By Charles A. Bishop, Ph.D., CEng., C.A.Bishop Consulting, Ltd.
This technical paper summarizes presentations given at the hybrid pro flex 2021 conference, organized by Fraunhofer Institute, FEP, in Dresden, Germany, in September 2021. The program highlights work done by Fraunhofer and its industrial partners as well as publicizes progress made by some of the collaborative European Research Initiatives. The conference theme was “Roll to product: What is coming?”
This report is my personal view of what I think were the highlights of the two days of papers presented. To see the full list of papers or to purchase access to the presentations, visit www.fep.fraunhofer.de/en/events.html and scroll down to pro flex 2021. The presentations were from a variety of collaborators ranging from material suppliers and research and development groups to equipment suppliers and final product producers.
Breakthroughs in ultra-thin glass
The opening session was largely taken up by presentations relating to the manufacturing and use of Ultra-Thin Glass (UTG). The first paper, by Riku Yamashiro (Nippon Electric Glass Co., Ltd.), covered the development and mass production of UTG. It was interesting to see the progress since they last reported on their ultra-thin glass products. They have improved the design of their manufacturing process so that now the flexible glass is available in 1-km rolls, widths up to 1,400 mm and a thickness down to 25 microns. The thin glass has a different thickness at the edges of the continuous sheet, which have to be trimmed off before winding up the flexible glass. They presented micrographs of the cut edges showing how their continuous laser slitting has improved the quality of the edges, reducing the micro-cracks at the edges. Scribing produces considerable micro-cracking, which likely limited the probability of consistently producing kilometer-long rolls. The laser beam heats the glass adjacent to a cooled zone so that thermal cracking occurs. This reduction in edge damage also means that the bending is improved with the glass being able to be bent around a smaller diameter before breaking occurs. This was shown in images of a two-point bending test showing the scribed glass failing at a bend diameter of 55 mm and, for the laser-cut glass, a bend diameter of 10 mm. This non-contact method of cutting minimizes the cracking which, in turn, makes the glass more robust for downstream processing.
It was interesting to hear that the ultra-thin glass has similar problems to polymer films with the edge trim needing to be removed during manufacture. Also, now that they can produce longer and wider rolls, Yamashiro disclosed the need to pay more attention to the tension and winding to prevent the rolls from telescoping.
There was a complementary paper by Sean Garner of Corning, Inc., [i] that gave an update of their UTG products, including some applications, and gave a brief description of their investigations into scaling up manufacturing of ultra-thin flexible ceramics. The flexible ceramic films they have been investigating include alumina (40 & 80 microns), silica (40 microns) and 3 mol% yttria stabilized zirconia (20 microns) with the alumina films being the initial focus. The benefits of the ceramics include a high-heat dissipation and thermal-shock resistance. The ceramics are made with high purity and fine grain size which helps to enable continuous sintering of the material into very thin flexible films. If we look at how long it has taken for the flexible glass to move from a laboratory, narrow, short film into the product now available, it provides a glimpse of some of the alternative ceramic films we can expect to become more widely available in 5-10 years’ time.
Lasers are another enabling technology in ultra-thin film processing. This includes coating patterning as well as scribing or cutting. Lasers can be controlled in line width and power and can be scanned across the web. This makes it possible to control the depth of material ablated so that individual layers can be removed without damage to the other layers, or multiple layers can be ablated to convert the web into sheets. The type of laser and wavelength may be used to optimize the conditions to minimize the amount of particles produced during ablation as well as melting and thermal damage. Using high-power, short-duration pulses maximizes the vaporization and minimizes the heating of the substrate. The options for this type of laser ablation are now either femtosecond or picosecond pulses. Maurice Clair of 3D-Micromac AG [ii] showed a number of different options and micrographs of cut edges of uncoated UTG cut at 0.8 m/s through to an OLED laminate. The OLED was sandwiched between two sheets of 50-micron flexible glass at a cutting speed of 0.4 m/s. The speed and accuracy of the scanned laser can be critical to retain registration and maximize production efficiency. As the ultra-thin glass width and length increases so to does the need for the speed of cutting while retaining the edge quality and minimization of any debris produced. I look forward to hearing about attosecond-pulsed laser ablation perhaps in the next pro flex conference.
