A Twist on Printed Electronics: Printing on 3D Shapes
Barry Matties speaks with Optomec’s Pascal Pierra about their LENS printer systems and Aerosol Jet technology, which allows manufacturers to print applications like sensors and antennas on 3D objects. The process is significantly faster and greener than competing technologies.
Barry Matties: Pascal, why don't you start by telling our readers a little bit about your company.
Pascal Pierra: Optomec is a U.S. company, headquartered in New Mexico. We have been focused on 3D printing/additive manufacturing for about 20 years with two product lines. One is the LENS product line, which is a directed energy deposition technology where we can print metal, like titanium and stainless steel. The main application for the LENS product line is adding materials to existing components—for example to restore, remanufacture, or add features to metal parts. Optomec offers additive and hybrid systems where we can combine LENS technology with a CNC machine to enable additive/subtractive metal processing all on one machine. We have collaboration already with CNC machine tool suppliers in the U.S., Mexico with VIWA, Taiwan with Tongtai, and we're working on other collaborations in Asia. We also have our own hybrid systems called the LENS Machine Tool series.
The other product line is the Aerosol Jet, a unique technology for printing electronics onto 3D structures. It allows users to aerosol spray a wide variety of electronic and even bio-materials onto substrates at feature resolutions ranging from 10 microns to a few millimeters. What is unique to Aerosol Jet technology is its ability to print high-resolution conformal features onto any 3D shape.
Matties: So, Aerosol Jet is about 3D printing electronics; the technology is interesting because you're using gas to focus this and you atomize the particulate.
Pierra: Yes. Aerosol Jet printing is using an Optomec patented technology to create a spray. We have two different types of atomizers available, pneumatic or ultrasonic, that nebulizes a source material, for example a silver nano-particle ink, into a dense mist of very fine droplets ranging in diameter from 2–5 microns. Each droplet can contain thousands of silver nano-particles. This dense cloud of droplets with entrained nanoparticle particles is then directed to the Aerosol Jet deposition nozzle where it is surrounded by a sheath gas. The big advantage of our technology is that sheath gas is used to dynamically focus the cloud into a tight stream to create features ranging from 10 microns to a few millimeters. Also, thanks to the sheath gas, the mist is accelerated to a velocity of 50 meters/second , so it remains in focus during its travel from the nozzle exit to the substrate for a distance up to 5 mm. This long focal length enables it to maintain the desired feature resolution when printing on non-planar surfaces. No other technology on the market has that capability.
With other printing technologies, such as inkjet, the nozzle must be very close to the substrate and therefore they are not able to print on non-planar surfaces.
Matties: The market for 3D surfaces is quite broad because there are all types of applications.
Pierra: That's correct. So, one of the main markets that we're addressing is manufacturing of antennas for smartphones as well as the IoT market with printing various types of sensors. We can print antennas and sensors together as well as interconnects including multi-level interconnects, because we also can print dielectric materials at the cross-over points, eliminating the need for multilayer circuit boards. This all means we can print more electronics into a smaller space which is ideal for hand-held mobile devices, aerospace, and smart devices used for IoT applications. We also have customers using Aerosol Jet printing for applications in life science. For instance, there is a university in Singapore printing heaters on bandages. So, you can apply the bandage containing printed electronics onto the wound and then you apply heat to the wound. That is an example of the ability to print any circuitry onto a 3D substrate, even one that is flexible.
Matties: For a circuit board designer, would this be a tool to use for rapid prototyping for just a standard circuit?
Pierra: Rapid prototyping would be a very good application for this technology because you can print components such as resistors, capacitors, sensors, antennas, and even thin film transistors as well as interconnects to complete the circuitry. Since the printing process is driven by digital CAD data, users can create and revise prototypes very quickly. For instance, we have many military customers in the U.S. using our technology for prototyping and low volume production.
Matties: On the 3D printing side, has this technology been around for several years?
Pierra: Yes, this technology has been around for more than 15 years. There's a little subtlety between 3D printing and Aerosol Jet. With the Aerosol Jet we actually print on 3D shapes.
Matties: Versus printing an actual part.
Pierra: Right. With microfluidic needles, we can print the entire structures, but the main application is to print conformal electronics onto a 3D shape. In some cases where you need to achieve a certain resistivity, you can make the printed trace a little bit thicker by printing multiple layers but it’s not really 3D printing as you read about in the press, such as printing any shapes or any objects of various materials. What we do with Aerosol Jet is more what we call printing “on” 3D parts and those parts are typically made using conventional manufacturing methods such as injection molding.
Matties: It's an additive process onto a base that already exists. So you're not shaping a new knee. You're printing a circuit on a knee.
