Plastic Electronics: The Challenges for Low Temperature Manufacturing

Citation

(2011), "Plastic Electronics: The Challenges for Low Temperature Manufacturing", Circuit World, Vol. 37 No. 3. https://doi.org/10.1108/cw.2011.21737caa.020

Plastic Electronics: The Challenges for Low Temperature Manufacturing

Article Type: Conferences and exhibitions From: Circuit World, Volume 37, Issue 3

The Hauser Forum, CambridgeMarch 15, 2011

The IeMRC and the Cambridge Integrated Knowledge Centre (CIKC) held their joint one-day plastic electronics seminar on the March 15 at the Hauser Forum in Cambridge. The seminar had a specific focus on the challenges and opportunities for new low-temperature manufacturing techniques in photonics and electronics and the clear level of interest in the subject was confirmed by the 123 people that attended the event. The full programme included ten speakers from both industry and academia, who covered a wide range of important subject matter relating to plastic and printed electronics (PEs).

The seminar was opened with a welcome from Dr Chris Rider, Director of the CIKC, who emphasised the focus on manufacturing that was shared by both the CIKC and the IeMRC. The CIKC had a key objective of moving projects out of the laboratory towards commercialisation. Several of the presentations during the day would highlight the work of both the IeMRC and the CIKC, while others were from small companies working on novel PE technology. Chris then introduced the first speaker of the day, Dr Steve Jones, of Printed Electronics Ltd (PEL), who gave a presentation titled “Printed electronics: where from here?” Steve began by giving an overview of his experience in the industry and a review of what actually constituted electronics. Electronics was about interconnection and integration of components to form functional devices. It was now becoming widespread, ubiquitous, and invisible. He then went on to define what constituted PE. In this case, circuitry was created with conductive and/or electroactive inks, using a wide variety of techniques. PE must enhance functionality and, if it did not, it would be doomed to failure. Steve emphasised how, in the last ten years, producers in the Far East had become extremely dominant in producing most of the world’s electronics. Taiwan’s Foxconn, for example, was one of the biggest producers and was intending to increase its workforce to 1.3 million people over the next 12 months. The big question was how other companies could compete with these large organisations, especially in Europe, where the supply chains barely existed any more. PE was still immature, i.e. it was still largely at the demonstrator/proof of concept stage. Therefore, there was a need to find niche, new products moving into areas where electronics had not gone before, e.g. for smart packaging and in anti-counterfeiting applications. PE had been over-hyped in the past, but there were fundamental gaps in the toolbox and the levels of integration that had been achieved were still considered to be poor. Collaboration was seen as a route to future success and Steve re-emphasised the huge potential that was available for PE. He then gave examples of some of the products that PEL had been able to produce, including displays and logic devices that could be embedded for anti-counterfeiting applications. Steve concluded by giving an overview of the UK’s PE capabilities.

The next presentation was given by Professor Henning Sirringhaus from the University of Cambridge, entitled “Organic transistors for applications in flexible electronics”. Henning began by introducing the performance achievable with printed organic transistors (thin-film transistors, TFTs), which was currently often inferior to traditional silicon devices. The organic semiconductors used had low mobilities, although substantial improvements had been made since the 1980s. He described the new organic materials that had been at the heart of this improvement and described the enhancements that could be achieved by downscaling. New technologies and equipment were needed for defining the requisite smaller channel lengths. A new technique known as self-alignment printing was then detailed which enabled organic transistors to be made that could switch in the megahertz range. The CIKC had studied this approach in detail to see if it could actually be used in a real life manufacturing process. It had been shown that arrays of transistors could indeed be made with good uniformity and high yield. Complementary n- and p-type devices could now be produced that gave approximately equal performances and work had been carried out to integrate these into logic-type devices. Work had also been undertaken on metal oxide semiconductors, as these could provide another route to even higher mobilities. In particular, the possibility of solution deposition had been investigated and this had required the development of appropriate alkoxide materials that could be spin coated onto amorphous substrates. This sol-gel route had been used to produce indium zinc oxide field-effect transistors that had then been characterised for performance and some of the data were shown, e.g. mobilities of 8-10 cm²/Vs had been recorded. Indium gallium zinc oxide TFTs had also been produced. Unlike the organic transistors, these metal oxide devices did, however, require some higher temperature processing stages.