Adhesive film manufacturer Fabian Unterste-Wilms of Tesa SE presented a paper on the use of adhesive films in electronic devices. There are many different applications of adhesive films in the assembly of devices using thin flexible films, including as protective surface films, shielding and grounding, lamination, component mounting, sealing and barrier. The adhesive films can be one of the first films to be laminated onto UTG rolls, and it was shown how this application has been improved for manufacturing. The laminated glass has to be laser-cut, and the technique needed to be developed such that the laser would produce a clean edge in both the adhesive and glass. The adhesive tapes are produced in a variety of thickness on a transparent polyester liner and can be either pressure-sensitive or UV-curable and can be transparent or, for barrier applications, may be opaque. To meet the changing demands of the flexible-electronics industry, the company has a technical center working in partnership with different users in the value chain, demonstrating how the tapes can be successfully laminated and wound onto the variants of ultra-thin glass.
Vacuum-coating systems
The cost of vacuum web-coating systems is high, and often the systems have been designed with a specific substrate or product in mind. It is not always easy to change substrates as the winding system may not be compatible. The substrate mechanical characteristics or thickness may require a tension range different to that available on the system. The paper by Dr. Hannes Klumbies of FHR Anlagenbau GmbH, an equipment builder, presented details of how one such vacuum-coating system built for depositing coatings onto flexible foil substrates was converted to enable coating onto ultra-thin flexible glass. Critical areas that need particular attention included roll diameters, front surface rollers and the use of liner film. Contamination, slipping and scratching of glass substrates is of paramount concern and so the target was to ensure the winding system included no front surface rollers. This meant there was not the option of using a high wrap roll to measure tension, and tension had to be controlled by measuring the torque on the drive motors.
As the tension require for the glass substrate was on the order of 100X less than for the metal foil, it was necessary to look at optimizing the drives. Using the more powerful drives for the foils meant that trying to control them at low tension is difficult with poor sensitivity, and the losses due in the rotating feedthroughs can add to control problems. To use linered rolls of glass required the winding system to have a winder to remove the liner before the glass could be processed around the deposition drum and an unwind for a liner to be included prior to the glass being rewound. The large deposition drum and the low tension required for the thin glass meant that the force holding the glass against the drum was very low, making the need for precise speed and tension control essential to prevent the glass from slipping and potentially becoming scratched. This paper highlighted a number of the potential difficulties and the costs arising from making modifications on an existing roll-to-roll, vacuum-deposition system when wanting to make a change of substrate outside the original design specification.
In-line gauging and inspection
Another session was dedicated to in-line monitoring using Hyper Spectral Imaging (HSI) either as a standalone system or in combination with complementary techniques such as X-ray diffraction (XRD) or reflectometry (XRR). This topic was introduced by Dr. Philipp Wollmann of Fraunhofer IWS on HSI and Dr. John Fahlteich of Fraunhofer FEP, the project co-ordinator of the NanoQI project to use combined HSI and XRD/XRR. The HSI system was introduced during the last pro flex conference where the basic technique was introduced with simple demonstrations showing how the technique might be used. Over the last couple of years, the HSI technique has been used on a number of demonstration projects to elaborate in more detail the advantages the technique offers. The coated web is illuminated and either the transmission or reflection light is collected and detected as a wavelength scan from the ultraviolet through to the near infrared. This full scan provides much more detailed information about the material than just a simple percentage reflection or transmittance value. The HSI enables information such as film thickness, morphology and chemical composition to be determined over the whole film and on multilayer films.
Julio Hernandez from Norsk Elektro Optikk AS, which makes hyperspectral cameras, showed how their system could be used to compare chemical composition of batch samples of pharmaceuticals, detection of foreign bodies in cotton textiles and polymer identification in recycled-polymer scrap. The performance of the HSI technique was enhanced by having baseline information and using calibration samples. In this presentation, the relationship of the width of web being interrogated and the distance between the camera and web, which determines the pixel size, was emphasized. As with other techniques that monitor webs for defects to resolve defects smaller than 100 microns means that the camera has to be close to the web, which limits the width that can be analyzed. This increases the number of systems required for full-width monitoring, thus increasing system cost.
