Last night I attended the IEEE Santa Clara Valley Consumer Electronics Society monthly meeting. The main presentation was by Joel Yocom, Business Development Manager for Conductive Inkjet Technology Ltd.
His presentation will be posted here later.
Joel presented an overview of inkjet technology and how they are applying it to printing circuits. They have developed a process that allows them to inkjet a catalytic ink which after UV curing allows the electroless (e-less) plating of copper. Given the choice of inkjet systems from scanning formats where the print head moves to fixed heads where the material moves past the head they have a wide range of potential substrate sizes and formats to choose depending on the end application. In addition, by using a piezoelectric drop-on-demand inkjet print head a wide range of ink chemistries can be used since the ink is not heated like most consumer desktop printers which use a thermal inkjet.
Even though they would prefer to be a technology platform company, they have built manufacturing equipment including the Metajet line of printers to deploy their technology. Joel described the general features of the these system which are high volume roll-to-roll printers which print circuits on 300 mm wide polyester films at speeds up to 30 meters / minute. Since the length of a roll of film can be up to 1 km long, the practical length of the circuit far exceeds any imaginable application. They do regular production runs of a circuit that exceeds 6 m in length.
With inkjets, print resolution drives both the cost of the print head (finer resolution requires smaller drop size which increases head cost and number of nozzles required) and speed (smaller the drop size the longer it takes to cover the area). For their current systems, Joel described the minimum feature size as approximately 200 µm. The design guide from their service bureau specifies minimum line width and minimum gaps in terms of CAD data and pixelation.
The resulting copper traces are 2 to 3 µm thick and have a resistivity of 30 mΩ/square. Since the traces are significantly more resistive than the typical 0.5 ounce copper per square foot foils (1 mΩ/square nominal resistivity) used in printed circuit boards (PCB), this limits the end applications. Initial high volume applications include LED strings and arrays, and RFID antennas (UHF and microwave).
At finer line widths such as 5 to 10 µm, the technology can be used to build meshes on transparent substrate for capacitive touch screens. Detailed information can be found on pages 21-23 of this company presentation. Potentially, conductive meshes print this way could replace Indium Tin Oxide (ITO) based solutions. [See also IEEE Nanotechnology Session 4 – Eric Granstrom – Cima NanoTech.]
In terms of cost, Joel quoted pricing of less than $1 per square foot in volume on polyester film. The cost for their process is 10 to 25% that of either flex circuits or rigid PCBs for applications that can accommodate the higher trace resistance. Operationally, since ink jetting is an additive process there is a substantial reduction in waste and related costs. Unlike traditional PCB processing, there is neither a used photo-resist nor a copper etchant waste stream. Currently their line runs approximately three to four million square feet of material a year and all their waste fits within a 55 gallon drum.
Fundamentally, this technology will not replace traditional PCBs but will enable lower specification or new applications at significantly lower costs. And their service bureau provides an inexpensive and easy way for potential users and product developers to evaluate the technology. Which is important since easy evaluation accelerates the adoption of new technology.