Final Report Summary - MULTIPLAT (Biomimetic ultrathin structures as a multipurpose platform for nanotechnology-based products)
Executive Summary:
The main scientific objectives of the project have been summarized as: The MultiPlat project will conceptualize, introduce and fabricate novel biomimetic, selective ion-conductive nanometric-thin membranes with highly ordered structure as a multipurpose platform for a range of applications.
After 36 Months duration MultiPlat finished with following main outcomes:
1. The novel proton conductive material was developed. The material has better conductivity than perfluorocarbon-sulfonic acid ionomer under stand-alone tests
2. Patented process for manufacturing of thin proton conductive membrane with robust porous supporting structure (patent application on the method to make asymmetric sol-gel nanomembranes, TUW 2011 - title: Asymmetric interpenetrating proton-conducting membrane.
3. The novel structure was integrated in a Fuel Cell breadboard model
Novel developed :
5. Novel nanofluidic devices, nanofluidic diodes and MOS transistors based on ions instead on electrons were developed and tested, high potential of the novel components for applications in nanomedicine, drug delivery and control circuits can be expected
6. The biomimetic proton conductive material is promising, and the material is considered as worth for follow-up development toward industrial application. These activities will be continued within the consortium via various means (national, industrial, and EU projects).
Project Context and Objectives:
Summary description of MultiPlat context and the main objectives
The main scientific objectives of the project have been summarized as: The MultiPlat project will conceptualize, introduce and fabricate novel biomimetic, selective ion-conductive nanometric-thin membranes with highly ordered structure as a multipurpose platform for range of applications. The functionalisation will be created by integration of selective ion- conducting nanochannels with nanomembranes in a manner analogous to that in biological cells, thus merging artificial and biological. In this way we impart the functionalities of a living structure to our artificial nano-building blocks, ensure smart behavior and enable a quantum leap in many applications of utmost importance for our present and future needs.
As the general concept described above is rather complex and worth decades of R & D, the work concept more specific has been broken down to:
- Development of nanomembranes as a matrix for the biomimetic structures.
- Parallel development of biomimetic channels comprising ion conductive functional groups.
- Integration of nanomembranes and ion conductive functional groups into biomimetic ultrathin membranes with a highly organized structure.
Within these general objectives, the particular steps toward reaching of the primary MultiPlat objectives are:
- Manufacturing of laminated composite nanomembranes samples as precursors for the biomimetic proton conductive structure.
- Experimental validation of proton conductivity in biomimetic structures without net water transport
- Integration/incorporation of ion (proton) conductive nanochannels into nano-membranes
- low proton resistivities
- Novel nanomembrane characterization techniques
- Proof of concept for proton conductive biomimetic nanomembranes at macro level
- Self cleaning of platinum catalyst
- Functional model of a new class of fuel cells
- Proof of concept of the biomimetic nanofluidic transistor
- Contribution to fundamental knowledge and innovation in the nanomembrane philosophy
Work Packages (WPs) of the Project
The essential work flow of the Projects (work tasks and subtasks) has been managed, according to the proposal, through Work Packages as:
WP1 Fabrication of nanomembranes as supports for biomimetic proton-conductive structures. The starting point is single-layer nanomembranes with homogeneous composition. Two procedures were done in parallel: fabrication of inorganic nanomembranes (TUW - ISAS) and organic nanomembranes (TUW - IAS). The next step was integration of nanofillers for improvement of mechanical, thermal and electrical properties. The final process was integration of inorganic and organic layers into a complex nanomembrane. All processes phases were permanently coordinated with characterization procedures (UNL and EPFL).
WP2 Fabrication of highly ordered biomimetic proton wire structures based on Grotthuss mechanism. The program included development of new classes of proton conductive structures: by bottom up nanotechnology and one by top-down approach. Catalyst chemical vapour deposition growth of CNTs as a scalable and easy route for their production is reported. Optimization of the CNT growth is important in order to provide the material suitable for the dispersion in polyelectrolytes in order to reinforce them mechanically and possibly improve their ionic conductivity. In the latter case, an additional surface functionalization with proton conductive groups is demanded. Additionally, the titanate nanotubes (TNTs) growth is described. TNTs were used as an alternative to the CNTs for Nafion based composite membranes preparation. The TNTs are electrically insulating and so the percolation threshold issue could be overcome. The TNTs doping via cation exchange mechanism is proposed as a convenient strategy for the TNTs surface modification. WP2 included the coordination in characterization and measurement with UNL and EPFL.
