Final Report Summary - EFECTS (Efficient environmental-friendly electro-ceramics coating technology and synthesis)
Project context and objectives:
This project proposes an alternative processing route for electroceramic layers based on chemical solution deposition (CSD) processes and inkjet printing at ambient pressure. This consortium aims to develop smart inks that are environmentally friendly and require lower energy input during processing, as part of work package two (WP2). Secondly, commercially available inkjet equipment, with in-house electronic and mechanical adaptations, will be tested and evaluated for use with these inks and both research scale and industrial scale printing equipment was installed by the different beneficiaries of the project (WP3). The main advantages of this approach are the lower investment cost, the faster deposition with higher yield and the processing under ambient pressure enabling a complete continuous processing.
The general scopes of this project are to:
1. introduce flexible and cost effective production systems based on inkjet printing in order to realize patterned functional coatings with improved qualities and nearly 100 % yield due to the accurate deposition of controlled droplets
2. implement innovative synthesis pathways and soft chemistry to avoid excessive energy input and to produce materials in semi-custom device manufacturing
3. enable an environmentally friendly process by promoting water as the preferred medium, avoid the release of toxic gases and reduce the amounts of precursor used
4. reduce the number of reaction steps by using 'clever pathways' during technological processing.
These goals are implemented in the following steps:
1. chemical formulation: avoid hazardous precursors and promote water as the preferred solvent
2. coating deposition: processing of the inks developed by the consortium in selected inkjet equipment and qualification of innovative deposition techniques based on modified inkjet printing (electro-magnetically assisted inkjet printing and inkjet assisted plasma spraying)
3. processing: using modifications to the standard inkjet parameters such as electromagnetic or thermal activation to obtain a more efficient deposition
4. transfer: optimised scaling up for specific industrial applications based on the protocols optimised in previous sections.
Inkjet printing was selected because of high throughput and deposition; need for low investment; low energy consumption; low raw material costs and material wastage; no contamination of precursor by closed feeding circuits; precise positioning of very small volumes of fluid, high resolution and controllable thickness; and, finally, functionality of the process under air at ambient pressure and temperature.
Inkjet printing is a general term describing several technologies for the controllable, non-contact deposition of small droplets of a liquid on a substrate. The main application of inkjet printing and the focus of technology development for inks and printing equipment have always been text and graphics printing. The inkjet approach also has strong advantages for industrial materials deposition and especially for printing functional coatings and structures, but this has only recently started to attract the development effort it requires. For most graphics printing applications, the requirement for high-resolution printing of coloured inks on paper has the freedom to design both the ink and substrate to achieve this goal. When depositing functional materials, the freedom to modify the ink and especially the substrate is greatly reduced and commercially available inkjet equipment is not always suitable. Therefore, most of the effort in this project was dedicated to the development of new, environmentally friendly inks and to the selection, evaluation, modification and design of inkjet printing equipment and characterisation tools and verifying their compatibility.
Based on the very promising outcomes of the first period of this project, the consortium selected the six most promising deliverables and focussed on further research, scaling and exploitation on this selection:
1. dense yttria-stabilised zirconia (YSZ) electrolyte layers for the production of solid oxide fuel cells (SOFCs) via inkjet printing
2. dense and porous YSZ layers for SOFCs and thermal barrier coatings (TBCs) via inkjet assisted plasma spraying or IJA-SPPS, a completely new technique, developed by the consortium
3. superconducting YBa2Cu3O7-x (YBCO) layers for inkjet printed coated conductors
4. inkjet printed, photocatalytically active titanium dioxide (TiO2) coating on tiles and glass
5. magneto-resistive devices, consisting of inkjet printed, patterned functional coatings
6. inkjet printed multiferroic architectures.
Project results:
A: Ink development
Types of inks
Within EFECTS, one can identify one general synthesis strategy that is leading to a large set of different electroceramic materials, each with their own specific applications and thus requirements. Some of the coatings in the project need to be dense and micrometres thick, while others need to be printed as only one hundred nanometre thick and epitaxially grown, high resolution patterns. Some of the layers can be heat treated at very high temperatures under air, while for others the substrate is narrowing down the possibilities. This creates the need and the opportunity for research towards a very wide range of possibly interesting precursor inks for each specific application in the project. Often, different inks were developed for one material.
Although a high variety of inks was studied during EFECTS, all formulations were prepared having the following criteria in mind:
1. environmentally friendly precursors, using preferably water as the main solvent,
2. easy scalability,
3. reduced energy input,
4. cost effectiveness and
5. printability.
In practice, three different types of precursor inks were developed:
1. Solution inks: inks containing metal ions stabilised in solution by addition of different chelating molecules.
2. Suspension inks: inks formulated based on commercially available ceramic nanoparticles dispersed in preferably water.
3. Bottom-up nanoparticle inks: colloidal suspension inks containing nanoparticles that were synthesised from molecular precursors.
Rheology testing
Once stable inks are obtained, they have to be optimised for printing. Although many issues can be overcome by careful selection of the printing equipment and the jetting parameters, it is necessary to study and adapt the ink rheology prior to printing. Viscosity and surface tension are the main parameters determining the jetting behaviour. They are often combined with printing parameters such as ejection speed and nozzle diameter to calculate the Reynolds and Weber number, expressing the fluidic behaviour and suitability for jetting. Of equal importance is the interaction between the inks droplets and the substrate: the contact angle being formed between the different phases is determining the wetting of the ink over the substrate and thus the quality of printing and the required printing patterns and droplet sizes. Wetting enveloped can be used to predict the wetting behaviour and to determine the additives necessary for optimisation.
Scaling
Finally, the consortium has installed a series of batch reactors, allowing the scaling of most types of titania and YSZ inks developed within the project: reactors for slurrie inks, reactor for solution based inks and ultrasonic reactor for breakdown of commercial nanopowders. These inks are now being finalised for commercialisation.
B: Inkjet printing
Equipment selection and evaluation
At the start of the project, the types of inkjet technology most suited for materials deposition were identified. Inkjet printing systems can be classified as drop-on-demand (DOD) and continuous inkjet printing (CIP). In DOD printers, droplets are generated only when needed for deposition on the substrate; for CIP, droplets are generated continuously and recovered for recirculation when they are not required to coat the substrate. CIP is typically chosen for high speed, low resolution industrial printing (e.g. for packaging); the need to recirculate and reconstitute unused ink and to deflect droplets in flight, reduces compatibility with some functional ink systems and increases size and complexity. It was therefore decided to use only DOD systems for this project.
Existing DOD printing systems are usually based on thermal (bubble-jet), piezoelectric or electromagnetic (solenoid micro-valve) technology. Thermal printheads use rapid local heating and vaporisation of the ink to generate droplets. This is not suitable for inks that gel or react when the temperature is increased and it risks compromising stability and compositional control, so thermal printheads were not considered here. Electromagnetic ('valvejet') printheads use a pressured ink feed and dispense a droplet when a small valve is opened by applying a pulse of current to the solenoid. This technology is suitable for relatively large droplet volumes above 10 nL and is typically used only in low-resolution packaging printers and dispensing systems, but the broad ink compatibility and robustness makes it of interest for functional materials deposition, particularly in turbulent environments or with coarse suspensions. In contrast, piezoelectric printheads operate with a small negative pressure, with the deflection of a piezoelectric element causing the ejection of droplets as small as 1 pL. This technology is ideally suited to high resolution printing and patterning.
A detailed review of commercially available inkjet printing equipment and specifically piezoelectric and electromagnetic DOD printing systems, was conducted. This included printheads, drive electronics, complete printers, positioning systems and characterisation/visualisation equipment. Products from 30 manufacturers were considered in detail and, based on their specifications, a shortlist prepared for experimental evaluation.
The target of environmentally friendly synthesis and deposition and particularly the preference for aqueous inks, constrained the selection of industrial piezoelectric printheads: very few are approved for reliable operation with water-based inks. On a development scale, the high cost of manufacturers' development platforms motivated the use of third-party electronics and software; and the need to be able to adjust printhead drive waveforms to suit new inks under development also discounted some products and suppliers.
Equipment development
For most printing applications in this project, commercially available printing devices were used: the novelty arose from the application, operating environment, ink system and ancillary equipment. Several new inkjet nozzles were, however, designed and produced to meet specific needs. These were all based on the solenoid micro-valve technology from a domino macrojet printer, which has the key advantages of being suitable for dismantling for cleaning and maintenance and of separating most of the coil and plunger mechanism from the ink flow, but which is commercially available in only a small number of 7 and 16 nozzle printhead designs. The first variant, show below, is a single-nozzle device in casing suitable for mounting on a pen plotter, a convenient and low-cost development-scale positioning system. Another body style was designed for robust mounting, e.g. in plasma spray applications. Both of these designs offer unique capabilities over commercial products and are well-suited when the application requires relatively large droplets (more than 10 nL), particularly of high-load or coarse suspensions.
Although commercially available printing devices were used, for DOD electromagnetic printing bespoke printing controllers were designed and produced. These controllers were designed to offer very flexible control of the drive voltage and timing, not available in commercial systems, to allow complete optimisation of jetting performance. To support troubleshooting in industrial environments, these controllers all had protection circuitry and fault indicators.
Purpose-written software was created to support every type of printing equipment, electromagnetic and piezoelectric, used in the project from a common interface. Unlike most commercial systems, this software is designed completely for materials deposition rather than graphics printing and is fully integrated with drop visualisation.
Drop visualisation equipment for inkjet printing usually illuminates an ink stream from behind with a high-brightness light emitting diode (LED), so that the droplets are viewed in shadow. The LED is flashed for a very short time to 'freeze' the droplet motion and the delay between ink ejection and the flash is controlled to record the evolving shape of ink droplets during jetting. Commercially available equipment has several disadvantages including high cost, proprietary software and interfacing and often a lack of flexibility with printing system support: typically they are designed only for industrial piezoelectric printheads.
For this project a system was designed based on a telecentric zoom lens (Moritex) and camera (allied vision technologies). Custom-made electronics was designed to support a very wide range of delay times (0 to 10 ms) with high time resolution (0.5 µs) and to form images either from a single droplet ejection or averaged over several ejections with the full range of jetting frequencies. This makes it equally suited to imaging nL-scale droplets from electromagnetic systems over a 4 mm flight distance and pL-scale droplets for piezoelectric systems over less than 1 mm. Software with purpose-written algorithms was created for the quantitative analysis of these results, particularly for determining automatically the time-dependence of droplet volumes and velocities, which are crucial parameters for many functional material applications.
Some laboratory-scale development, particularly with electromagnetic DOD nozzles, can be performed with commercially-developed pen plotters. Customised nozzles and adapters for commercially available solenoid micro-valves were developed to allow these to be conveniently mounted. However, high-resolution printing (particularly with piezoelectric DOD systems) requires better resolution and repeatability and industrial use requires much improved robustness. Printing systems to meet these needs were produced by Soul-Kozak (Poland) to a specification developed in the project. A key innovation is the ability to use a variety of printheads interchangeably and simultaneously, including electromagnetic DOD nozzles, single-nozzle piezoelectric dispensers and industrial printheads. Sophisticated motion control hardware, supporting coordinated motion for printing arcs as well as conventional rastering modes, was implemented in combination with custom-made electronics and software.
Applications and industrial installations
The novel technologies developed in this project consist of the combination of these printing components with new inks (with demonstrated materials compatibility), software, optimised printing parameters and an appropriate processing environment to meet the needs of an application. Whilst inkjet printing has been reported for some of these applications previously, these particular optimised solutions have not and in most cases represent a significant advance in the state of the art. These are discussed in detail elsewhere in this report.
Inkjet equipment has been installed in pilot production facilities for three of these applications in the premises of three of the industrial partners: Zenergy Power, Instytut Energetyki (CEREL) and Cereco.