Pierra: That's exactly right. Then the other thing that we do, for instance we can print a coating on a stent; the addition of the material allows the body to better accept the stent and therefore reducing the medication required.
Another application that is very successful was developed in collaboration with General Electric. GE is now printing strain sensors onto metal blades used in their industrial gas turbine engines. The sensors are printed using a ceramic ink that can withstand the high operating temperature inside the engine. Each sensor has a unique digital code and a precisely printed pattern that can be used to measure expansion in the blade—a condition known as creep. During maintenance cycles the sensors are scanned and the data is sent to Predix, the GE Industrial Internet of Things analytics platform. Then just the blades that require replacement are identified and replaced reducing maintenance costs and downtime. The advantage of this is now they know which blades on the turbine need to be replaced and which don't. In reality very few blades need to be replaced during the early in-service life of the engine. Reliability engineers call the early life failure mode “infant mortality.” Prior to incorporating creep sensors, GE replaced all of the blades based on the predicted infant mortality failure time.
Matties: Regardless of condition.
Pierra: Yes. Now they can identify and replace just those blades that are about to experience failure. So, there's a huge market here that people are not yet aware of, which is to help extend the life of the critical parts by adding sensors on the parts. We are also printing strain sensors on shafts. And the same, you can measure the sensors and see if that shaft needs to be replaced or not. This opens the door to a very wide range of applications for many industries. Printing the sensors adds a few dollars to the cost of each blade, but saves thousands of dollars in unneeded blade replacement and downtime.
Matties: Now, if a customer already has materials and inventory in their warehouse, and they say, "We want to add a sensor to this," they don't need to go out to a manufacturer. They could just literally do it inside their own factory.
Pierra: That's exactly right.
Matties: This changes the way circuits could be built in the future.
Pierra: That's correct. You can add the sensor afterwards. You can add the sensors on the shaft. The shaft is a good example of a 3D shape. So, it doesn't matter. You can print anything on anything.
Matties: What about automobile applications?
Pierra: For automobiles we have a few applications, especially with printing heaters. For instance, new cars have assistance for drivers, right? Some luxury cars, if you start coming out of your lane, the car takes you back into the lane. All this technology is assisted by cameras. You don't want the camera to be obstructed by ice in the winter. So, we have applications where we can print heaters on the window that the camera is seeing through. Then, you can also print an ice sensor. You can sense that the heater needs to be turned on and turn it on automatically. This application is trickier than it looks because you send the same currents through the window, but you see the trace length is different. So, if you want the same current to go through each of the traces, you need to adjust. You don't see it, but some traces here are thicker than others so that all have the same resistance creating uniform heating/melting. So, it's not as trivial as it looks.
Matties: We were talking earlier about airplane wings or icy conditions. Can you explain more about that?
Pierra: We also have development with some airplane companies about printing the same type of heaters on flexible electronics so you can adjust it from the wings and then with the same ice sensor, you can detect the ice and then switch on the heater automatically to make sure the airplane has no icing problem before take-off.
Matties: Are you actually printing on a wing surface or are you printing a membrane that they apply?
Pierra: We can do both.
Matties: What's the size? Because the wing is awfully large. How does a wing get through your process?
Pierra: In this particular application that I was discussing, we were doing prototypes on drawings. Then it didn't get to the actual application, but in practice we could print on flexible circuitry and just apply it on the wing. It would not be a problem.
Matties: But the machine itself, it's not a machine that you could go mount somewhere and cover the whole wing, literally?
Pierra: Well, we could. The Aerosol Jet technology has been packaged as a stand-alone entity; we call it the Aerosol Jet Print Engine (AJPE). The AJPE can be mounted on a robot arm to access the wing in this example or on any other automation platforms to address whatever high volume manufacturing application you may have.
Matties: I'm wondering if there's a business of portability with this where someone can have it in a truck and go provide service to companies, or the military can have it in jeeps out in the battlefields or other locations. Is there thinking along those lines or any interest in that? Is there any practical purpose for that? Repair radios or something in a combat situation?
Pierra: Not at the moment but we did have a project with the U.S. Army to print various types of sensors and power circuits directly on 3D printed drones depending on the mission. This was a collaboration with a U.S. company called Aurora Flight Sciences.
Matties: You guys sound like you're very innovative. I'm just thinking about a wing, if I have an existing airplane and I want to convert all my planes to this heater technology in this area. So it works that way. This is really interesting technology and it's already changed the way people are thinking and doing business. Since it's not really 3D printing, how would you describe this?