WP3 focused on fabrication of soft proton-conductive structures and nanochannel arrays. Proton conductive, electron blocking selective nanomembranes can be used.for FCs and nanosensors. Innovative structures were manufactured on nanomembrane support. All free standing and assembled nano-membranes where characterized using a variety of techniques to obtain the properties deemed necessary for their applications in PEM fuel cells. These experiments were carried out as part of project work, but also as a tool to monitor and control the processes involved as well as to verify the products of the different phases. The results where used to assess, control and make the necessary modifications.
Emphasis was laid on investigating the membranes in condition as close to working condition as possible. This resulted in samples which could be readily built into test fuel cells and evaluated. The results obtained therefore indicate a promising potential for biomimetic membranes in fuel cell applications. Further a nanofluidic transistor, diodes and logical circuits based on the ions transport within nanochannels were conceptualized, developed and tested.
WP4 dealed with the assembly of hard NTs as proton wires into nanomembranes, their vertical alignment and opening of their ends to form ion channels. Next, the proton conductive nanomembranes were assembled with electrodes. The functionalisation of the NTs with platinum nanoparticles as catalyst was developed. A Catalyst Covered Membrane (CCM) based on the innovative proton conductors was assembled with Gas Diffusion Layers (GDL) and gaskets. MEAs manufacturing processes was studied within this WP. Final assembling of the new fuel cell in the breadboard model followed by demonstration of the electrochemical performances of MEA prepared with gas diffusion layer materials containing carbon nanotubes developed by Nanocyl.
WP5 The objective of this WP was to evaluate new MEA produced in WP4 into the fuel cell breadboard model. The activities within this WP are: MEA and FC testing procedures, electrochemical measurements and cost evaluation. This evaluation aimed at summarizing the cost of the technologies developed within the project Multiplat and at comparing to the state of the art and to the future requirements. A cost breakdown has been made in order to identify at which cost target are the different elements of the PEMFC stack. This is of interest for assessing the impact of components integration and manufacturing aspects toward different time-scales and/or production rates.
WP6 The MultiPlat is a fully new strategy and has high potential of use in the field of nanomembranes. The idea of nanomembrane biomimetic structures was disseminated throughout the project in order to tailor the developments of the project to the expectations of the market. The work performed within Multiplat has led to a significant number of publications (35 scientific publications) and 43 dissemination activities,
12 prominent fields of exploitation were identified in the latest version of the Plan for use and dissemination of results PUDF-2012. The PUDF was updated regularly.
One patent application on the method to make asymmetric sol-gel nanomembranes was filed in June 2011 by TUW (title: Asymmetric interpenetrating proton-conducting membrane).
Roadmapping activities were performed. This covered the different types of fuel cells, and indicated more specifically for the applications targeted by MultiPlat, the different potential applications like automotive or portable. Among other informations the study tried to predict the future market for PEMFC and other sectors of fuel cells for the near future. Within WP6 a user-group was founded: 11 entities became member and participated in various events like MultiPlat Project workshop, received information etc.
WP7 The management of the project interacted with all the work packages and followed the work performed by the partners to guarantee the achievement of the project and its results.
Throughout the project, WP7 aimed at achieving the following objectives:
- To ensure identification of challenges and stakes to consider project complexity and scale in the organisational structure for efficient decision-making.
- To secure the allocation and coordination of all resources (human and financial) in order to reach project objectives within the pre-defined contractual and time frames.
- To assure all contractual matters and to manage all knowledge and communication issues
This could be done through clearly defined processes that concern: operational management (administrative, financial and contractual coordination) and technical management (progress follow-up, risk management, partnership management).
Project Results:
MultiPlat M36 Science and Technology
In the first part of the Report results are presented which are related to the proton conductive structures solely related to application in fuel cells. This result basically contains principal work and results from entire project Tasks and Subtasks. The second part of the Report presents the important results on the ions transport in ultra-thin channels (nanochannels), the area where the technology meet biological systems. Most of processes within the metabolism of a biological cell are actually based on control of the ions and molecules flow in and out of the cell. Hence, the second part of the Report is devoted to our results in the control of ions in a nanochannel.
The Report has a pyramidal structure. The most important results in a field are presented at the beginning of a Chapter and the partial results are presented beneath, according to its importance. The Report contains two pyramids.
At the begin MultiPlat's significant results are given compared with the claims in the proposal.
- Manufacturing of laminated composite nanomembranes samples
The Project strongly contributes to the novel filed in nanotechnology - the nanomembranes. A nanomembrane may be defined as a nanostructure with a thickness under 100 nm and a high aspect ratio exceeding 1.000,000 with areas reaching several cm2. Within MultiPlat numerous novel nanomembranes structures were developed including complex ones: layer-by-layer carbon nanotubes membranes, polymer/nanotubes multilayered composite membranes, metal oxide fibers/polymer membranes.