C. Inkjet printing YSZ electrolyte layers for SOFCs
SOFCs are attractive electric power generators due to their high energy conversion efficiency, expected environmental benefits and the flexibility of fuels used. Fully YSZ is the most commonly used electrolyte material for intermediate and high temperature operation (650 to 1 000 °C) due to its unique properties such as excellent mechanical strength, high chemical and thermal stability and pure oxygen ionic conductivity over a wide range of conditions. To produce efficient, low ohmic loss SOFCs, however, the YSZ electrolyte must typically have a thickness less than 10 µm. Thin YSZ electrolytes have previously been successfully prepared by a number of techniques including RF magnetron sputtering, chemical vapour deposition (CVD) and electron-beam deposition. Yet, such processes are not cost effective and are difficult to scale up. Ceramic powder processing routes like tape-casting and screen printing are widely used and produce excellent coatings. They are inexpensive and readily commercialised, but reliable deposition of coatings with thicknesses less than 10 µm presents a challenge. Liquid phase deposition techniques like spray coating, electrostatic spray deposition (ESP), dip coating or spin coating are comparatively inexpensive and suitable for preparing thin coatings, but they also have disadvantages: spin coating is difficult to scale to high-throughput production, dip coating requires additional processing steps if only one surface is to be coated and spray methods are difficult to control to avoid material wastage and coating non-uniformity. In addition, multiple coatings and complex sintering procedures are often required and the strain induced during thermal cycling can result in the appearance of defects such as cracks, pores or delamination.
The inkjet technology however, combines excellent scalability and cost-effectiveness with far greater control and precision than conventional liquid phase deposition methods, reproducibly dispensing droplets with volumes from one pL to tens of nL at rates of several kHz on demand and is in that way advancing the state of the art in scalable, cost-effective fabrication of SOFCs.
The chosen SOFC design was an anode-supported configuration, in which the YSZ electrolyte must be deposited by inkjet printing on a porous nickel oxide (NiO) YSZ substrate. After heat treatment, this substrate becomes a highly porous cermet, acting as the anode and mechanically supporting the cell. The porosity of the substrate and the need to prevent YSZ from entering and blocking the pores during deposition, required the development of an inkjet printing approach significantly different from those commonly used. Both for conventional graphics and industrial printing applications and for many of the other novel applications developed in this research project, inks are produced from nano-scale suspensions (pigments) or solutions (dyes or sol-gel precursors) and there is a strong motivation to deposit very small-volume ink drops (e.g. from piezoelectric printheads). In contrast to many other fields of this project where very precise and small volume printing is envisaged, for this application, a comparatively coarse YSZ suspension (particle size around 1 µm) needed to be developed and a printing technology with comparatively large nozzle orifices based on solenoid micro-valves was selected to allow reliable and scalable deposition. This combined printing approach is referred to as electromagnetic (EM) DOD direct ceramic inkjet printing (DCIJP).
Laboratory-scale printing studies were performed in the University of Cambridge (UCAM). A suspension ink was prepared from commercial zirconium dioxide (ZrO2) 8 mol % Y2O3 powder dispersed in a-terpineol, with the addition of plasticizers and binders and methanol to adjust the viscosity. Printing parameters, including ink pressure, micro-valve opening time and droplet spacing, were studied in order to optimize the surface quality of the YSZ coating. It was found that moderate overlapping between adjacent droplets and multiple coatings produced the desired membrane quality.
SOFCs with NiO+YSZ/YSZ/LSM+YSZ/LSM architecture have been successfully prepared and tested in the Instytut Energetyki (IEn) using humidified hydrogen/nitrogen mixtures as the fuel and ambient air as the oxidant. Performance of these cells was comparable to SOFC fabricated using conventional fabrication methods. Scanning electron microscopy (SEM) revealed a highly coherent, dense YSZ electrolyte layer with no open porosity.
Electrical characterisation was then performed for each cell. The fuel cell housing was heated to 700 to 800 °C and the NiO/YSZ anode cermets were reduced in the presence of sequentially increased hydrogen flow. As anticipated, the open cell voltage (OCV) of the SOFC cell increased steadily with an increase in hydrogen flow. An open cell voltage in the range 1.01 to 1.02 V was measured for the SOFCs operated at 800 °C with humidified hydrogen (H2) fuel, i.e. consisting of 95 % H2 and 5 % water (H2O). Although the theoretical OCV value is 1.077 V, the measured values are typical for gas-tight cells operated without any seals with an electrolyte fabricated by conventional methods. The OCV measurements are consistent with the SEM surface and cross-section appearance (no open porosity). After the cell voltages had stabilised, current-voltage cell performance was measured. These results demonstrate that electromagnetic inkjet printing technique can be successfully implemented, using 10 layers of the 5 wt % YSZ suspension, to fabricate electrolyte coatings for SOFCs thinner than 10 µm and comparable in quality to those fabricated by more conventional ceramic processing methods. Microscopy and the SOFCs electrical performance confirm that a gas-tight electrolyte layer was formed using this novel inkjet printing route. In addition, electrochemical impedance spectra (EIS) have been obtained and analysed with the equivalent circuit model to perform deconvolution of the impedance spectra. The results indicate that losses associated with the ohmic resistance of the electrolyte layer and the contact resistances of the electrode/electrolyte interface constitute only 15 % of the cell resistance at 800 °C; the anode concentration polarisation appears to be the major loss factor. The low ohmic resistance of the inkjet-printed electrolyte and the low contact resistances at its electrode interfaces further demonstrate that dense and adherent electrolyte layers suitable for SOFC applications can be produced by the inkjet route developed here.
The technology was initially developed and demonstrated on the laboratory scale to fabricate 2.5 times 2.5 cm SOFCs, with printing performed in UCAM. To demonstrate the successful transfer of this technology, a printing system was developed for IEn. To support a higher degree of sample throughput and semi-automation, a positioning stage was developed with height adjustment for the printhead, front and rear access for multiple substrates and the ability to print a total area of 0.2 m2 per batch. For the jetting parameters and droplet spacings typically used, this system allows total coating areas of several square metres to be printed routinely without interruption, at a rate of up to 0.15 m2/min.
Using this equipment, the SOFC design was also scaled up to 4 times 4 cm2 and 16 cm2 active area SOFCs, fabricated at IEn, with electrolyte layers fabricated using inkjet printing technology. Similar characteristics of the complete fuel cell have been measured for the cells with scaled-up active surface area.
Further systematic optimisation of the SOFC design and inkjet printing parameters has now been performed using the Taguchi matrix approach for two different cathode materials. The successful transfer and optimisation of this new technology for YSZ electrolyte deposition prepares the way for future commercial exploitation. There are also opportunities to exploit the same core technology for the deposition of other SOFC components, including functional layers and electrodes and the deposition of a CGO layer based on novel ink formulations developed for other applications in this collaborative project has already been demonstrated. The use of inkjet printing technology for the majority of the SOFC architecture is expected to deliver additional improvements in cost-effectiveness and SOFC performance beyond the state of the art in continuing development after the conclusion of this project.
D. Dense 8YSZ layers for SOFCs and porous 5YSZ layers for TBCs via inkjet assisted plasma spraying
Within EFECTS a substantial amount of work was performed by Cereco in collaboration with other partners towards the development of two different thermal spraying techniques, i.e. inkjet assisted solution plasma spraying (IJA-SPPS) and suspension plasma spraying (SPS), both using the by the consortium developed water-based inks as feedstock materials. Ink introduction to the plasma flame was performed either using inkjet printing (IJP) technology or atomisation. Use of an atomizer creates a spray of droplets with a broad size range distributed over a significant spray angle, while IJP creates a mono-sized stream of droplets on demand with controlled volume and velocity. The feeding equipment has to obtain high feed rates and at the same time to be robust enough to withstand the dusty, high temperature plasma spraying environment. Furthermore, the IJP nozzle has to perform stable and consistent printing behaviour, for prolonged jetting periods. The refinement of splat size using novel liquid-injection plasma spraying methods has been shown to form adherent YSZ coatings on very smooth substrates, which is not possible with the conventionally used plasma spraying methods; and the deposition of a YSZ TBC on an aircraft part using these novel plasma spraying methods has been demonstrated to meet key specifications required by the aerospace industry.
IJA-SPPS set-up and testing
Within this task, the main focus was to demonstrate the novel concept of using inkjet technology as a liquid injection method for plasma spraying. Jetting in a plasma spray environment introduces additional requirements and many inkjet printing systems are unsuitable because of changes in ambient pressure, temperature and the high gas flow rates. Key requirements are the achievement of high solid fraction feed rates and the control of the droplet size and velocity, even under these harsh conditions. Inkjet equipment evaluation by UCAM identified three electromagnetic inkjet systems for this application: one modified by UCAM from a Domino MacroJet printer ('Domino') and two commercial products: one from Lee Products ('Lee') and one from Fritz Gyger ('Gyger').
It was shown that the droplet velocity is controlled mostly by the ink pressure while single droplet formation could be achieved without satellites. The droplet volume and position measured in flight by quantitative drop visualisation confirmed very stable jetting with constant velocity during high frequency operation, as well as for long continuous operating periods at constant frequency. This is a result of the rapid response of the valve, its internal fluidic design and the hard ceramic seal. Gyger valves were therefore selected as the final choice for IJA-SPPS and IJA-SPS development in Cereco.
Dense 8YSZ layers for SOFCs
The development of 8YSZ thin layers appropriate for use as electrolytes in anode supported SOFCs, was one of the final goals for Cereco in collaboration with UGent, UCAM and IEn. For the conventional state-of-the-art deposition of SOFC electrolyte layers, screen printing, doctor blading and dip coating are mainly used, with inkjet printing as an emerging technology in this field. These processes all require a post-deposition sintering step. This can be avoided by employing sputtering and, at considerably lower cost, plasma spraying. Unlike conventional 8YSZ deposition, IJA-SPPS deposits the electrolyte layer in a single step procedure without sintering. The reason for this is that the precursor materials, during their travel in the hot plasma flame, react to produce the desired YSZ phase. IJA-SPPS coatings are formed from the solidification of micron and submicron splats on the substrate. These splats are developed by particles created from the precursor solution droplets during their flight in the plasma flame prior to substrate impact and consequent solidification. Critical plasma spraying parameters (primary and secondary gas flow rate, plasma power, spraying distance, atomising gas flow rate, ink feed rate) were extensively studied so that a high quality microstructure could be achieved. Thin coatings developed from 8YSZ solution inks using IJA-SPPS with a Gyger valve (200 µm orifice) on anode-support substrates (from IEn) present a homogeneous lamellae-free microstructure consisting of a molten matrix mixed with some unmolten particles and homogeneously dispersed micropores without any microcracks. Thin coatings deposited from 8YSZ suspensions using SPS with an atomizer on the same anode-support substrates had a thickness of 10-50 µm, presented uniform thickness, a good quality microstructure and 7% closed porosity. Both techniques (IJA-SPPS & SPS) produced 8YSZ coatings with good adhesion to the substrate although the initial anode substrate surface is very smooth (roughness Ra less than 1.5 µm) which is attributed to coating development from small sized splats; this is not possible for conventional APS coatings, formed by mechanical anchorage of large splats on high roughness substrates. The innovative techniques used in this project hence advance the state of the art in thermal spraying giving the opportunity to achieve 8YSZ thin layers adhesion not only on ceramic anodes but also on polished metal substrates, which is an area of active SOFC research.