Fabiano Rimediotti of Nordmeccanica SPA-Vacuum Div. brought us back to vacuum web-coating problems by describing the limitations in the measurement of a simple aluminum- or aluminum-oxide-coated polymer film. The transmittance, reflectance and electrical conductivity are routinely measured ,and newer systems also measure large defects of >100 microns, but there is no information on the stoichiometry of the coating or the structure. The NanoQI project uses a combination of HSI and XRR/XRD. The XRR, once calibrated with suitable sample information including simulated spectra, can then be used to compare the sample information by curve fitting and then information about thickness, density and roughness can be determined. If XRD is used, information about the crystal size or structure can be added. This information is obtained offline and is then used by the HSI to further improve the information on the coating as it is measured on-line in the vacuum system. The target for this work is to replace the current optical monitor with an HSI camera and to use the output to control the aluminum thickness and oxidation level in the aluminum-oxide coating. I, for one, will be interested to hear the results of this work at the next pro flex conference. I still have a reservation about the predictive use of this system with respect to barrier coatings. It is known that defects smaller than 100 microns affect the barrier performance of coatings. If they are not measured, then they will be an unknown error to the predicted barrier performance, thus giving the prediction a random uncertainty.
This was followed by a presentation on the production of perovskite coatings by Dr. Giulia Lucarelli from TNO. The perovskites are produced from an ink coated onto a carrier film which is dried to promote crystallization and then quenched and annealed to produce the final structure. Critical to the process is the crystallization and annealing of the coating. The speed of the process is compatible with having both the HSI and XRD performed on-line. The HSI is positioned at the quenching station and monitors the quality of the coating at the start of the process. The XRD is situated after the quenching where it can determine the crystal size and coating density, which means it can provide feedback on any process changes that are made to the initial drying that, in turn, affects the nucleation density, which affects the crystal size. This is a sheet-production line and not roll-to-roll, and so the sheet can be annealed and reversed back to the XRD to be monitored a second time to evaluate the final structure and provide feedback about the annealing process. This combined monitoring enables optimization of the process to be speeded up as the coating monitoring is moved from off-line to on-line. The use of HSI either as a standalone system or in combination with XRD/XRR has the potential to both speed up development as well as monitor critical steps in manufacturing. It is still in development with respect of where it can best be used, and it will be interesting to see how this development progresses. I would still like to see improvements in the resolution to be able to monitor smaller defects but in a way that did not simply multiply the systems and the cost.
Energy-harvesting developments
The other focused session was on the “Smart2Go” European project, introduced by Dr. Matthias Fahland, project coordinator for Smart2Go. This research initiative is tasked with producing energy-harvesting devices coupled to energy storage and management that could be used by a wide range of applications including “wearables.” There many energy-harvesting options from photovoltaics, triboelectric or piezoelectric generators, thermoelectric or electromagnetic harvesters. Organic photovoltaic (OPV) and thermoelectric thin-film devices were shown, and final applications such as organic light emitting diodes (OLEDs) were connected and fitted to safety jackets that, once the light level falls, become illuminated, helping to make the wearer more visible. The energy produced by the harvesting devices needs to be stored and Bernd Fuchsbichler from VARTA Microinnovation presented the work being done to optimize and develop batteries to meet the current safety specification.
This was followed by Markus Tuomikoski of VTT Technical Research Centre of Finland, Ltd., who presented his paper on the development of an autonomous energy-management and supply platform (ESP) which adds a couple of integrated circuits to the whole process of managing the incoming energy harvested to be stored by an integrated battery or supercapacitor and then managing the output and communication to the external sensors. I found the description of some of the ideas of wearables disappointing. The target seems to still be to produce discreate harvesting devices which would then be added to clothing, such as by sticking them onto the clothing. Attaching devices to clothing is not ideal as the devices are never as flexible as the clothing itself and, although they might be tolerable in some industrial clothing, would be less acceptable to the wider public if use in everyday clothing. As all the methods of energy harvesting are already being developed onto fibers to allow then to be woven into clothing, this does appear to be a project that is lagging behind work being done elsewhere in the world. Where sensors and control systems are required, micro devices have been developed that are small enough to be included into spun fibers, allowing them to be fully integrated into clothing.
