However, considerable progress is being made in raising the performance of materials in key areas like conductivity, which dictates the mobility of an electric current within a device and the speed with which it operates.
A lot is also being done to improve printing processes so that they can provide the degree of resolution necessary for the large scale manufacture of electronic products.
Without the ability to achieve a higher level of resolution than that required for conventional graphics work, printing would not have the long-term potential to create a massive market for low-cost electronics. Within 20 years, this could be bigger than the present global market for silicon chips.
A Look at the Future
Details of some of these latest trends were outlined at two recent conferences held at Cambridge and London and organized by the consultancies IDTechEx and Pira International.
Several speakers at the meeting arranged by IDTechEx in Cambridge discussed the high cost of ink ingredients and substrates. They claimed that materials as a result accounted for a disproportionate amount of the total costs of new printed electronic products.
Jie Zhang, principal staff engineer at Motorola, calculated that materials absorbed 87 percent of the total costs of developing printed electronic products, while equipment accounted for 10 percent and facilities 3 percent.
But she conceded that “the biggest proportion of the high material costs” comes from the cost of polymer substrates, like polyethylene terephthalate (PET), polyester, polycarbonate and polyethylene naphthalate (PEN), which require a lot of pre-treatment.
“The proportion of material costs should be as low as possible if we are going to eventually sell these products in large quantities,” said professor Arved Huebler, director of the Institute for Print and Media Technology at Chemnitz University of Technology, Germany, and founder of a start-up printed electronics company.
The price of some materials has been pushed up by outside influences. The cost of indium tin oxide (ITO), a conductive anode material in organic light emitting diodes (OLEDs), has rocketed by a factor of 15 to around $1,000 per kilo over the last four years due to its use in liquid crystal displays. H.C. Starck Group, the electronic materials subsidiary of Bayer MaterialScience, is marketing its Baytron conductive polymer as a less expensive alternative to ITO.
The price of silver, which is used widely in electronic inks because of its high conductivity, soared by 50 percent in the first half of this year because of demand from investment funds. Robert Oberle, president of the RFID tags producer TCD Technology, Quakertown, PA, told the London meeting of Pira International that silver is now “very expensive, which does not make it a viable material in the long term.”
Edward Voncken, chief executive of Orgatronics B.V., a Dutch technology company for OLEDs, pointed out that materials like conductive polymers would inevitably be sold at a high price because they are used in such small amounts. “If developers are purchasing only a few kilograms annually, how do the producers get a return on their own development costs?” he asked.
Partly as a result of the cost and quality of materials, IDTech Ex has been revising its forecasts for the printed electronics market, although it still expects the sector will be worth approximately $300 billion worldwide by 2025.
It now predicts that by 2015, only 10 percent of RFID tags will be fully printed, when the total number produced could amount to more than 500 billion. A year ago, it thought that around a third would be.
“Some companies have been delaying the start of commercial production of printed RFIDs because of the problems they are having with the quality of materials,” said Peter Harrop, IDTechEx chairman. “There is plenty of printing capacity out there. At the moment, the developers of printed electronics cannot make things work in the way they expected.”
Insufficient conductivity in polymers could be resolved by the introduction of hybrid materials which incorporate both organic and inorganic components. “Already some chemical companies are doing work on printable inorganic conductors,” said Raghu Das, IDTechEx chief executive. “We consider that printed electronics will be both organic and inorganic.”
Although there have been some recent significant advances in the conductivity of polymers, most conductive polymers still achieve mobility levels well below that of amorphous silicon, electronically the weakest of the silicons.
R&D programs are focusing on finding ways of boosting the conductivity of organic materials by combining them with inorganic substances with a much higher mobility. The challenge has been to achieve a combination that has enough solubility to be printed at high speeds and volumes to guarantee low-cost manufacturing.
Thomas Lindner, R&D manager at Hewlett Packard, Palo Alto, CA, told the IDTechEx meeting that his company had been working since 2003 on a printable zinc oxide sol gel transistor, or liquid transformable into a solid, which has a higher conductivity that amorphous silicon.
Reuben Rieke, president and chief executive of Rieke Metals, Lincoln, NE, told the conference how his company had been using a variety of metals to give, at a low cost, higher levels of conductivity to polymers like polythiophenes. It has recently made a number of polythiophenes with nickel and zinc contents of 0.02 percent and 0.2 percent, respectively.
However, Vivek Subramanian of the electric engineering and computer sciences department of University of California, Berkeley, warned that inadequate mobility was not the only problem with organic materials.
“With printed organic transistors, stability is a challenge,” he said. “Organic devices almost all show substantial operational degradation, i.e. frequency instability. Consequently, substantial material engineering is required.”
Bruce Kahn, a consultant at Printed Electronics Consulting, Rochester, NY, told the Pira conference that upscaling of R&D work could pose problems when materials have to undergo the stresses of a fully operational printing process. Current improvements in organic materials could give them a conductivity higher than amorphous silicon by 2008.
“These materials could be better by a factor of 100,” he said. “But we cannot be sure that just because it has been achieved in the laboratory, we can make similar transistors on the production line.”
Limitations of Inkjet
Because of worries about the effects of upscaling, both conferences were told that more R&D is being directed at printing processes other than inkjet, which so far has been the predominant method used in printed electronics.
Mr. Lindner said that inkjet has already become an established process because it is low-cost and is compatible with a variety of substrates and provides direct control of ink composition.
Karel Vanheusen, strategic program manager for printed electronics and displays at Cabot Corp. pointed out that inkjet had shown itself to have the potential to serve a range of markets from electronic components and printed circuit boards to RFID tags, flat-panel displays, solar panels, fuel cells and batteries.
Steve Jones, business development director, Printed Electronics Ltd., UK, said there were doubts about the efficiency of inkjet in the manufacture of commercial quantities of electronic products.
“(Inkjet) offers the possibility of producing one-off concepts and prototypes through pre-production and conceptually into higher volume production,” he said. “The major limitation of inkjet printing has been the inability to print reliably at high resolution and precision, i.e. where a micron is meaningful.”
The big problem with inkjet is the necessity for inks with low viscosity, which restricted resolution, he noted. “An ink’s functionality (in printed electronics) is inversely proportional to its jettability,” he said. “Anything that jets well will be useless in the process and vice versa.”
However, Colin Marsh, a scientist at The Technology Partnership (TTP), revealed details of advances in inkjet processes at his Cambridge-based technology development company which seemed to show that uncertainties about the resolution of inkjet could be unjustified. TTP has developed an inkjet system with high viscosity inks which is able to print lines of a 20-micron width and 3 microns apart.
Another drawback for inkjet could be that despite its current low costs it may not be in the long term cheap enough, particularly in sectors like RFID where the objective is print tags at a unit cost of 1 US cent.
“I don’t think a 5-cent RFID tag is achievable with inkjet,” John Attard, business developer at Xaar, the UK inkjet printhead manufacturer and developer, told the Pira conference. “Inkjet can certainly help to reduce costs but you are not going to get down to that level with it.”
Mr. Kahn said that inkjet is not a process appropriate for mass production. “If you want to print something in high volume, it does not make sense if you do it at a drop at a time,” he said.
Wolfgang Mildner, managing director of PolyIC of Germany, which aims to make commercial quantities of printed RFID tags next year, said his company has concentrated on using different printing processes, none of them inkjet, within a roll-to-roll (R2R) system. These include flexo, offset, gravure and screen printing, all of which need to be modified for printed electronics.
“You cannot use standard methods without adapting them to the special needs of printed electronics,” Mr. Mildner explained to the Pira meeting. “You have to combine several different processes because of the different layers that have to be printed.”