Porous 5YSZ layers for TBCs
The development of thick 5YSZ layers appropriate for use as TBCs using IJA-SPS and SPS was a second final goal for Cereco in collaboration with UCAM. The endorser Hellenic Aerospace Industry (HAI) will further test and evaluate the trial article (demonstrator) prepared by Cereco. For the development of TBCs, in industrial practice, two dry route processes are used: conventional APS and electron beam physical vapour deposition (EBPVD). Thick APS coatings exhibit low thermal conductivity, a lamellar microstructure, cracks parallel to the substrate induced by rapid solidification and randomly distributed large pores (some tens of microns) induced by stacking defects. Thick EBPVD coatings compared to APS ones present a column-like microstructure perpendicular to the substrate inducing better mechanical performance and higher durability, but with simultaneously higher cost and higher thermal conductivity. Thermal spraying using liquid feedstock inks for the development of these coatings is expected to bring improvements due to refined microstructural characteristics originating from the sub-micron and nano-sised starting materials. Thick coatings developed from 5YSZ suspensions using SPS with atomisation on polished or grit blasted Hastelloy X substrates with a NiCrAlY bond coat reveal different characteristics depending on the selected plasma spraying parameters. The coating quality requirements were specified by HAI. The optimum set of parameters resulted in 200 µm thick coatings with a featureless microstructure, homogeneously dispersed micropores, 5% porosity and 400 HV0.1 microhardness. An increase in the plasma power results in 270 µm thick coatings with 17% porosity, micron and sub-micron sized pores, vertical to the substrate pore distributions and 287 HV0.1 microhardness. Simultaneous increase of the plasma power and reduction of the atomising gas flow rate results in coatings with 3% microporosity and 380 HV0.1 microhardness. Coatings developed under optimum conditions exhibit high bond strength values while durability investigations (thermal cycling testing) have reached 350 cycles with no sign of detachment, spalling or delamination. The maximum achieved deposition rate is 1.14 µm/pass. Thick coatings deposited from 5YSZ suspensions using IJA-SPS present a white coloured smooth top surface while their micrographs present a uniform and microcrack free microstructure with a thickness up to 10 µm. The achieved deposition rates were 0.05 µm/pass.
Scaling - TBC on exhaust nozzle divergent seal for F-16 engine
Large scale demonstration was performed by Cereco on a simulating exhaust nozzle divergent seal from the engine model F110-GE of the F-16 aircraft (55 cm x 15 cm), provided by the aerospace industry endorser. The part was bond coated (around 140 µm thick) with the atmospheric plasma spraying (APS) technique using commercially available NiCrAlY with spheroidal gas atomised particles. Although the IJA-SPS technique proved feasible for the deposition of TBC coatings, very low deposition rates were achieved. Thus, for the deposition of a 250 µm thick coating, 5 000 passes are necessary, which last up to 121 consecutive hours using a single inkjet printing Gyger nozzle. Spraying for such a long period results in undesirable side effects, including accumulation of stresses in the coating. On the other hand, use of SPS with the atomizer technique results in deposition rates two orders of magnitude higher. This means that with this technique only 220 passes (almost 5.5h) are required for the same thickness. Thus, it was decided that the real article would be coated using SPS with atomizer for a total of 287 passes (almost 7h) resulting in a coating over 300 µm thick. This part was delivered to Aerospace Industry for further testing - sectioning for evaluation of coating characteristics and uniformity - where it will be compared with standard TBCs developed using conventional APS. Comparing thick 5YSZ conventional APS coatings developed by EFECTS with optimum ones developed using SPS with atomizer, it is verified that the latter exhibit a featureless microstructure instead of a lamellar one; splats of significantly smaller size, in the micron and sub-micron range, instead of splats of the order of some tens of microns; homogeneously dispersed micron and sub-micron pores mainly oriented vertically to the substrate, depending on the plasma spray parameters, instead of large ones randomly distributed. Such coatings could effectively reduce the thermal expansion coefficient mismatch with the metal substrate and find applications as TBCs.
E. Inkjet printed high temperature superconducting (HTS) wires
In electrical engineering applications, the current carrying capacity Jc of HTS wires is the decisive criterion for the technical applicability. All HTS are ceramic, oxidic materials with copper-oxide layers as a common basis. This layered crystalline structure causes a strong anisotropy of the critical current density. Thus, the critical current density, parallel to the copper-oxide-layers, is several times higher than it is vertical to the layers. In the same way, the current carrying capacity through high-angle grain boundaries is strongly restricted. As it is not possible to produce a kilometre-long monocrystal, the industrial manufacture of HTS tapes is faced with the challenge of equally aligning all small superconductor crystals within a HTS wire. The critical current of a HTS tape therefore depends on the texture degree and thus on the manufacture of the wire.
The HTS wires of the actual generation - second generation (2G) - consist of a metal substrate tape subsequently coated with buffer and superconductor layers. The superconductor layer is textured by an aligned growth of the superconductor layer on the buffer layer lying underneath and acting as a template. In the case of CSD processing, an already textured metal substrate is used as an economical and scalable approach. The metal substrate can be manufactured in large lengths by standard rolling and annealing processes and serve as texture template for the aligned growth of the buffer layer and the following HTS layer. The material approach selected in this project was nickel-tungsten alloy (NiW) as substrate, lanthanumzirconate (LZO) and ceria (CeO2) as buffer layers and yttrium-barium-copper-oxide (YBCO) as HTS material.
The real dimensions of the coatings and the substrate can be visualised by SEM cross sectional analysis. It indicates a total thickness of the HTS wire of about 80 µm, where the ceramic coating of the buffer and superconducting layer is less than 1µm.
Main advantages of our CSD process, in combination with innovative inkjet printing at ambient pressure are the lower investment, the faster deposition with higher yield and the processing under ambient pressure enabling a completely continuous processing. Common CSD techniques currently used include spin coating and dip coating steps. It is clear however, that spin coating cannot be used for the production of long lengths, especially when continuous production is required, something that dip coating is perfectly capable of. Drawbacks presented by dip-coating include the coating of both sides of the tape, which represent a waste of precursor solution and can potentially clog the pulleys used to transport the tape. Another important drawback is that the sol usually lies in an open tray exposed to the ambient atmosphere and therefore a loss of solvent due to evaporation is likely to destabilize the precursor solution and/or change its concentration and rheological properties, introducing undesired variations in the deposition process.
In order to overcome these drawbacks of conventional coating techniques, innovative inkjet printing approaches were successfully developed within this project. Inkjet printing offers a drop wise and therefore accurate coating of one side of the substrate. The ink reservoir is constantly sealed and under inert gas preventing pollution and evaporation of the coating solution.
Zenergy Power developed and built continuous coating devices including variable inkjet print-heads and subsequent annealing furnaces. The devices allow the mostly automatic continuous production of HTS tapes with lengths exceeding 100m at 10mm width. For different layer types (buffer and superconducting layers) actually three continuous coating devices are in operation.
The main challenge for a successful operation of inkjet printing technology is the understanding and the optimisation of the ink - print-head interaction. Thus several hundred single experiments were conducted by Zenergy Power within the EFECTS project in order to achieve a stable operation of the inkjet printing. The project partner UCAM provided an upgraded version of the printer software and operated pretests of print-head technologies (electro-magnetic and piezo-electric) leading to a clear suggestion for installation at Zenergy Power. In-house, various parameters were tested and analysed in order to achieve best homogeneity and highest performance in the printed layers: printing pattern (hexagonal and square shaped patterns were tested), printing parameters (nozzle size, opening time and frequency were optimised) and inks (viscosity, surface tension and wetting angle) were optimised.
In particular, for the ink-development a large number of inks with various additives, surfactants, solvents and mixtures thereof were synthesised, analysed and tested leading to a formulation with the best ever reported properties for printed superconducting films worldwide. The impressive homogeneity of the printed films can be seen in a movie clip available at the project homepage http://www.efects.eu
In addition to the in-house developed inks several formulations of water based inks from the project partner UGhent were tested in the continuous coating devices.
Another issue faced within the project is the commercial availability of inks. Here additional requirements such as raw materials access and stability have to be considered. Zenergy Power cooperated with a special chemistry supplier in order to investigate the commercial access to the optimised inks developed within the EFECTS project. After some successful tests Honeywell Specialty Chemicals GmbH, Seelze, Germany, was qualified to deliver large quantities of metal-organic salts and ready-made coating solutions to Zenergy Power. All results obtained within the last year of the project were based on these commercial chemicals partially slightly modified by Zenergy Power. The delivery in larger lots also has the advantage of higher reproducibility over time compared to the previously used lab scale prepared batches.
Using the commercial coating chemicals and the optimised printing technology, coated conductors with the so far best performance regarding to length and superconducting properties worldwide were manufactured.
Critical current densities in the superconducting layer well above 1.3 MA per cm2 on 10 mm wide tapes illustrate the nearly perfect growth of the ceramic layers on the metal substrate. Within the EFECTS project Zenergy Power succeeded in transferring the lab scale results towards continuous processing of tape lengths exceeding 100 m (equal to more than 1 m2 coated area). The value actually achieved of more than 9000 Am is the best ever reported for an all CSD inkjet processed HTS wire.
F. Inkjet printed photocatalytically active titania layers on glass/tiles
Self-cleaning titania
Although the photocatalytic activity of TiO2 semiconductors is known since 1972, enhancement of this phenomenon was realised years later with the progress of nanotechnology. Using the energy delivered by the sun, titania is activated, changing the material in a photocatalyst, that can break down almost any organic molecule you can thinks of. Being a colourless and in many cases, perfectly transparent material, titania can be used to apply on many underground as self cleaning coating. Probably the most well-known application is the commercialised self-cleaning window glass. Yet, one can think of applying titania on door knobs, sanitary ceramics, ceramic tiles, pavements, architectural steel, bricks, etc.
Hundreds of publications on general TiO2 nanoparticles synthesis with application on self cleaning technology or photovoltaics existed before the EFECTS project. Yet, TiO2 inks were present in the ink industry limited for colouring pigment applications. In addition, ink manufactures were using exclusively the toxic and expensive solvent based inks mainly due to the good wetting behaviour and the quick evaporation. The last years, the tendency of water based inks preparation for the printing technology is clear due to regulations and environmentally friendly requirements. Within this project, the development of water-based titania precursors, compatible with inkjet technology is the main focus. Printing titania instead of applying it with expensive vacuum methods or less controlled spraying techniques, would mean a major breakthrough from industrial point of view, opening the market for a whole new range of applications. One of the main groups of costumers there are the manufacturers of ceramic tiles, glass and metal industries.
Top-down ink synthesis
Within the project, different types of titania precursor inks were developed by Nanophos and UGhent. NanoPhos evolved a top-down technique for the low cost preparation of totally water based TiO2 suspensions suitable for inkjet printing. In particular, commercial TiO2 nanopowders are processed with ultrasounds in water medium and stabilizers are added to retain the nanoparticles dispersed in the aqueous solution. The technique is environmentally friendly with no by products during the production process and has been applied for patent protection. Deposition of these suspensions transforms the coated surfaces into self cleaning, superhydrophilic and self sterilising only by harnessing the ultraviolet (UV) light emitted by the sun.
Ti4+ solution inks
Next to this, UGhent has developed a solution based titania precursor ink, containing Ti4+ ions stabilised in water by addition of different chelating agents. Traditionally, this type of solution inks has to be prepared under strict, water-free conditions to avoid hydrolysis of the titania precursor and thus uncontrolled precipitation. Yet, the use of smartly selected chemicals allows to create a highly inert, water insensitive precursor which can be used to prepare water-based inks with long time stability. These inks were tested both in piezo-electric and electromagnetic printing setups and completely optimised for printing in electromagnetic or piezoelectric setup. Inkjet printing of this kind of inks leads to 100 to 500 nm thick titania films after heat treatment to 500 °C under air. When printed on glass, these layers are completely transparent, defectless and exhibiting very smooth surfaces with roughness values below 4nm as measured from atomic force microscopy (AFM).
The photocatalytic activity of these layers was tested by following the decomposition of a methylene blue solution, contacted with the layer upon exposure to UV light. The inkjet printed layers show a photocatalytic performance that is even better than the commercial window glass that is for sale on the construction market. Currently, wheathering and other simulation tests are being performed in order to study the performance of these layers when exposed to real outside conditions.
Low temperature processing
As discussed above, sintering temperatures as high as 500 °C are necessary to convert the aqueous titania inks into functional, selfcleaning, photocatalytic layers. Yet, for many applications on more heat sensitive undergrounds such as for example polymers, wood or some types of metals, it is important to develop perfectly transparent titania layer that can be converted into a self cleaning coating at much lower temperatures. Furthermore, reducing the energy of any industrial process is of importance given the current global environmental policies and of course continuously growing energy cost.