This session ended with a presentation about “Project SYMPHONY to Develop Printed Piezoelectric Energy Harvesting and Sensing for Energy Autonomous Condition Monitoring” by Dr. Jonas Groten of JOANNEUM RESEARCH Forschungsgesellschaft mbH. This project is looking at everything from new materials and new devices to enable sensors to be used in remote or difficult places such as pressure sensors or stress sensors on turbine blades. The use of the sensors is aimed at improving the use, reliability and lifetime of the items they are sensing which will help reduce the carbon footprint of the technology. The results of this project also should provide technology that could be of use to the “Smart2Go” project.
All this...and more
There was also a scattering of other disparate papers that were offered throughout the conference. The paper, “Development of Innovative OLED-Patch for Photodynamic Therapy” by Dr. Sebastian Luci of Endomedica GmbH, described the development of an OLED that was designed to be small and provide topical irradiation of skin cancers, replacing the large lighting arrays currently used. The use of an activation cream on the skin improves the results, and the close proximity of the OLED helps give a more precise dose of UV light, leading to a more effective treatment with less total irradiation used.
Two other papers highlighted the range of papers in that one by Prof. Stephen Forrest of the University of Michigan covered the “Deposition of organic thin films from the vapor phase,” and the other from Ziam Ghaznavi of Emerson & Renwick was of a technique to remove coatings in the paper, “Roll-to-roll reactive ion-etching of nanoscale features in Si for next-generation flexible electronics.” The UofM paper showed various organic vapor-deposition configurations and roll-to-roll systems for the manufacture of cost-effective OLEDs that are predicted to meet the US Dept. of Energy cost goal of $100/m2. The presentation included details of organic vapor-jet printing where the vapor can be jetted out through an array of nozzles to provide patterned coatings. For anyone interested in organic vapor deposition this provides a good glimpse of the technology.
In contrast, the E&M presentation was all about how to etch coatings and surfaces. This is well established technology for silicon wafers but is still new to roll-to-roll systems, and the paper delivered results on the work that has been done to enable this to be included as a roll-to-roll technique. Using specially designed plasma sources, located facing a cooled drum, and sliced silicon wafers that were attached, using Kapton tape, to a stainless-steel web could be wound past the plasma sources and the silicon etched. The silicon wafers were patterned using standard wafer processing ready for the etch step. Using design of experiments and statistical process modeling, the gas composition was investigated to understand how the gas composition would affect the etching process. By changing gas composition, the shape of the pillars could be changed from cylindrical to conical. Also, the position of the wafers from one side of the drum to the other was checked for uniformity of process as were any effects relating to the movement of the wafers as they passed through the plasma. Using four plasma heads around the drum, an etch of >1micron of Si was achieved at a winding speed of 1 mpm. It will be very interesting to see some real applications emerge from this first step into roll-to-roll reactive ion etching.
Other papers also looked at producing fine lines or patterns onto flexible substrates at different scales from coating, embossing and UV-curing gratings or diffusers onto 750-mm-wide ultra-thin glass to the paper presented by Piotr Kowalczewski XTPL SA which developed a high silver-content ink (85 wt%) that was extruded out of a fine printing nozzle (0.5-10.0 microns) to produce fine lines of 1-10 microns using precision-controlled substrate movement. Not only could this technology be used to produce grids, both regular or shaped, but could be used to repair damaged, broken or missing lines on circuits that have complex topography.
Conclusion
I reiterate, this review does not comment of every paper. I hope it has given a flavor of how broad the topics covered were. pro flex is to be recommended as a conference well worth attending, and I look forward to the next one.
Footnotes
i. See “Establishing the flexible-glass R2R coating ecosystem,” Converting Quarterly 2019 Q1, p. 64
ii. See “Roll-to-roll laser processing of flexible thin-film devices,” Converting Quarterly 2021 Q4, p. 49
Charles A. Bishop, principal at UK-based C.A.Bishop Consulting, Ltd., holds a Bachelor’s degree in Materials Engineering with a Diploma in Industrial Studies. His research led to developing a process for manufacturing titanium-based bone implants for tendon location. He went on to obtain a Master’s degree and Ph.D. following further research into vacuum-deposition processes. Charles has 40+ years of experience in vacuum deposition, mainly onto flexible webs. He writes the “Vacuum Verbiage” Q&A technical column for this publication and moderates the online “Vacuum Web Coating” Technical Topics Channel. Charles can be reached at +44-1509-502076, email: cabuk8@btinternet.com.
コメント