After a networking coffee break, Hazel Assender from the University of Oxford gave a presentation on the IeMRC-funded flagship project to develop the “high-speed vacuum deposition of organic TFTs in a roll-to-roll (R2R) facility”. The project that Hazel described was running at four universities, Oxford, Leeds, Manchester and Bangor and also involved eight industrial partners. The R2R manufacturing capability of the project was then detailed and this used the evaporation of an organic monomer to deposit a polymer on films such as PET to form transistor structures. The R2R processing could run at 5 m/s and handle films up to 350-mm wide. A larger and faster capability was also available via the industrial partner, Camvac. The advantages and disadvantages of vacuum deposition were compared with solvent techniques. Vacuum was a low-energy rapid process in which the deposition of multi-layers was relatively straightforward. The basic circuit design being investigated was then explained. The polymer deposited was based on an acrylate chemistry and metals were selectively deposited using an evaporation process. Hazel then focused on the deposition of pentacene and described the impact of the deposition parameters on the subsequent material performance. For example, pentacene grown in nitrogen ambient gave the best crystalline material. The transistor’s insulator layer was formed by the vacuum deposition of a monomer that was subsequently cured. Curing of the material had been studied using FTIR and by monitoring the dielectric constant of the polymerised film. The effect of insulator thickness on device performance and the shelf life of fabricated transistors had also been studied. In conclusion, the project, which had been running since summer 2010, had demonstrated the ability to fabricate organic electronics in a R2R environment and it was already possible to build complete devices from multiple layers.

Dr Steve Thomas from Conductive Inkjet Technologies Ltd (CIT) then gave a presentation entitled “Additive manufacture of interconnects”. He began by emphasising the importance of interconnects in electronics and stating that they were traditionally based on the copper tracks of PCBs. However, these needed multiple process stages were labour and equipment intensive and wasteful of materials. An alternative approach was screen printing but, again, it also tended to be relatively expensive and also used a lot of material. As a result, CIT had moved away from these conventional processes and focused on a route that separated the printability from the conductivity. One possible approach was to use an electroless copper plating solution and Steve then described the basic chemistry of the electroless deposition process. The concept was to inkjet deposit catalytic metals that were capable of initiating the deposition of copper in an additive process. The key challenge had been to develop a suitable catalytic ink formulation and one such example also incorporated a UV curing system containing photoinitiator materials that generated free radicals after exposure. This approach gave a very rapid transition to a cured film. Much work had been carried out to give good microstrucutral control to enable the metal catalyst to be accessible and which had a roughened surface morphology for enhanced metal adhesion. The process developed was suitable for reel-to-reel applications which enabled reduced processing costs via the ability to have continuous processing and less labour input. One innovative feature in the equipment was the use of fluidic bearings. Typical feature sizes produced were 220 μm and larger, with a printing speed of 20 m/min giving sheet resistances of 20-100 mࡎ/square. The copper was deposited at a plating speed of 2-12 m/min and the metal was solderable using conventional surface mount technology and low-temperature solders. The processing of finer features had also been developed using a photolithographic patterning process and this enabled features down to 3μm to be produced for use in range of applications. These included making fine meshes that could be used with PEDOT to replace indium tin oxide. Steve concluded by saying that the CIT “print and plate” method offered a more economical solution for interconnect metallisation.

The final paper of the morning session was given by Dr Andrew Flewitt from the University of Cambridge and this was on the selection of combinations of semiconductor/dielectric metal oxides deposited without substrate heating for transparent TFTs. He began by explaining why amorphous silicon continued to dominate in terms of large area electronic materials. This was because it was a known technology based on a stable non-toxic material that could be processed at low temperatures. However, it also had low carrier mobility, incurred patterning costs and was metastable when in device operation. A key alternative was metal oxide technology. Although this tended to be more of an unknown, it did offer amorphous microstructured materials via low-temperature processing that could give high carrier mobilities. There was a range of materials that could be used such as zinc and tin oxides. The oxides could also be used to form dielectric structures. For a given application, the appropriate material technology had to be married with the optimum deposition and patterning processes. Using metal oxides, it was possible to deposit transparent TFTs and these would find increasing use in a wide range of applications such as sensors, touch screens, etc. The novel deposition process developed for these materials was described and was known as the HiTUS sputtering system. This gave a much higher deposition rate than typical magnetron sputtering and produced stress-free films without substrate heating. Materials such as zinc tin oxide, cuprous oxide, indium zinc oxide, hafnium oxide, and alumina had been deposited as pinhole-free films that could be used in a range of devices. A four mask process for producing TFTs was then described and the performance characteristics of a range of devices were presented.