Therefore, UGent developed a titania precursor, again water-based, that can be converted into a aqueous suspensions containing titania nanoparticles by microwave treatment for less than five minutes at 140 °C. After this bottom-up synthesis, only solvents i.e. mainly water, need to be removed from the ink after deposition while preformed titania particles are present. This means a tremendous decrease of the temperatures needed for coating, yet still resulting in a completely transparent and very thin coating. This is in contrast to most commercial titania paints that can be simply applied without heating, which offer self cleaning power yet come with a white appearance, limiting the applications. On the other hand, the inkjet printing approach, combined with smart nano-inks, offer the possibility of creating transparent layers with similar appearance to those obtained from expensive vacuum techniques such as CVD which is currently widely used to create self cleaning glass.
The bottom-up inks were loaded into the ink cartridge of a Dimatix materials printer and printed on microscopy glass slides for testing. Heat treatments varying from 100 °C to 500 °C were applied to the printed layers, all resulting in hard, non sticky and transparent layers. Focussed ion beam (FIB) SEM cross-sectional analysis was performed as a function of sintering temperature in order to determine the sample thickness. For layers sintered below 250 °C, the amount of left-over organics is still large and therefore, the electron beam was melting the surface samples. The layer thickness further decreases upon heating to 350 °C showing that only at this temperature, most organics have disappeared from the layer. The films exhibit a very smooth surface, with roughness below 4 nm over 5 times 5µm surfaces. The photocatalytic activity of the TiO2 films was studied by following the degradation of a methylene blue dye in contact with the TiO2 as a function of exposure time to UV light. Clearly, the sample heated at 500 °C shows the highest efficiency and performance is decreasing with decreased heating temperature. Against expectations, the layer that was only heated to 150 °C, performs very well at the beginning, but is clearly levelling off at longer exposure times and is not resulting in a linear response in the plot. This is certainly related to damaging of the layers after longer times submersed in water. Optimisation and outdoor wheathering test are currently being performed.
The method to produce transparent coatings at low temperature, made from these inks combined with nanoparticles and using the inkjet technology has been the subject of a patent application.
G. Magnetoresistive device
Icmab-Cisc has produced environmental friendly, stable and efficient inks for printing both La0.7Sr0.3Mn0.3 (LSMO) magnetoresistive coatings and magnetoresistive devices. The inks designed allow coating and patterned coating over single crystal substrates, amorphous and polycrystalline with functional ceramics in a fully computer aided design (CAD) and computer aided manufacturing (CAM) scope. It allows producing customised electronic magnetoresistive devices for short series changing the design on fly, during the same production line.
The designed precursors have been tested to produce epitaxial coatings, when single crystalline substrates, by the chemical route based on the decomposition of metalorganic salts, the so called CSD-MOD process. These functional ferromagnetic and magnetoresistive epitaxial coatings have been successfully tested to produce multiferroic devices by IFW.
Magnetoresistive LSMO inks have been tested in a piezoelectric driven single nozzle printer developed in the framework of EFFECTS with a mechanical resolution of 2 mm and repeatability better than 5 mm. The nozzle of 60 mm in diameter has allowed jetting droplets between 30 and 40 pL. The use of such a large nozzle has not been a limit for the resolution expected in the drawings; it has been possible to create a smaller diameter by using second harmonic excitation. A patterned coating has been developed in order to produce magnetoresistive contactless analog encoders as an example of the capacity of the procedure.
Encoders are extensively used in the feedback of the control loop of servo-mechanisms. Feed back of the position allows a very high precision positioning by using low cost motors. Although digital encoders are available, the analogue systems allows a lower cost considering both, the encoder and the driving electronics and are extensively used from very low cost applications as in positioning back-view mirrors in the cars or seat position memory up to servo-actuators in tracers or valves and a large number of applications. Basic analogue sensors consist of variable resistors, typically carbon resistors with sliding contact. The sliding system has a short life that can be estimated in some tens of thousands of operations, in the best cases. Systems requiring larger life expectancy should avoid mechanical contact. Contactless systems show a life expectancy larger than 50 million operations. Magnetoresistance has been suggested for position sensing in substitution of mechanical encoders, including also applications for the well known potentiometres. Customisation and fabrication flexibility are two essential points for the market and inkjet printing technology allows both in a very easy way controlling the production directly from the CAD system in a real CAD/CAM conception.
A full bridge could be considered with four contact pads equi-distributed around a round path of magnetorresistive material. The magnetic assembling rotates over the circle. In order to establish the main parameters for the device, magnetoresistive behaviour has been determined. According to the results the expected change of resistance of the LSMO path for a field in the range of 0.3 T which can be achieved by NdFeB magnets is nearly a 0.3 % at the room temperature.
With this values, full bridge and half bridge Wheatston configuration allows a linear response. The viability test, previous to scaling has been realised over a LAO substrate of 10 times 10 mm2. Three geometries have been considered for printing, a square closed path, a circle and a round undulated path.
The printing regularity can affect the resistance of the patterned paths. In spite of the high precision in the positioning of the drops, local defects could introduce a great mismatch of the printed device. In order to get a better control of the position of the drops, increasing of the surface energy in the interaction with the substrate has been promoted by a treatment of the substrate for hydrophilic behaviour. The result is a continuous covering.
The resistance of each branch of the bridge is in the range of 9 kW with a mismatch lower than 10 %. A full electronics testing system has been also developed. Scaling up has been performed in association with OXOLUTIA S.L. to achieve series production. A number of 512 nozzles printheads have been tested for medium scale production achieving a 720 times 1440 dpi resolution, large enough to obtain a reliable fusion of droplets. In order to enhance the magnetic field, a magnetic shell covering the LSMO tracks has been designed. Therefore, a set of inks based on magnetic nanoparticles was synthesised. Solutions of magnetic nanoparticles have been also synthesised for the development of ferromagnetic inks able to perform patterned and continuous coatings by using direct printing over the chosen substrate. The solutions have been adapted to work as inks for direct printing of ferrite based patterns being tested the most used compositions: magnetite and manganese, cobalt, copper nickel and zinc.
The solutions showed long time stability (tested up to more than three months). Applications in electronics are in the way in order to test the possibility to construct magnetic couplings and magnetic yokes in electronics circuits.
The deposition of a magnetic shell over the magnetite increases the mean flux density over the magnetoresistive element with 15 %. This increment supposes increasing the magnetoresistive coefficient from 0.3 to more than 0.4 30 % with respect the initial value.
H. Multiferroics: proof-of-concept
The term multiferroics stands in most cases for a layer system consisting of a ferroelectric and a ferromagnetic layer, where the layers are essentially elastically coupled. Multiferroic layer systems are interesting materials for future information-technology devices in which data can be written to magnetic memory elements by applied electric fields. Furthermore, multiferroics are of great interest for fundamental research because the magnetic order can be controlled through the ferroelectric order and the other way around at a microscopic level. The elastic coupling of the layers allows to control the electric polarisation of the ferroelectric layer by a magnetic field or to alter the magnetisation of the ferromagnetic layer by an electric field. The coupling of magnetic and ferroelectric orders in a material is known as the 'magnetoelectric' effect.
There are different preparation routes of multiferroic systems which are a physical deposition, e.g. pulsed laser ablation (PLD), and a chemical deposition, e.g. CSD, technique. The physical deposition route of multiferroics is well established and multiferroic devices of good quality can be produced in a laboratory scale. The physical deposition process is based on energy and investment-intensive vacuum processes requiring vacuum conditions, the use of energy-consuming evaporation systems and the low yield and deposition speed. CSD is a promising process for a low cost fabrication of complicated coatings and high-end decorated patterns at fast deposition rates.
Inkjet printing or ink plotting of multiferroic materials is a completely new technology. Most of the literature and patents on multiferroics are related to the design and manufacture of physical deposition techniques as for instance physical laser deposition. Chemical solution plotting routes are an efficient alternative due to a low cost fabrication of completed multiferroics.
The most widely used solution preparation technique of such perovskite materials is based on methoxyethanol in combination with metal alkoxides as reagents. The control over precursor characteristics is good in this precursor preparation but the process simplicity is low and the chemicals used toxic and expensive. Processes focusing on water based precursors are much less explored but a very promising alternative, holding the promise of much more simple and cheap processing.
Barium titanate BaTiO3 (BTO) is a very intensively studied ferroelectric crystal, with a lattice constant of the tetragonal BTO crystal structure, aBTO = 3.99 Å. La0.7Sr0.3MnO3 or shortly LSMO is a ferromagnetic material with a lattice constant of aLSMO = 3.87 Å, which is very close to the value of BaTiO3 leading to tensile strain in LSMO films. LSMO offers the advantage that its lattice parameter can be finely tuned by the polarisation of the ferroelectric BTO crystal. Consequently, the magnetic properties of the LSMO film can be influenced by the polarisation of the BTO crystals.
The consortium developed aqueous, environmentally friendly precursor solutions for the deposition of BaTiO3 and La0.7Sr0.3MnO3 epitaxial layers on SrTiO3 substrates. These inks were deposited on SrTiO3 (100) substrates by ink plotting using a 'Sonoplot GIX Microplotter'. The Sonoplot ink plotter is capable of applying picoliters of fluid, continuously creating features onto a surface. The ink is loaded by capillary forces into a hollow glass needle, which is attached to a piezoelectric element. At the resonant frequency of the loaded dispenser the fluid is sprayed out of a 10 or 30 µm diameter opening at the end of the needle.
Firstly, LSMO as well as BTO single layers were plotted separately on STO substrates. The wet films were dried and annealed in a tube furnace. LSMO films were pyrolised at temperatures between 250 and 500 °C in air and crystallised at 900 °C for 5 hours in oxygen atmosphere. BTO films were pyrolised and crystallised at 500 °C and 700 °C in air. Crack formation in BTO layers can be avoided by using a slow heating rate of 60 K/h. The desirable thickness of BTO layers was obtained through multiple coating and annealing processes. In a next step BTO layers were coated on top of a LSMO film. Such a multiferroic device consists of a stack of LSMO / triple layer BTO / silver.
The requirements for the fabrication of such architectures are a very flat and homogeneous La-0.7Sr0.3MnO3 film as well as a crack free, non porous and fully isolating BaTiO3 layer on top. The magnetic properties of the LSMO layer reveal a Curie temperature TC of 360 K and a saturation magnetisation of about 280 emu/cm3, assuming film thickness of 100 nm. Polarisation measurements were carried out to demonstrate the switching processes of the BTO film when applying an electric field.
Potential impact:
General overview, scientific exploitation
Given the very successful scientific progress made during the project, the consortium as a whole has been able to generate 14 published papers in international peer-reviewed journals, another 14 submitted (or in preparation) papers in international peer-reviewed journals, 33 oral presentations on international conferences, 38 poster presentations on international conferences, two patents filed and one patent under preparation.
Furthermore, there were some collaborative dissemination activities where the entire (or the largest part) of the consortium took part:
1. NN11, Thessaloniki, Greece 2011: EFECTS session with two plenary and five talks
2. sustainable, intelligent manufacturing (SIM) 2011, Leiria, Portugal: EFECTS session with one plenary and three talks
3. Eucas2011: European conference on applied superconductivity, Den Hague, the Netherlands: invited lectures on results EFECTS related to superconductors, plus three talks and many posters
4. Seventh Framework Programme (FP7) success stories: creating economic value through effective collaboration, Suschem - European technology platform for sustainable chemistry, Poland 2011: presentations and flyers at the Polish European Presidency event.
Finally, the coordinator planned a press release on the outcomes of the project by the end of 2011. For the actual commercial exploitation of the project, the plans were organised per partner.