Pierra: It is printed electronics in that we can print on 3D shapes. We do have the capability to make thicker lines, which is effectively 3D printing microstructures. We also have the capability to print 3D microstructures using photopolymers. The main application is printing electronic antennas and/or sensors on 3D objects. Of course you have the same flexibilities that you have with traditional 3D printers where it's all digital. It's very easy to change. It's very quick to make rapid prototyping. One of our customers, Lite-On Mobile, is a contract manufacturing company, based in Guangzhou. They have 16 of our printers and they print millions of conformal antennas and sensors for cellphones. The big advantage that they have with our technology is they're able to deliver prototypes to the customers in 3–4 days, where with the previous technology it took them 2–3 weeks.
Matties: What is the print speed? Is it a very slow or is it a robust process?
Pierra: The print speed depends on the feature size and layer thickness. There are a few parameters. I would say the print speed varies from 10 millimeters per second all the way to about 20 millimeters per second. For the Lite-On application they’re printing on four cellphone cases simultaneously on one machine. Depending on the pattern, print time can take one or two minutes to print on all four cases. So the process has high throughput and is cost effective for mass production. One million of these a month can be achieved with 16 print heads.
Matties: How does that compare to other competing technologies?
Pierra: A competing technology, for instance, in the cellphone world, is a method that's called laser direct structuring (LDS). It's a seven-step process, which includes etching and plating that is typically outsourced to a third-party vendor. Also, it's not very green and it is a longer process. By comparison ours is a very simple two-step process. You print the antenna and you cure it. It's a very quick process; it's green; it uses very little material; it doesn't involve any harsh chemicals. As far as the cost, it's actually cheaper than the current technology like LDS technology. Especially as you print larger parts. LDS is not very cost effective for large parts. For us, it makes no difference.
Matties: Is the raw material something that is purchased through you, or is this a product with multiple suppliers?
Pierra: Actually, Optomec believes in an open approach for both of our product lines, where we do not sell materials to our customers. Our customers buy powders and inks directly from material suppliers. For Aerosol Jet there are a very large number of suppliers that are selling nano-particle inks. Optomec evaluates these inks to make sure they can be used with the technology. The ink viscosity range at room temperature can be in the range of 1 to 1000 cp. And for nano-particle inks, it is preferred that the particles are less than 200 nanometers in diameter. So in essence, if the material can be suspended in a solution and atomized, it can be printed with Aerosol Jet technology.
Matties: I see you have on your screen here about 30 or so suppliers. So supply is not an issue. What about investment? What would someone look at for basically getting up and running?
Pierra: Just the last word on the materials; Aerosol Jet supports a wide variety of materials used in the electronics industry. For example, the system supports printing of non-metallic conductors such as PDOT:PSS and carbon nanotubes, resistive inks, dielectrics, adhesives, conductive adhesives, semiconductors, and even biologic materials. Most recently we introduced support for copper and copper/nickel alloy (Constantan) inks.
Matties: There's a lot of flexibility in this.
Pierra: There are endless possibilities.
Matties: One unit will handle multiple materials. They don't have to buy independent machines. You can use silver one day and carbon the next?
Pierra: That's exactly right. It's not a problem. Actually, when you print multi-layer circuitry, you can first print the silver, then you print the dielectric, then you print silver again, and then you print conductive epoxy to attach the components.
Matties: There's a big market for that. We're watching a couple other companies that are really focused on this multi-layer circuit printing from vapor to final board.
Pierra: So to answer your question on investment, Optomec offers a variety of system configurations for use in R&D to high volume production. Entry prices start at around $250,000.
Matties: When you take that investment, and compare it to the advantage GE has, the ROI must be quick.
Pierra: Correct. ROI will depend on the applications, of course. You are right to highlight that the ROI is not only on the manufacturing cost saving itself but also on the overall savings associated with the entire application.
Matties: I expect, depending on individual circumstances, it would be a pretty quick ROI.
Pierra: Yes, many new prospective customers are interested. The price is not really an issue.
Matties: How many units do have already in play?
Pierra: Today, Optomec has sold more than 300 systems in 20 countries. About 70% are Aerosol Jet and the remainder are LENS.
Matties: So it's proven technology. Well tested. It's been in the market for many years. People can buy this with confidence.
Pierra: Absolutely. The technology has been on the market or more than 15 years.
Matties: What sort of maintenance concerns would they have with your technology?
Pierra: For the aerosol jet, the maintenance is very minimal. Of course, because it's an ink-based technology, you must clean it periodically. For our production systems the cleaning cycle is once every four hours. However, downtime is minimized because the dirty print head can be quickly removed (without tools) and a clean head can be swapped in. The system is then back up and running quickly, while the dirty head is placed in a cleaning bath and ready to go again in just a few hours.
Matties: Great. Pascal, I certainly appreciate you taking time to talk with us today. Thank you.
Pierra: Thank you.