The afternoon session was chaired by Professor Martin Goosey, Industrial Director of the IeMRC and, after giving a brief overview of the IeMRC’s activities, he introduced Richard Young from Brunel University who gave a presentation on “Microcontact printing and applications”. Richard began by giving an introduction to the related work that had been carried out at Brunel in recent years on offset lithography and printed electroluminecent displays. He then described the concept of soft lithography and microcontact printing, using a polydimethylsiloxane (PDMS) elastomeric stamp. A single silicon wafer could be used for producing a range of individual PDMS stamps used in the microcontact process. The work carried out to develop the thiol-based inks used in the process was also described. A range of conductor track and gap sizes down to 25 μm had been produced for use in display structures, the stimulating voltages were reduced from 130 V to around 70 V in certain more electrically stable structures with 75 μm tracks and gaps. A range of frequencies had also been used to stimulate the phosphor particles. Richard had also looked at using the conductors in thermistor structures based on copper oxide. For the future, it was envisaged that sub-micron dimensions would be possible.

The second presentation of the session was on “Electronics via imprint” and was given by Scott White of Pragmatic Printing Ltd The company had been formed in 2010, but was originally an earlier spin out from Manchester University. The company had a unique technology for making 2D semiconductors that was thus ideal for printing applications. Scott then outlined the challenges for printed logic. One of these was being able to produce the smaller features required for increased performance and to reduce overall size. Imprint was said to be the oldest form of printing and had been performed for thousands of years. The processes used today were basically the same but with the capability to produce much smaller features at low cost and high volume. Features down to a few nanometres could now be produced. The process was detailed and the first stages were similar to microcontact printing. There was then a range of process options, including direct embossing into functional materials, patterning of a resist for subsequent pattern transfer and pre-patterning of a substrate for selective deposition. An example of an imprint stamp was shown that exhibited nanometre-scale features. These approaches could then be used to produce novel devices with complex 3D structures and multiple material layers. The 3D imprint processes enabled self-aligned structures to be formed. Simple 2D planar nanotransistors could also be produced with sub-micron features and a single semiconductor layer. Scott then described the printed logic platform and the need for a full range of device architectures based on a common platform for printing them. Initial applications were shown and the example of low-cost non-invasive printed logic labels was cited. These required easy assembly on polyester, paper or card. The demand for electronic smart packaging devices had been valued at around $7.7 billion in 2010 and there was a growing need for new printed security devices. There were other applications in consumer goods for brand enhancement and smart packaging, as well as in novelty products such as greeting cards, toys and games. Scott concluded by summarising that imprint technology provided a unique platform for printed logic and that there were many new opportunities emerging.

Dr Daping Chu of the University of Cambridge then gave a talk entitled “Laminated electro-active foils – liquid crystal transflective panels”. Dr Chu outlined the trends in display technology from 1980 out to 2040 and it was clear that there was going to be a very large growth in demand for reflective display types. He then described electro-optic bistability in smectic liquid crystals and its potential use in e-ink “newsprint” displays. By using a black-dyed cell, it was possible to achieve a reflective contrast of >7:1 with a very large number of pixels that could be multiplexed. Much work had been carried out to develop high stability dyes for these applications and, in particular, there was a need for coloured dyes that could be used in full colour displays. Dr Chu described the technical and commercial objectives of the “LEAF” project that was addressing a number of specific aspects of display manufacturing. Its key objective was to develop technology for plastic envelope liquid crystal displays. This work included the characterisation of a wide range of materials. He then showed the facilities available within the Centre for Advanced Photonics and Electronics at Cambridge, before giving some examples of displays that had been made, including coloured plastic cells. These could be combined in stacks to provide full colour reflective displays and a possible process flow for making such displays was described. Work had also been undertaken to produce UV durable colourants and to replace indium tin oxide with graphene in transparent electrodes. Plastic electroactive foils would find large area applications in smart windows but, if full colour displays could be produced, there would be many more applications, e.g. in large area advertising signage and for changing a building’s décor. Other potential applications included transparent antennae and photovoltaics. A number of UK manufacturing opportunities were then described.