Project website:
http://www.efects.eu
Coordinator:
Prof. Dr Isabel Van Driessche, isabel.vandriessche@ugent.be +32-926-44433
Dr Petra Lommens, petra.lommens@ugent.be +32-926-44449
UGhent
SCRiPTs, Department of inorganic and physical chemistry
Krijgslaan 281, S3
9000 Ghent, Belgium
This project proposes an alternative processing route for electroceramic layers based on chemical solution deposition (CSD) processes and inkjet printing at ambient pressure. This consortium aims to develop smart inks that are environmentally friendly and require lower energy input during processing, as part of work package two (WP2). Secondly, commercially available inkjet equipment, with in-house electronic and mechanical adaptations, will be tested and evaluated for use with these inks and both research scale and industrial scale printing equipment was installed by the different beneficiaries of the project (WP3). The main advantages of this approach are the lower investment cost, the faster deposition with higher yield and the processing under ambient pressure enabling a complete continuous processing.
The general scopes of this project are to:
1. introduce flexible and cost effective production systems based on inkjet printing in order to realize patterned functional coatings with improved qualities and nearly 100 % yield due to the accurate deposition of controlled droplets
2. implement innovative synthesis pathways and soft chemistry to avoid excessive energy input and to produce materials in semi-custom device manufacturing
3. enable an environmentally friendly process by promoting water as the preferred medium, avoid the release of toxic gases and reduce the amounts of precursor used
4. reduce the number of reaction steps by using 'clever pathways' during technological processing.
These goals are implemented in the following steps:
1. chemical formulation: avoid hazardous precursors and promote water as the preferred solvent
2. coating deposition: processing of the inks developed by the consortium in selected inkjet equipment and qualification of innovative deposition techniques based on modified inkjet printing (electro-magnetically assisted inkjet printing and inkjet assisted plasma spraying)
3. processing: using modifications to the standard inkjet parameters such as electromagnetic or thermal activation to obtain a more efficient deposition
4. transfer: optimised scaling up for specific industrial applications based on the protocols optimised in previous sections.
Inkjet printing was selected because of high throughput and deposition; need for low investment; low energy consumption; low raw material costs and material wastage; no contamination of precursor by closed feeding circuits; precise positioning of very small volumes of fluid, high resolution and controllable thickness; and, finally, functionality of the process under air at ambient pressure and temperature.
Inkjet printing is a general term describing several technologies for the controllable, non-contact deposition of small droplets of a liquid on a substrate. The main application of inkjet printing and the focus of technology development for inks and printing equipment have always been text and graphics printing. The inkjet approach also has strong advantages for industrial materials deposition and especially for printing functional coatings and structures, but this has only recently started to attract the development effort it requires. For most graphics printing applications, the requirement for high-resolution printing of coloured inks on paper has the freedom to design both the ink and substrate to achieve this goal. When depositing functional materials, the freedom to modify the ink and especially the substrate is greatly reduced and commercially available inkjet equipment is not always suitable. Therefore, most of the effort in this project was dedicated to the development of new, environmentally friendly inks and to the selection, evaluation, modification and design of inkjet printing equipment and characterisation tools and verifying their compatibility.
Based on the very promising outcomes of the first period of this project, the consortium selected the six most promising deliverables and focussed on further research, scaling and exploitation on this selection:
1. dense yttria-stabilised zirconia (YSZ) electrolyte layers for the production of solid oxide fuel cells (SOFCs) via inkjet printing
2. dense and porous YSZ layers for SOFCs and thermal barrier coatings (TBCs) via inkjet assisted plasma spraying or IJA-SPPS, a completely new technique, developed by the consortium
3. superconducting YBa2Cu3O7-x (YBCO) layers for inkjet printed coated conductors
4. inkjet printed, photocatalytically active titanium dioxide (TiO2) coating on tiles and glass
5. magneto-resistive devices, consisting of inkjet printed, patterned functional coatings
6. inkjet printed multiferroic architectures.
Project results:
A: Ink development
Types of inks
Within EFECTS, one can identify one general synthesis strategy that is leading to a large set of different electroceramic materials, each with their own specific applications and thus requirements. Some of the coatings in the project need to be dense and micrometres thick, while others need to be printed as only one hundred nanometre thick and epitaxially grown, high resolution patterns. Some of the layers can be heat treated at very high temperatures under air, while for others the substrate is narrowing down the possibilities. This creates the need and the opportunity for research towards a very wide range of possibly interesting precursor inks for each specific application in the project. Often, different inks were developed for one material.
Although a high variety of inks was studied during EFECTS, all formulations were prepared having the following criteria in mind:
1. environmentally friendly precursors, using preferably water as the main solvent,
2. easy scalability,
3. reduced energy input,
4. cost effectiveness and
5. printability.
In practice, three different types of precursor inks were developed:
1. Solution inks: inks containing metal ions stabilised in solution by addition of different chelating molecules.
2. Suspension inks: inks formulated based on commercially available ceramic nanoparticles dispersed in preferably water.
3. Bottom-up nanoparticle inks: colloidal suspension inks containing nanoparticles that were synthesised from molecular precursors.
Rheology testing
Once stable inks are obtained, they have to be optimised for printing. Although many issues can be overcome by careful selection of the printing equipment and the jetting parameters, it is necessary to study and adapt the ink rheology prior to printing. Viscosity and surface tension are the main parameters determining the jetting behaviour. They are often combined with printing parameters such as ejection speed and nozzle diameter to calculate the Reynolds and Weber number, expressing the fluidic behaviour and suitability for jetting. Of equal importance is the interaction between the inks droplets and the substrate: the contact angle being formed between the different phases is determining the wetting of the ink over the substrate and thus the quality of printing and the required printing patterns and droplet sizes. Wetting enveloped can be used to predict the wetting behaviour and to determine the additives necessary for optimisation.
Scaling
Finally, the consortium has installed a series of batch reactors, allowing the scaling of most types of titania and YSZ inks developed within the project: reactors for slurrie inks, reactor for solution based inks and ultrasonic reactor for breakdown of commercial nanopowders. These inks are now being finalised for commercialisation.
B: Inkjet printing
Equipment selection and evaluation
At the start of the project, the types of inkjet technology most suited for materials deposition were identified. Inkjet printing systems can be classified as drop-on-demand (DOD) and continuous inkjet printing (CIP). In DOD printers, droplets are generated only when needed for deposition on the substrate; for CIP, droplets are generated continuously and recovered for recirculation when they are not required to coat the substrate. CIP is typically chosen for high speed, low resolution industrial printing (e.g. for packaging); the need to recirculate and reconstitute unused ink and to deflect droplets in flight, reduces compatibility with some functional ink systems and increases size and complexity. It was therefore decided to use only DOD systems for this project.
Existing DOD printing systems are usually based on thermal (bubble-jet), piezoelectric or electromagnetic (solenoid micro-valve) technology. Thermal printheads use rapid local heating and vaporisation of the ink to generate droplets. This is not suitable for inks that gel or react when the temperature is increased and it risks compromising stability and compositional control, so thermal printheads were not considered here. Electromagnetic ('valvejet') printheads use a pressured ink feed and dispense a droplet when a small valve is opened by applying a pulse of current to the solenoid. This technology is suitable for relatively large droplet volumes above 10 nL and is typically used only in low-resolution packaging printers and dispensing systems, but the broad ink compatibility and robustness makes it of interest for functional materials deposition, particularly in turbulent environments or with coarse suspensions. In contrast, piezoelectric printheads operate with a small negative pressure, with the deflection of a piezoelectric element causing the ejection of droplets as small as 1 pL. This technology is ideally suited to high resolution printing and patterning.
A detailed review of commercially available inkjet printing equipment and specifically piezoelectric and electromagnetic DOD printing systems, was conducted. This included printheads, drive electronics, complete printers, positioning systems and characterisation/visualisation equipment. Products from 30 manufacturers were considered in detail and, based on their specifications, a shortlist prepared for experimental evaluation.
The target of environmentally friendly synthesis and deposition and particularly the preference for aqueous inks, constrained the selection of industrial piezoelectric printheads: very few are approved for reliable operation with water-based inks. On a development scale, the high cost of manufacturers' development platforms motivated the use of third-party electronics and software; and the need to be able to adjust printhead drive waveforms to suit new inks under development also discounted some products and suppliers.
Equipment development
For most printing applications in this project, commercially available printing devices were used: the novelty arose from the application, operating environment, ink system and ancillary equipment. Several new inkjet nozzles were, however, designed and produced to meet specific needs. These were all based on the solenoid micro-valve technology from a domino macrojet printer, which has the key advantages of being suitable for dismantling for cleaning and maintenance and of separating most of the coil and plunger mechanism from the ink flow, but which is commercially available in only a small number of 7 and 16 nozzle printhead designs. The first variant, show below, is a single-nozzle device in casing suitable for mounting on a pen plotter, a convenient and low-cost development-scale positioning system. Another body style was designed for robust mounting, e.g. in plasma spray applications. Both of these designs offer unique capabilities over commercial products and are well-suited when the application requires relatively large droplets (more than 10 nL), particularly of high-load or coarse suspensions.
Although commercially available printing devices were used, for DOD electromagnetic printing bespoke printing controllers were designed and produced. These controllers were designed to offer very flexible control of the drive voltage and timing, not available in commercial systems, to allow complete optimisation of jetting performance. To support troubleshooting in industrial environments, these controllers all had protection circuitry and fault indicators.
Purpose-written software was created to support every type of printing equipment, electromagnetic and piezoelectric, used in the project from a common interface. Unlike most commercial systems, this software is designed completely for materials deposition rather than graphics printing and is fully integrated with drop visualisation.
Drop visualisation equipment for inkjet printing usually illuminates an ink stream from behind with a high-brightness light emitting diode (LED), so that the droplets are viewed in shadow. The LED is flashed for a very short time to 'freeze' the droplet motion and the delay between ink ejection and the flash is controlled to record the evolving shape of ink droplets during jetting. Commercially available equipment has several disadvantages including high cost, proprietary software and interfacing and often a lack of flexibility with printing system support: typically they are designed only for industrial piezoelectric printheads.
For this project a system was designed based on a telecentric zoom lens (Moritex) and camera (allied vision technologies). Custom-made electronics was designed to support a very wide range of delay times (0 to 10 ms) with high time resolution (0.5 µs) and to form images either from a single droplet ejection or averaged over several ejections with the full range of jetting frequencies. This makes it equally suited to imaging nL-scale droplets from electromagnetic systems over a 4 mm flight distance and pL-scale droplets for piezoelectric systems over less than 1 mm. Software with purpose-written algorithms was created for the quantitative analysis of these results, particularly for determining automatically the time-dependence of droplet volumes and velocities, which are crucial parameters for many functional material applications.
Some laboratory-scale development, particularly with electromagnetic DOD nozzles, can be performed with commercially-developed pen plotters. Customised nozzles and adapters for commercially available solenoid micro-valves were developed to allow these to be conveniently mounted. However, high-resolution printing (particularly with piezoelectric DOD systems) requires better resolution and repeatability and industrial use requires much improved robustness. Printing systems to meet these needs were produced by Soul-Kozak (Poland) to a specification developed in the project. A key innovation is the ability to use a variety of printheads interchangeably and simultaneously, including electromagnetic DOD nozzles, single-nozzle piezoelectric dispensers and industrial printheads. Sophisticated motion control hardware, supporting coordinated motion for printing arcs as well as conventional rastering modes, was implemented in combination with custom-made electronics and software.
Applications and industrial installations
The novel technologies developed in this project consist of the combination of these printing components with new inks (with demonstrated materials compatibility), software, optimised printing parameters and an appropriate processing environment to meet the needs of an application. Whilst inkjet printing has been reported for some of these applications previously, these particular optimised solutions have not and in most cases represent a significant advance in the state of the art. These are discussed in detail elsewhere in this report.
Inkjet equipment has been installed in pilot production facilities for three of these applications in the premises of three of the industrial partners: Zenergy Power, Instytut Energetyki (CEREL) and Cereco.
