As silicon (Si) MEMS construction has become a more popular approach to building industrial piezo printheads, a number of questions and areas of confusion have arisen for potential integrators and OEMs. Below is my perspective on a variety of topics, formed by over a decade of experience engaged in nearly every aspect of Si MEMS inkjet printhead development.
Why all the interest in Si MEMS piezo printheads for industrial applications?
The advent of single pass printer architectures to address the analog-to-digital transformation of industrial print applications has generated a need for printheads with higher nozzle counts, tighter packing densities and smaller drop sizes. This need for miniaturization fits well with precise feature size control inherent in the photolithographic and micromachining techniques used in MEMS processes. Further, silicon and silicon oxide provide excellent chemical compatibility with most ink families used in commercial and industrial inkjet printing. Careful selection of upstream construction materials and bonding epoxies help to push the envelope for applications requiring compatibility with complex crosslinking inks, functional materials and aggressive maintenance fluids. Finally, silicon MEMS manufacturing holds out the promise of economies of scale enjoyed by the semiconductor industry. As the total number of units shipped grows and printhead manufacturers learn how to take advantage of the silicon area consumed, the high fixed cost of operating a MEMS fab can be spread across a larger number of units, lowering the per unit cost.
Do Si MEMS piezo printheads have a short service life?
Many manufacturers of Si MEMS printhead have adopted thin film PZT to construct their piezoelectric actuators. The use of thin film PZT allows smaller, more efficient actuators, but at the cost of increasing electrical field stresses – stresses that de-pole piezoelectric materials over large number of cycles. Fortunately, not all thin films are vulnerable to this phenomenon; the sputtered PZT film used in Samba printheads from FUJIFILM Dimatix are naturally self-poling and do not experience any reduction in activity level, even when operated over trillions of firing cycles. A more common thin film piezo construction method, sol-gel, is more susceptible to de-poling, so more attention must be made during jet design to limit the level of electrical field stress.
Can Si MEMS piezo printheads be operated at elevated temperatures?
Two considerations come to mind when operating Si MEMS heads at higher temperature: construction material suitability and thermal gradients. All Si MEMS heads will accommodate the operating temperatures needed to jet the aqueous inks most commonly used in emerging commercial print, textile, packaging, and industrial decoration applications. Most piezo MEMS heads should also be able to tolerate the typical operating points of UV-curable inks. However, as the temperature differential between ambient and the printhead jetting condition increases, the potential for undesirable thermal gradients and the associated uniformity variation grows. Fortunately silicon is relatively thermally conductive, so thermal gradients can be avoided by using printhead architecture features like recirculating ink flows through the silicon die. Higher operating temperature inks, like UV paste and phase change (hot melt) formulations, are likely to be limited by the glass transition temperature of upstream epoxies, construction materials, and, in the case of sol-gel heads, increased susceptibility to PZT de-poling. Perhaps the biggest obstacle for pursuing these applications with Si MEMS head are commercial considerations – the market demand for high temperature inks is not well established – rather than insurmountable technical challenges.
Do Si MEMS piezo printheads self-heat, do they require cooling?
From an electrical perspective, operation of a piezo inkjet printhead is like charging up and discharging a capacitor, in some cases, over 200,000 times per second. Sadly, very little of this energy gets converted into kinetic energy for the ejected drop. However, a small fraction can be lost by the piezo actuator itself, which would provide a mechanism for self-heating except for one important consideration – the flow of ink through the printhead. Enough local heat generation is absorbed by the ink ejected to prevent a temperature rise. This is especially true for heads with ink recirculation, a feature that works to prevent any localized heat buildup and reduce thermal gradients, as well as keeping ink particles in suspension, maintaining a constant ink viscosity at the nozzle and eliminating air bubbles. The only other possible source of localized heating is from onboard electronics besides the piezoelectric actuator. At a minimum, Si MEMS heads used for graphic applications incorporate digital switches to allow independent selection of each jet per firing cycle while reducing the number of interconnects to upstream electronics. The losses on these switches are low enough not to cause problems for properly thermally coupled controller chips. However if the printhead manufacturer includes the drive power electronics in the printhead assembly, then water cooling or high ink circulation flow is necessary to remove the excess heat generation for any print application requiring high duty cycle, high frequency operation. For this reason, many printhead manufacturers, including FUJIFILM Dimatix, have designed their printhead products such that the drive electronics are located external to the printhead assembly.
During the IMI Europe Digital Printing Conference, from Wednesday 30 November – Thursday 1 December in Amsterdam, FUJIFILM Dimatix will discuss the progress since the concept of Si MEMS Printhead Factory was originally introduced three years ago at the IMI European Ink Jet Conference by the company’s CEO, Martin Schoeppler. The topics addressed in the FAQ above, plus several more, will be covered in the Wednesday afternoon session.
Marc K. Torrey
VP, Product R&D and Marketing, FUJIFILM Dimatix, Santa Clara, CA USA