Following a networking and refreshment break there were two final presentations. The first of these was given by Dr Keiran Reynolds of eight19 Ltd and he gave a talk entitled “Towards roll-to-roll production of organic photovoltaic modules”. It was clear that new designs and printing processes would revolutionise the way that photovoltaics were built and eight19 was developing the technology and manufacturing processes needed to make plastic photovoltaics a reality. The solar market was growing rapidly. In 2009, 1.45 billion people lacked access to electricity; there was thus a huge opportunity for providing so-called “off-grid power”. One example was to develop a lighting unit that could provide 2-5 W for 5 h/day. There could be a demand for up to two million of these in India alone by 2017. Another opportunity was for the “off-grid charging” of mobile phones. Organic photovoltaics had the potential to offer a number of real advantages over conventional approaches and these included solution processing, robustness and lightweight. The power conversion efficiency of these devices was gradually increasing and was currently at ∼8.3 percent. The structure of such an excitonic solar cells was described and shown to be different to that of a conventional solar cell. Kieran then discussed some of the deposition challenges, e.g. for forming the semiconductor composite structures and the associated morphology control. Another key challenge was to achieve maximum area utilisation in the cells in order to obtain the best power generating potential.

The final presentation of the seminar was given by Professor Richard Penty of Cambridge University and was on “Polymer interconnects for datacom and sensing”. As data rates and frequencies continued to increase, there was an growing need to use optical interconnects to augment traditional copper technology. Optical interconnects offered a number of advantages, such as lower EMI, lower power consumption and higher data rate capability. However, the technology also needed low-cost components that were compatible with conventional PCB manufacturing processes. A siloxane-based material from Dow Corning was available that exhibited suitable mechanical, thermal, and optical properties for wave guide applications. Multimode waveguides had been deposited on FR4 substrates with cross-sections of a few tens of microns. The fundamental transmission properties of these waveguides were described and it had been found that very low crosstalk could be achieved between adjacent waveguides. Also, waveguides crossing each other at 90° were effectively lossless and crosstalk was better than 30 dB. A polymer back plane application had been considered for utilising the optical waveguide approach and a ten card optical design was described that had 100 waveguides in total. There was widespread industrial interest in the use of optical back planes and many of the larger electronics companies now had their own design concepts. The challenges of providing optical coupling schemes were described and the aim had been to produce a low complexity, low-cost optoelectric PCBs. The design adopted was based on a low-cost double-sided FR4 board in which the electrical layout and vias were processed first. Waveguides were then fabricated on the bottom board surface and the components were attached using a solder reflow process. A through board L connector was then used to provide the optical interconnections. The optical performance of such an assembly shown. The presentation concluded with a discussion of the use of sensitised waveguides that could act as gas sensors. Such a device had been built that was sensitive to ammonia; the dye used changed colour upon exposure to ammonia. Measurement of the optical absorption was able to give the ammonia concentration and it was a reversible process. Work had also been carried out to develop a screen printing method for polymer waveguides and this had been successful to a certain extent, although there were issues with surface roughness and the quality of the waveguides was currently deemed to be poor. Further work was underway to improve the processing and quality of these printed waveguides.

Martin Goosey then gave the concluding remarks for the seminar and thanked the speakers, who he noted had covered a diverse range of plastic and PE applications including, amongst others, interconnects, semiconductors, sensors, photovoltaics, optical waveguides, and displays. Throughout the day there had been a strong focus from all of the presenters on addressing the manufacturing challenges and on taking their work towards real commercial applications. In summary, this was an excellent seminar that attracted a large audience from both industry and academia. It is to be hoped that the success of this seminar will provide the catalyst for future-related events.

Martin Goosey