C. Inkjet printing YSZ electrolyte layers for SOFCs
SOFCs are attractive electric power generators due to their high energy conversion efficiency, expected environmental benefits and the flexibility of fuels used. Fully YSZ is the most commonly used electrolyte material for intermediate and high temperature operation (650 to 1 000 °C) due to its unique properties such as excellent mechanical strength, high chemical and thermal stability and pure oxygen ionic conductivity over a wide range of conditions. To produce efficient, low ohmic loss SOFCs, however, the YSZ electrolyte must typically have a thickness less than 10 µm. Thin YSZ electrolytes have previously been successfully prepared by a number of techniques including RF magnetron sputtering, chemical vapour deposition (CVD) and electron-beam deposition. Yet, such processes are not cost effective and are difficult to scale up. Ceramic powder processing routes like tape-casting and screen printing are widely used and produce excellent coatings. They are inexpensive and readily commercialised, but reliable deposition of coatings with thicknesses less than 10 µm presents a challenge. Liquid phase deposition techniques like spray coating, electrostatic spray deposition (ESP), dip coating or spin coating are comparatively inexpensive and suitable for preparing thin coatings, but they also have disadvantages: spin coating is difficult to scale to high-throughput production, dip coating requires additional processing steps if only one surface is to be coated and spray methods are difficult to control to avoid material wastage and coating non-uniformity. In addition, multiple coatings and complex sintering procedures are often required and the strain induced during thermal cycling can result in the appearance of defects such as cracks, pores or delamination.
The inkjet technology however, combines excellent scalability and cost-effectiveness with far greater control and precision than conventional liquid phase deposition methods, reproducibly dispensing droplets with volumes from one pL to tens of nL at rates of several kHz on demand and is in that way advancing the state of the art in scalable, cost-effective fabrication of SOFCs.
The chosen SOFC design was an anode-supported configuration, in which the YSZ electrolyte must be deposited by inkjet printing on a porous nickel oxide (NiO) YSZ substrate. After heat treatment, this substrate becomes a highly porous cermet, acting as the anode and mechanically supporting the cell. The porosity of the substrate and the need to prevent YSZ from entering and blocking the pores during deposition, required the development of an inkjet printing approach significantly different from those commonly used. Both for conventional graphics and industrial printing applications and for many of the other novel applications developed in this research project, inks are produced from nano-scale suspensions (pigments) or solutions (dyes or sol-gel precursors) and there is a strong motivation to deposit very small-volume ink drops (e.g. from piezoelectric printheads). In contrast to many other fields of this project where very precise and small volume printing is envisaged, for this application, a comparatively coarse YSZ suspension (particle size around 1 µm) needed to be developed and a printing technology with comparatively large nozzle orifices based on solenoid micro-valves was selected to allow reliable and scalable deposition. This combined printing approach is referred to as electromagnetic (EM) DOD direct ceramic inkjet printing (DCIJP).
Laboratory-scale printing studies were performed in the University of Cambridge (UCAM). A suspension ink was prepared from commercial zirconium dioxide (ZrO2) 8 mol % Y2O3 powder dispersed in a-terpineol, with the addition of plasticizers and binders and methanol to adjust the viscosity. Printing parameters, including ink pressure, micro-valve opening time and droplet spacing, were studied in order to optimize the surface quality of the YSZ coating. It was found that moderate overlapping between adjacent droplets and multiple coatings produced the desired membrane quality.
SOFCs with NiO+YSZ/YSZ/LSM+YSZ/LSM architecture have been successfully prepared and tested in the Instytut Energetyki (IEn) using humidified hydrogen/nitrogen mixtures as the fuel and ambient air as the oxidant. Performance of these cells was comparable to SOFC fabricated using conventional fabrication methods. Scanning electron microscopy (SEM) revealed a highly coherent, dense YSZ electrolyte layer with no open porosity.
Electrical characterisation was then performed for each cell. The fuel cell housing was heated to 700 to 800 °C and the NiO/YSZ anode cermets were reduced in the presence of sequentially increased hydrogen flow. As anticipated, the open cell voltage (OCV) of the SOFC cell increased steadily with an increase in hydrogen flow. An open cell voltage in the range 1.01 to 1.02 V was measured for the SOFCs operated at 800 °C with humidified hydrogen (H2) fuel, i.e. consisting of 95 % H2 and 5 % water (H2O). Although the theoretical OCV value is 1.077 V, the measured values are typical for gas-tight cells operated without any seals with an electrolyte fabricated by conventional methods. The OCV measurements are consistent with the SEM surface and cross-section appearance (no open porosity). After the cell voltages had stabilised, current-voltage cell performance was measured. These results demonstrate that electromagnetic inkjet printing technique can be successfully implemented, using 10 layers of the 5 wt % YSZ suspension, to fabricate electrolyte coatings for SOFCs thinner than 10 µm and comparable in quality to those fabricated by more conventional ceramic processing methods. Microscopy and the SOFCs electrical performance confirm that a gas-tight electrolyte layer was formed using this novel inkjet printing route. In addition, electrochemical impedance spectra (EIS) have been obtained and analysed with the equivalent circuit model to perform deconvolution of the impedance spectra. The results indicate that losses associated with the ohmic resistance of the electrolyte layer and the contact resistances of the electrode/electrolyte interface constitute only 15 % of the cell resistance at 800 °C; the anode concentration polarisation appears to be the major loss factor. The low ohmic resistance of the inkjet-printed electrolyte and the low contact resistances at its electrode interfaces further demonstrate that dense and adherent electrolyte layers suitable for SOFC applications can be produced by the inkjet route developed here.
The technology was initially developed and demonstrated on the laboratory scale to fabricate 2.5 times 2.5 cm SOFCs, with printing performed in UCAM. To demonstrate the successful transfer of this technology, a printing system was developed for IEn. To support a higher degree of sample throughput and semi-automation, a positioning stage was developed with height adjustment for the printhead, front and rear access for multiple substrates and the ability to print a total area of 0.2 m2 per batch. For the jetting parameters and droplet spacings typically used, this system allows total coating areas of several square metres to be printed routinely without interruption, at a rate of up to 0.15 m2/min.
Using this equipment, the SOFC design was also scaled up to 4 times 4 cm2 and 16 cm2 active area SOFCs, fabricated at IEn, with electrolyte layers fabricated using inkjet printing technology. Similar characteristics of the complete fuel cell have been measured for the cells with scaled-up active surface area.
Further systematic optimisation of the SOFC design and inkjet printing parameters has now been performed using the Taguchi matrix approach for two different cathode materials. The successful transfer and optimisation of this new technology for YSZ electrolyte deposition prepares the way for future commercial exploitation. There are also opportunities to exploit the same core technology for the deposition of other SOFC components, including functional layers and electrodes and the deposition of a CGO layer based on novel ink formulations developed for other applications in this collaborative project has already been demonstrated. The use of inkjet printing technology for the majority of the SOFC architecture is expected to deliver additional improvements in cost-effectiveness and SOFC performance beyond the state of the art in continuing development after the conclusion of this project.
D. Dense 8YSZ layers for SOFCs and porous 5YSZ layers for TBCs via inkjet assisted plasma spraying
Within EFECTS a substantial amount of work was performed by Cereco in collaboration with other partners towards the development of two different thermal spraying techniques, i.e. inkjet assisted solution plasma spraying (IJA-SPPS) and suspension plasma spraying (SPS), both using the by the consortium developed water-based inks as feedstock materials. Ink introduction to the plasma flame was performed either using inkjet printing (IJP) technology or atomisation. Use of an atomizer creates a spray of droplets with a broad size range distributed over a significant spray angle, while IJP creates a mono-sized stream of droplets on demand with controlled volume and velocity. The feeding equipment has to obtain high feed rates and at the same time to be robust enough to withstand the dusty, high temperature plasma spraying environment. Furthermore, the IJP nozzle has to perform stable and consistent printing behaviour, for prolonged jetting periods. The refinement of splat size using novel liquid-injection plasma spraying methods has been shown to form adherent YSZ coatings on very smooth substrates, which is not possible with the conventionally used plasma spraying methods; and the deposition of a YSZ TBC on an aircraft part using these novel plasma spraying methods has been demonstrated to meet key specifications required by the aerospace industry.
IJA-SPPS set-up and testing
Within this task, the main focus was to demonstrate the novel concept of using inkjet technology as a liquid injection method for plasma spraying. Jetting in a plasma spray environment introduces additional requirements and many inkjet printing systems are unsuitable because of changes in ambient pressure, temperature and the high gas flow rates. Key requirements are the achievement of high solid fraction feed rates and the control of the droplet size and velocity, even under these harsh conditions. Inkjet equipment evaluation by UCAM identified three electromagnetic inkjet systems for this application: one modified by UCAM from a Domino MacroJet printer ('Domino') and two commercial products: one from Lee Products ('Lee') and one from Fritz Gyger ('Gyger').
It was shown that the droplet velocity is controlled mostly by the ink pressure while single droplet formation could be achieved without satellites. The droplet volume and position measured in flight by quantitative drop visualisation confirmed very stable jetting with constant velocity during high frequency operation, as well as for long continuous operating periods at constant frequency. This is a result of the rapid response of the valve, its internal fluidic design and the hard ceramic seal. Gyger valves were therefore selected as the final choice for IJA-SPPS and IJA-SPS development in Cereco.
Dense 8YSZ layers for SOFCs
The development of 8YSZ thin layers appropriate for use as electrolytes in anode supported SOFCs, was one of the final goals for Cereco in collaboration with UGent, UCAM and IEn. For the conventional state-of-the-art deposition of SOFC electrolyte layers, screen printing, doctor blading and dip coating are mainly used, with inkjet printing as an emerging technology in this field. These processes all require a post-deposition sintering step. This can be avoided by employing sputtering and, at considerably lower cost, plasma spraying. Unlike conventional 8YSZ deposition, IJA-SPPS deposits the electrolyte layer in a single step procedure without sintering. The reason for this is that the precursor materials, during their travel in the hot plasma flame, react to produce the desired YSZ phase. IJA-SPPS coatings are formed from the solidification of micron and submicron splats on the substrate. These splats are developed by particles created from the precursor solution droplets during their flight in the plasma flame prior to substrate impact and consequent solidification. Critical plasma spraying parameters (primary and secondary gas flow rate, plasma power, spraying distance, atomising gas flow rate, ink feed rate) were extensively studied so that a high quality microstructure could be achieved. Thin coatings developed from 8YSZ solution inks using IJA-SPPS with a Gyger valve (200 µm orifice) on anode-support substrates (from IEn) present a homogeneous lamellae-free microstructure consisting of a molten matrix mixed with some unmolten particles and homogeneously dispersed micropores without any microcracks. Thin coatings deposited from 8YSZ suspensions using SPS with an atomizer on the same anode-support substrates had a thickness of 10-50 µm, presented uniform thickness, a good quality microstructure and 7% closed porosity. Both techniques (IJA-SPPS & SPS) produced 8YSZ coatings with good adhesion to the substrate although the initial anode substrate surface is very smooth (roughness Ra less than 1.5 µm) which is attributed to coating development from small sized splats; this is not possible for conventional APS coatings, formed by mechanical anchorage of large splats on high roughness substrates. The innovative techniques used in this project hence advance the state of the art in thermal spraying giving the opportunity to achieve 8YSZ thin layers adhesion not only on ceramic anodes but also on polished metal substrates, which is an area of active SOFC research.
