Press Release

Printed Electronics, Inkjets and Silver Ink

I have been working with silver ink and ink jet printing for many years. In the 80s, my company was involved in a yearlong R&D program for Nortel.

We were researching whether or not silver/copper inks could be applied to plastic and FR-4 to create workable low-cost circuits. The program's objectives: To find a suitable plastic material that could be molded and then screen-printed with silver ink and a dielectric material. The plan was to create a very low-cost green plastic phone. The components were attached with silver ink, and the keypad was composed of carbon pads and conductive buttons.

The research was both a success and a letdown. We were able to produce a working clear plastic phone that sat on the CEO's desk for years. The silver ink portion was very successful; the plastic molding was not .

The art of silk-screening silver ink is well-known and millions of such circuits are made every year. The highest use is in silver ink thin plastic keypads and front-panel applications. The circuits tend to be simple, wide trace-width single- or double-layer applications. They represent a far cry from the high-technology multilayer PCBs with fine lines and small vias that the market presently uses.

The question persists: Is the present silk-screening silver ink technology good enough to ever manufacture an acceptable PCB? Because of a few production problems, I do not believe it is possible. First, the best line widths typically obtainable with silk screening is too wide, at 7/7. Yes, you can tweak them to get 5/5, but not on every pass.

The main problem is gravity, which will flatten out of the ink, causing bleeding and reduced spaces. We tried and failed at a flash UV-cured ink to stop the ink from flowing. Even with thixotropic inks, the ink still flowed and was not usable. The thinning of the track also presented a second problem; the higher resistance and weakening structure of the line caused cracking.

To lower the resistance of the silver ink, and reduce the possibility of cracking and dendrite growth, I used a process that covered the top layer of silver ink with a dense copper/palladium metal. The process used a simple green two-tank setup which worked well.

Resistance is our enemy. Silver ink is composed of about 92% silver particles in the 12 micron size, with a thermal cure epoxy-based binder. Even though silver has one of the lowest resistance numbers, when we mix it with epoxy the resistance rises quickly. The epoxy insulates the particles from each other, reducing the number of contact points and raising the resistance.

Manufacturers try to buy silver particles that feature many facets and points rather than flat platelets. The size of the silver particle also affects the resistance: The smaller the particle, the lower the resistance but the higher the price. Oven cycles above the recommended maximum temperature can cause more silver points to contact each other, as you reduce the epoxy volume due to outgassing and volumetric reduction, and you may damage the epoxy binder.

Screening silver or copper conductors is not up to speed to replace existing PCBs. But inkjet technology does have a good chance of becoming mainstream in the very near future, depending on the specific application. Most inkjet printer heads have a hard time laying down enough ink to create a suitable conductive line for any electronics that uses low power.