoplastic polymer). You can have reproducibility and conformity and resistors at interesting geometries. There are lots of advancements being made.”

Technology
So far, most of the printed electronics have been used in flexible hybrid electronics (FHE), typically at large dimensions—something on the order of 1 micron. That is changing rapidly, however.

“ITRI (Industrial Technology Research Institute) in Taiwan has developed technology at about 0.25 microns,” said Brewer. “You cannot print at the level of litho processes, so you’re probably not going to want to use this to print memory or logic at a level that’s done today. But in the next 10 years, we’re going to see fractions of a micron. There also are a couple a main methodlogies for putting down inks, which are inkjet or screen. That could change. And there are limitations at these geometries to the types of structures that can be used. In the future, a lot of this will be driven by materials, and that’s where we need to start. If we have new materials, we can have new unique geometries, so that anywhere you need a structure where the form factor needs to be adjusted, this will work.”

Getting to that point requires a mix of disciplines, ranging from electrical engineering to material science, chemistry, and nanoparticle research.

“There is a lot of ink formulation work to achieve copper interconnect-like performance,” said Melissa Grupen-Shemansky, CTO for Flexible Electronics & Advanced Packaging at SEMI. “We’re looking at electrical properties that can be modified and more complex sensors like MEMS.”

PARC has a project underway with NASA’s Jet Propulsion Laboratory to develop a sensor sheet made up of hundreds of thousands of sensors. Some are sensitive to temperature, while others are sensitive to light. “These can be printed in a space station, and they can be made regardless of conditions at launch,” said Veres. “You also can use this in laminates on boards. So instead of building up a board, you can add more into every layer. This gives you new design freedom, and it also works for early stage prototyping. We’re also looking at analog printing for things like batteries, where more and more scale is needed. This allows us to address cost while creating batteries that are more conformal than coin cells.”

Challenges ahead
While the possible applications are seemingly limitless, printed electronics need to undergo the same kind of rigorous verification, validation and testing as other chips. And that’s where reality begins seeping in.

“This requires multi-physics simulation—electrical, thermal and chemical—said Norman Chang, chief technologist in ANSYS‘ semiconductor business unit. “With flexible-hybrid electronics, there is 3D geometrical input, so you need tools to visualize the while design. This is way different than a finFET, where you have multiple layers but each layer is planar. So there is less correlation with a finFET than you might expect, and you need to know what kind of angle you want to wrap around.”

ANSYS is working with Hewlett Packard Enterprise on solving these kinds of problems, which can be extraordinarily complex. “One of the major applications is a sensor film, but with continued vibration the thermal gradient can differ in different places,” said Chang. “With a wrist band or a body monitor, you also need co-simulation due to the flexibility. That can also change because there is an abundance of different kinds of substrates, and each case will need a different approach. Hopefully we will see some standardize products emerging where the design base is on a common platform. That would save a lot of cost.”

That is particularly true for antenna arrays, which could be printed on packages for applications such as 5G. The problem there is an inability to test those devices using standard test equipment, and solutions might apply to printed electronics.

“With millimeter wave, the antenna wavelengths are short enough that you need small antenna arrays, but you mount those on the package so there is no longer an RF port,” said David Hall, chief marketer at National Instruments. “That is going to require over-the-air measurements of RF power out of the device. And if you’re using beam-forming, you have 16 antennas to test. You can test them one at a time, or measure at some incident angle. The goal is to measure process defects in RF power, or anything affecting modulation quality.”

Different materials also would require different equipment for simulation and test. “We’re going to see this with a smart phone that is foldable, IIoT applications and body monitors,” said ANSYS’ Chang. “There will be different sensors for different markets, and different materials will need different simulation tools. There is a lot of activity between government and industry on this today.”

Reliability is a key concern in this area, and strips of sensors might be easy enough to replace, it’s not clear how long they will last or how they will deteriorate.

“Durability is always a problem in printed electronics,” said Brewer. “In some ways, it’s more durable than silicon or metal and it’s very stable. But you do have to worry about the physical durability of a device, which could include everything from de-lamination to physical degradation like cracking when it is exposed to a harsh environment. You also have to figure out if heat is going to be a problem—do you need to create a heat-resistant flexible substrate—and does it need to be transparent.”

Conclusion
Printed electronics largely has been a science project in the past. But there is enough momentum building around this approach—and enough end-user demand across a variety of market segments—that it’s very likely the electronics industry will see much more of this kind of technology in the future.

How quickly, for which applications, and where problems may erupt are still to be determined. But the ability to put electronics into any environment, regardless of the shape or environmental conditions, and to be able to obtain more data than through traditional chips is a very attractive set of attributes for many applications. That set of attributes is only likely to grow as the kinks are worked out of the supply chain and reliability simulation and testing are proven to be sufficient for quality assurance.

Related Stories

Flexible Devices Drive New IoT Apps

Printed Electronics Materials