Porous 5YSZ layers for TBCs
The development of thick 5YSZ layers appropriate for use as TBCs using IJA-SPS and SPS was a second final goal for Cereco in collaboration with UCAM. The endorser Hellenic Aerospace Industry (HAI) will further test and evaluate the trial article (demonstrator) prepared by Cereco. For the development of TBCs, in industrial practice, two dry route processes are used: conventional APS and electron beam physical vapour deposition (EBPVD). Thick APS coatings exhibit low thermal conductivity, a lamellar microstructure, cracks parallel to the substrate induced by rapid solidification and randomly distributed large pores (some tens of microns) induced by stacking defects. Thick EBPVD coatings compared to APS ones present a column-like microstructure perpendicular to the substrate inducing better mechanical performance and higher durability, but with simultaneously higher cost and higher thermal conductivity. Thermal spraying using liquid feedstock inks for the development of these coatings is expected to bring improvements due to refined microstructural characteristics originating from the sub-micron and nano-sised starting materials. Thick coatings developed from 5YSZ suspensions using SPS with atomisation on polished or grit blasted Hastelloy X substrates with a NiCrAlY bond coat reveal different characteristics depending on the selected plasma spraying parameters. The coating quality requirements were specified by HAI. The optimum set of parameters resulted in 200 µm thick coatings with a featureless microstructure, homogeneously dispersed micropores, 5% porosity and 400 HV0.1 microhardness. An increase in the plasma power results in 270 µm thick coatings with 17% porosity, micron and sub-micron sized pores, vertical to the substrate pore distributions and 287 HV0.1 microhardness. Simultaneous increase of the plasma power and reduction of the atomising gas flow rate results in coatings with 3% microporosity and 380 HV0.1 microhardness. Coatings developed under optimum conditions exhibit high bond strength values while durability investigations (thermal cycling testing) have reached 350 cycles with no sign of detachment, spalling or delamination. The maximum achieved deposition rate is 1.14 µm/pass. Thick coatings deposited from 5YSZ suspensions using IJA-SPS present a white coloured smooth top surface while their micrographs present a uniform and microcrack free microstructure with a thickness up to 10 µm. The achieved deposition rates were 0.05 µm/pass.
Scaling - TBC on exhaust nozzle divergent seal for F-16 engine
Large scale demonstration was performed by Cereco on a simulating exhaust nozzle divergent seal from the engine model F110-GE of the F-16 aircraft (55 cm x 15 cm), provided by the aerospace industry endorser. The part was bond coated (around 140 µm thick) with the atmospheric plasma spraying (APS) technique using commercially available NiCrAlY with spheroidal gas atomised particles. Although the IJA-SPS technique proved feasible for the deposition of TBC coatings, very low deposition rates were achieved. Thus, for the deposition of a 250 µm thick coating, 5 000 passes are necessary, which last up to 121 consecutive hours using a single inkjet printing Gyger nozzle. Spraying for such a long period results in undesirable side effects, including accumulation of stresses in the coating. On the other hand, use of SPS with the atomizer technique results in deposition rates two orders of magnitude higher. This means that with this technique only 220 passes (almost 5.5h) are required for the same thickness. Thus, it was decided that the real article would be coated using SPS with atomizer for a total of 287 passes (almost 7h) resulting in a coating over 300 µm thick. This part was delivered to Aerospace Industry for further testing - sectioning for evaluation of coating characteristics and uniformity - where it will be compared with standard TBCs developed using conventional APS. Comparing thick 5YSZ conventional APS coatings developed by EFECTS with optimum ones developed using SPS with atomizer, it is verified that the latter exhibit a featureless microstructure instead of a lamellar one; splats of significantly smaller size, in the micron and sub-micron range, instead of splats of the order of some tens of microns; homogeneously dispersed micron and sub-micron pores mainly oriented vertically to the substrate, depending on the plasma spray parameters, instead of large ones randomly distributed. Such coatings could effectively reduce the thermal expansion coefficient mismatch with the metal substrate and find applications as TBCs.
E. Inkjet printed high temperature superconducting (HTS) wires
In electrical engineering applications, the current carrying capacity Jc of HTS wires is the decisive criterion for the technical applicability. All HTS are ceramic, oxidic materials with copper-oxide layers as a common basis. This layered crystalline structure causes a strong anisotropy of the critical current density. Thus, the critical current density, parallel to the copper-oxide-layers, is several times higher than it is vertical to the layers. In the same way, the current carrying capacity through high-angle grain boundaries is strongly restricted. As it is not possible to produce a kilometre-long monocrystal, the industrial manufacture of HTS tapes is faced with the challenge of equally aligning all small superconductor crystals within a HTS wire. The critical current of a HTS tape therefore depends on the texture degree and thus on the manufacture of the wire.
The HTS wires of the actual generation - second generation (2G) - consist of a metal substrate tape subsequently coated with buffer and superconductor layers. The superconductor layer is textured by an aligned growth of the superconductor layer on the buffer layer lying underneath and acting as a template. In the case of CSD processing, an already textured metal substrate is used as an economical and scalable approach. The metal substrate can be manufactured in large lengths by standard rolling and annealing processes and serve as texture template for the aligned growth of the buffer layer and the following HTS layer. The material approach selected in this project was nickel-tungsten alloy (NiW) as substrate, lanthanumzirconate (LZO) and ceria (CeO2) as buffer layers and yttrium-barium-copper-oxide (YBCO) as HTS material.
The real dimensions of the coatings and the substrate can be visualised by SEM cross sectional analysis. It indicates a total thickness of the HTS wire of about 80 µm, where the ceramic coating of the buffer and superconducting layer is less than 1µm.
Main advantages of our CSD process, in combination with innovative inkjet printing at ambient pressure are the lower investment, the faster deposition with higher yield and the processing under ambient pressure enabling a completely continuous processing. Common CSD techniques currently used include spin coating and dip coating steps. It is clear however, that spin coating cannot be used for the production of long lengths, especially when continuous production is required, something that dip coating is perfectly capable of. Drawbacks presented by dip-coating include the coating of both sides of the tape, which represent a waste of precursor solution and can potentially clog the pulleys used to transport the tape. Another important drawback is that the sol usually lies in an open tray exposed to the ambient atmosphere and therefore a loss of solvent due to evaporation is likely to destabilize the precursor solution and/or change its concentration and rheological properties, introducing undesired variations in the deposition process.
In order to overcome these drawbacks of conventional coating techniques, innovative inkjet printing approaches were successfully developed within this project. Inkjet printing offers a drop wise and therefore accurate coating of one side of the substrate. The ink reservoir is constantly sealed and under inert gas preventing pollution and evaporation of the coating solution.
Zenergy Power developed and built continuous coating devices including variable inkjet print-heads and subsequent annealing furnaces. The devices allow the mostly automatic continuous production of HTS tapes with lengths exceeding 100m at 10mm width. For different layer types (buffer and superconducting layers) actually three continuous coating devices are in operation.
The main challenge for a successful operation of inkjet printing technology is the understanding and the optimisation of the ink - print-head interaction. Thus several hundred single experiments were conducted by Zenergy Power within the EFECTS project in order to achieve a stable operation of the inkjet printing. The project partner UCAM provided an upgraded version of the printer software and operated pretests of print-head technologies (electro-magnetic and piezo-electric) leading to a clear suggestion for installation at Zenergy Power. In-house, various parameters were tested and analysed in order to achieve best homogeneity and highest performance in the printed layers: printing pattern (hexagonal and square shaped patterns were tested), printing parameters (nozzle size, opening time and frequency were optimised) and inks (viscosity, surface tension and wetting angle) were optimised.
In particular, for the ink-development a large number of inks with various additives, surfactants, solvents and mixtures thereof were synthesised, analysed and tested leading to a formulation with the best ever reported properties for printed superconducting films worldwide. The impressive homogeneity of the printed films can be seen in a movie clip available at the project homepage http://www.efects.eu
In addition to the in-house developed inks several formulations of water based inks from the project partner UGhent were tested in the continuous coating devices.
Another issue faced within the project is the commercial availability of inks. Here additional requirements such as raw materials access and stability have to be considered. Zenergy Power cooperated with a special chemistry supplier in order to investigate the commercial access to the optimised inks developed within the EFECTS project. After some successful tests Honeywell Specialty Chemicals GmbH, Seelze, Germany, was qualified to deliver large quantities of metal-organic salts and ready-made coating solutions to Zenergy Power. All results obtained within the last year of the project were based on these commercial chemicals partially slightly modified by Zenergy Power. The delivery in larger lots also has the advantage of higher reproducibility over time compared to the previously used lab scale prepared batches.
Using the commercial coating chemicals and the optimised printing technology, coated conductors with the so far best performance regarding to length and superconducting properties worldwide were manufactured.
Critical current densities in the superconducting layer well above 1.3 MA per cm2 on 10 mm wide tapes illustrate the nearly perfect growth of the ceramic layers on the metal substrate. Within the EFECTS project Zenergy Power succeeded in transferring the lab scale results towards continuous processing of tape lengths exceeding 100 m (equal to more than 1 m2 coated area). The value actually achieved of more than 9000 Am is the best ever reported for an all CSD inkjet processed HTS wire.
F. Inkjet printed photocatalytically active titania layers on glass/tiles
Self-cleaning titania
Although the photocatalytic activity of TiO2 semiconductors is known since 1972, enhancement of this phenomenon was realised years later with the progress of nanotechnology. Using the energy delivered by the sun, titania is activated, changing the material in a photocatalyst, that can break down almost any organic molecule you can thinks of. Being a colourless and in many cases, perfectly transparent material, titania can be used to apply on many underground as self cleaning coating. Probably the most well-known application is the commercialised self-cleaning window glass. Yet, one can think of applying titania on door knobs, sanitary ceramics, ceramic tiles, pavements, architectural steel, bricks, etc.
Hundreds of publications on general TiO2 nanoparticles synthesis with application on self cleaning technology or photovoltaics existed before the EFECTS project. Yet, TiO2 inks were present in the ink industry limited for colouring pigment applications. In addition, ink manufactures were using exclusively the toxic and expensive solvent based inks mainly due to the good wetting behaviour and the quick evaporation. The last years, the tendency of water based inks preparation for the printing technology is clear due to regulations and environmentally friendly requirements. Within this project, the development of water-based titania precursors, compatible with inkjet technology is the main focus. Printing titania instead of applying it with expensive vacuum methods or less controlled spraying techniques, would mean a major breakthrough from industrial point of view, opening the market for a whole new range of applications. One of the main groups of costumers there are the manufacturers of ceramic tiles, glass and metal industries.
Top-down ink synthesis
Within the project, different types of titania precursor inks were developed by Nanophos and UGhent. NanoPhos evolved a top-down technique for the low cost preparation of totally water based TiO2 suspensions suitable for inkjet printing. In particular, commercial TiO2 nanopowders are processed with ultrasounds in water medium and stabilizers are added to retain the nanoparticles dispersed in the aqueous solution. The technique is environmentally friendly with no by products during the production process and has been applied for patent protection. Deposition of these suspensions transforms the coated surfaces into self cleaning, superhydrophilic and self sterilising only by harnessing the ultraviolet (UV) light emitted by the sun.
Ti4+ solution inks
Next to this, UGhent has developed a solution based titania precursor ink, containing Ti4+ ions stabilised in water by addition of different chelating agents. Traditionally, this type of solution inks has to be prepared under strict, water-free conditions to avoid hydrolysis of the titania precursor and thus uncontrolled precipitation. Yet, the use of smartly selected chemicals allows to create a highly inert, water insensitive precursor which can be used to prepare water-based inks with long time stability. These inks were tested both in piezo-electric and electromagnetic printing setups and completely optimised for printing in electromagnetic or piezoelectric setup. Inkjet printing of this kind of inks leads to 100 to 500 nm thick titania films after heat treatment to 500 °C under air. When printed on glass, these layers are completely transparent, defectless and exhibiting very smooth surfaces with roughness values below 4nm as measured from atomic force microscopy (AFM).
The photocatalytic activity of these layers was tested by following the decomposition of a methylene blue solution, contacted with the layer upon exposure to UV light. The inkjet printed layers show a photocatalytic performance that is even better than the commercial window glass that is for sale on the construction market. Currently, wheathering and other simulation tests are being performed in order to study the performance of these layers when exposed to real outside conditions.
Low temperature processing
As discussed above, sintering temperatures as high as 500 °C are necessary to convert the aqueous titania inks into functional, selfcleaning, photocatalytic layers. Yet, for many applications on more heat sensitive undergrounds such as for example polymers, wood or some types of metals, it is important to develop perfectly transparent titania layer that can be converted into a self cleaning coating at much lower temperatures. Furthermore, reducing the energy of any industrial process is of importance given the current global environmental policies and of course continuously growing energy cost.
Therefore, UGent developed a titania precursor, again water-based, that can be converted into a aqueous suspensions containing titania nanoparticles by microwave treatment for less than five minutes at 140 °C. After this bottom-up synthesis, only solvents i.e. mainly water, need to be removed from the ink after deposition while preformed titania particles are present. This means a tremendous decrease of the temperatures needed for coating, yet still resulting in a completely transparent and very thin coating. This is in contrast to most commercial titania paints that can be simply applied without heating, which offer self cleaning power yet come with a white appearance, limiting the applications. On the other hand, the inkjet printing approach, combined with smart nano-inks, offer the possibility of creating transparent layers with similar appearance to those obtained from expensive vacuum techniques such as CVD which is currently widely used to create self cleaning glass.
The bottom-up inks were loaded into the ink cartridge of a Dimatix materials printer and printed on microscopy glass slides for testing. Heat treatments varying from 100 °C to 500 °C were applied to the printed layers, all resulting in hard, non sticky and transparent layers. Focussed ion beam (FIB) SEM cross-sectional analysis was performed as a function of sintering temperature in order to determine the sample thickness. For layers sintered below 250 °C, the amount of left-over organics is still large and therefore, the electron beam was melting the surface samples. The layer thickness further decreases upon heating to 350 °C showing that only at this temperature, most organics have disappeared from the layer. The films exhibit a very smooth surface, with roughness below 4 nm over 5 times 5µm surfaces. The photocatalytic activity of the TiO2 films was studied by following the degradation of a methylene blue dye in contact with the TiO2 as a function of exposure time to UV light. Clearly, the sample heated at 500 °C shows the highest efficiency and performance is decreasing with decreased heating temperature. Against expectations, the layer that was only heated to 150 °C, performs very well at the beginning, but is clearly levelling off at longer exposure times and is not resulting in a linear response in the plot. This is certainly related to damaging of the layers after longer times submersed in water. Optimisation and outdoor wheathering test are currently being performed.
The method to produce transparent coatings at low temperature, made from these inks combined with nanoparticles and using the inkjet technology has been the subject of a patent application.
G. Magnetoresistive device
Icmab-Cisc has produced environmental friendly, stable and efficient inks for printing both La0.7Sr0.3Mn0.3 (LSMO) magnetoresistive coatings and magnetoresistive devices. The inks designed allow coating and patterned coating over single crystal substrates, amorphous and polycrystalline with functional ceramics in a fully computer aided design (CAD) and computer aided manufacturing (CAM) scope. It allows producing customised electronic magnetoresistive devices for short series changing the design on fly, during the same production line.
The designed precursors have been tested to produce epitaxial coatings, when single crystalline substrates, by the chemical route based on the decomposition of metalorganic salts, the so called CSD-MOD process. These functional ferromagnetic and magnetoresistive epitaxial coatings have been successfully tested to produce multiferroic devices by IFW.
Magnetoresistive LSMO inks have been tested in a piezoelectric driven single nozzle printer developed in the framework of EFFECTS with a mechanical resolution of 2 mm and repeatability better than 5 mm. The nozzle of 60 mm in diameter has allowed jetting droplets between 30 and 40 pL. The use of such a large nozzle has not been a limit for the resolution expected in the drawings; it has been possible to create a smaller diameter by using second harmonic excitation. A patterned coating has been developed in order to produce magnetoresistive contactless analog encoders as an example of the capacity of the procedure.
Encoders are extensively used in the feedback of the control loop of servo-mechanisms. Feed back of the position allows a very high precision positioning by using low cost motors. Although digital encoders are available, the analogue systems allows a lower cost considering both, the encoder and the driving electronics and are extensively used from very low cost applications as in positioning back-view mirrors in the cars or seat position memory up to servo-actuators in tracers or valves and a large number of applications. Basic analogue sensors consist of variable resistors, typically carbon resistors with sliding contact. The sliding system has a short life that can be estimated in some tens of thousands of operations, in the best cases. Systems requiring larger life expectancy should avoid mechanical contact. Contactless systems show a life expectancy larger than 50 million operations. Magnetoresistance has been suggested for position sensing in substitution of mechanical encoders, including also applications for the well known potentiometres. Customisation and fabrication flexibility are two essential points for the market and inkjet printing technology allows both in a very easy way controlling the production directly from the CAD system in a real CAD/CAM conception.
A full bridge could be considered with four contact pads equi-distributed around a round path of magnetorresistive material. The magnetic assembling rotates over the circle. In order to establish the main parameters for the device, magnetoresistive behaviour has been determined. According to the results the expected change of resistance of the LSMO path for a field in the range of 0.3 T which can be achieved by NdFeB magnets is nearly a 0.3 % at the room temperature.
With this values, full bridge and half bridge Wheatston configuration allows a linear response. The viability test, previous to scaling has been realised over a LAO substrate of 10 times 10 mm2. Three geometries have been considered for printing, a square closed path, a circle and a round undulated path.
The printing regularity can affect the resistance of the patterned paths. In spite of the high precision in the positioning of the drops, local defects could introduce a great mismatch of the printed device. In order to get a better control of the position of the drops, increasing of the surface energy in the interaction with the substrate has been promoted by a treatment of the substrate for hydrophilic behaviour. The result is a continuous covering.
The resistance of each branch of the bridge is in the range of 9 kW with a mismatch lower than 10 %. A full electronics testing system has been also developed. Scaling up has been performed in association with OXOLUTIA S.L. to achieve series production. A number of 512 nozzles printheads have been tested for medium scale production achieving a 720 times 1440 dpi resolution, large enough to obtain a reliable fusion of droplets. In order to enhance the magnetic field, a magnetic shell covering the LSMO tracks has been designed. Therefore, a set of inks based on magnetic nanoparticles was synthesised. Solutions of magnetic nanoparticles have been also synthesised for the development of ferromagnetic inks able to perform patterned and continuous coatings by using direct printing over the chosen substrate. The solutions have been adapted to work as inks for direct printing of ferrite based patterns being tested the most used compositions: magnetite and manganese, cobalt, copper nickel and zinc.
The solutions showed long time stability (tested up to more than three months). Applications in electronics are in the way in order to test the possibility to construct magnetic couplings and magnetic yokes in electronics circuits.
The deposition of a magnetic shell over the magnetite increases the mean flux density over the magnetoresistive element with 15 %. This increment supposes increasing the magnetoresistive coefficient from 0.3 to more than 0.4 30 % with respect the initial value.
H. Multiferroics: proof-of-concept
The term multiferroics stands in most cases for a layer system consisting of a ferroelectric and a ferromagnetic layer, where the layers are essentially elastically coupled. Multiferroic layer systems are interesting materials for future information-technology devices in which data can be written to magnetic memory elements by applied electric fields. Furthermore, multiferroics are of great interest for fundamental research because the magnetic order can be controlled through the ferroelectric order and the other way around at a microscopic level. The elastic coupling of the layers allows to control the electric polarisation of the ferroelectric layer by a magnetic field or to alter the magnetisation of the ferromagnetic layer by an electric field. The coupling of magnetic and ferroelectric orders in a material is known as the 'magnetoelectric' effect.
There are different preparation routes of multiferroic systems which are a physical deposition, e.g. pulsed laser ablation (PLD), and a chemical deposition, e.g. CSD, technique. The physical deposition route of multiferroics is well established and multiferroic devices of good quality can be produced in a laboratory scale. The physical deposition process is based on energy and investment-intensive vacuum processes requiring vacuum conditions, the use of energy-consuming evaporation systems and the low yield and deposition speed. CSD is a promising process for a low cost fabrication of complicated coatings and high-end decorated patterns at fast deposition rates.
Inkjet printing or ink plotting of multiferroic materials is a completely new technology. Most of the literature and patents on multiferroics are related to the design and manufacture of physical deposition techniques as for instance physical laser deposition. Chemical solution plotting routes are an efficient alternative due to a low cost fabrication of completed multiferroics.
The most widely used solution preparation technique of such perovskite materials is based on methoxyethanol in combination with metal alkoxides as reagents. The control over precursor characteristics is good in this precursor preparation but the process simplicity is low and the chemicals used toxic and expensive. Processes focusing on water based precursors are much less explored but a very promising alternative, holding the promise of much more simple and cheap processing.
Barium titanate BaTiO3 (BTO) is a very intensively studied ferroelectric crystal, with a lattice constant of the tetragonal BTO crystal structure, aBTO = 3.99 Å. La0.7Sr0.3MnO3 or shortly LSMO is a ferromagnetic material with a lattice constant of aLSMO = 3.87 Å, which is very close to the value of BaTiO3 leading to tensile strain in LSMO films. LSMO offers the advantage that its lattice parameter can be finely tuned by the polarisation of the ferroelectric BTO crystal. Consequently, the magnetic properties of the LSMO film can be influenced by the polarisation of the BTO crystals.
The consortium developed aqueous, environmentally friendly precursor solutions for the deposition of BaTiO3 and La0.7Sr0.3MnO3 epitaxial layers on SrTiO3 substrates. These inks were deposited on SrTiO3 (100) substrates by ink plotting using a 'Sonoplot GIX Microplotter'. The Sonoplot ink plotter is capable of applying picoliters of fluid, continuously creating features onto a surface. The ink is loaded by capillary forces into a hollow glass needle, which is attached to a piezoelectric element. At the resonant frequency of the loaded dispenser the fluid is sprayed out of a 10 or 30 µm diameter opening at the end of the needle.
Firstly, LSMO as well as BTO single layers were plotted separately on STO substrates. The wet films were dried and annealed in a tube furnace. LSMO films were pyrolised at temperatures between 250 and 500 °C in air and crystallised at 900 °C for 5 hours in oxygen atmosphere. BTO films were pyrolised and crystallised at 500 °C and 700 °C in air. Crack formation in BTO layers can be avoided by using a slow heating rate of 60 K/h. The desirable thickness of BTO layers was obtained through multiple coating and annealing processes. In a next step BTO layers were coated on top of a LSMO film. Such a multiferroic device consists of a stack of LSMO / triple layer BTO / silver.
The requirements for the fabrication of such architectures are a very flat and homogeneous La-0.7Sr0.3MnO3 film as well as a crack free, non porous and fully isolating BaTiO3 layer on top. The magnetic properties of the LSMO layer reveal a Curie temperature TC of 360 K and a saturation magnetisation of about 280 emu/cm3, assuming film thickness of 100 nm. Polarisation measurements were carried out to demonstrate the switching processes of the BTO film when applying an electric field.
Potential impact:
General overview, scientific exploitation
Given the very successful scientific progress made during the project, the consortium as a whole has been able to generate 14 published papers in international peer-reviewed journals, another 14 submitted (or in preparation) papers in international peer-reviewed journals, 33 oral presentations on international conferences, 38 poster presentations on international conferences, two patents filed and one patent under preparation.
Furthermore, there were some collaborative dissemination activities where the entire (or the largest part) of the consortium took part:
1. NN11, Thessaloniki, Greece 2011: EFECTS session with two plenary and five talks
2. sustainable, intelligent manufacturing (SIM) 2011, Leiria, Portugal: EFECTS session with one plenary and three talks
3. Eucas2011: European conference on applied superconductivity, Den Hague, the Netherlands: invited lectures on results EFECTS related to superconductors, plus three talks and many posters
4. Seventh Framework Programme (FP7) success stories: creating economic value through effective collaboration, Suschem - European technology platform for sustainable chemistry, Poland 2011: presentations and flyers at the Polish European Presidency event.
Finally, the coordinator planned a press release on the outcomes of the project by the end of 2011. For the actual commercial exploitation of the project, the plans were organised per partner.
Project website:
http://www.efects.eu
Coordinator:
Prof. Dr Isabel Van Driessche, isabel.vandriessche@ugent.be +32-926-44433
Dr Petra Lommens, petra.lommens@ugent.be +32-926-44449
UGhent
SCRiPTs, Department of inorganic and physical chemistry
Krijgslaan 281, S3
9000 Ghent, Belgium