- Experimental validation of proton conductivity in biomimetic structures
The proton conductivity in organized nanochannels, similar to the biological channels is positively confirmed in macromolecular selfassembly structures and confined proton conductive nanochannels. Proton conductivity was also confirmed in carbon nanotubes. As the CNTs are electrically conductive they finally were replaced by electrically isolating titanium dioxide nanotubes.
- Integration/incorporation of ion (proton) conductive nanochannels into nano-membranes
The nanochannels in form of water-filled titanium dioxide nanotubes, externally oriented sulfonic groups in methylpropane sulfonic acid AMPS polymers and sulfonic groups of perfluorooctyl acrylate (PFOA) polymer are successfully incorporated in nanomembranes.
-.Low proton resistivities
The proton resistivities of best membranes amounted to 0.03 ohm/cm at 80 degC - 100 degC, the value
60 % lower than for comparable perfluorocarbon-sulfonic acid ionomers. From the practical point of view i.e. applications in fuel cells, the total resistivity of the developed material is lower than for known materials, considering the thickness ratio.
- Self leaning of precious platinum catalyst
An innovative, efficient method of leaning of precious platinum catalyst on the activated L-by-L nanomembranes was developed, tested and published.
- Novel nanomembrane characterization techniques
Characterization techniques were developed and published in Journals.
- Proof of concept for proton conductive biomimetic nanomembranes at macro level. minimal structure area 5 cm2.
The breadboard models are routinely manufactured with membranes of 25 cm2. Moreover, the novel fabrication method shows to be simple and efficient and hence applied to the patent.
- Functional model of a new class of fuel cells
The maximal operating temperature in our fuel cells was 95 - 1000C, The water flow was generally low, the effective thickness of nanomembrane before integration into the support was 1 - 5 ?µm.
- Proof of concept of the biomimetic nanofluidic transistor, exclusively operating on ionic fluids.
A functional biomimetic transistor based on ions could be demonstrated. Moreover an biomimetic array of fluidic diodes was shown. The parameters of those structures exceed the contemporary devices developed elsewhere.
- Breakthrough in the fundamental knowledge
A series of publications in recognized relevant journals evidences the contributions to fundamental knowledge.
1. Biomimetic proton conductive structure
Two concepts of biomimetic membranes were planned within the Project, stated in the Project Proposal. The first one was membrane fabrication by a sol-gel process and the second one was preparation of asymmetric membranes by delayed onset of liquid-liquid demixing using the process of immersion preparation (with and without electrical filed).
1.1 Sol-gel L3 sponge structure
First concept was based on the method of preparing a flexible nanoporous membrane starting from Sol-Gel L3 sponge phases. The necessary flexibility of the membrane was achieved using polyethylene glycol based precursors. It has been confirmed by experiments (AFM and SEM) that the pore diameter can be easily tuned within the 1-100 nm scale by adjusting the ratio of the components.
Poly (sulfone) (Mw 35,000, Mn 16,000) (PSf) and poly (ether imide) (melt index 9 g/10 min at 337o C/6,6 kg) (PEI) were obtained from Sigma-Aldrich. 2-Acrylamido-2-methylpropane sulfonic acid, 99 % purity (AMPS) and poly (ethylene glycol) 700 diacrylate (PEG 700 DA) were obtained from Sigma-Aldrich. Irgacure 2959 was obtained from Ciba. Poly (ethylene glycol) 400 (PEG 400) and n-methyl-pyrrolidone, 99 % purity (NMP) were obtained from Fluka and Sigma-Aldrich, respectively. In all experiments deionized water was used in the second gelation bath to prepare nanomembranes.
Concentrated solutions of PEI (27 wt %) and PSf (30 wt %) in NMP were prepared by mixing on a hot plate overnight. Solutions of AMPS in NMP were prepared fresh by mixing solutions in vials cooled in ice. Specified quantities of AMPS solutions were mixed with the PEI and PSf solutions, respectively. PEG 700 DA and Irgacure 2959 were added to these solutions at 10 mol % and 1 wt % based on AMPS, respectively. Polymer concentration is expressed as weight percentage of polymer per 100 g of polymer and solvent (NMP). AMPS concentration is expressed as mol of AMPS per L of solvent (NMP).
Prepared solutions were casted on a glass plate using a film applicator with a 120 mm gap (Erichsen Quadruple Applicator, Model 360), immersed in a PEG 400 bath, and quickly transferred into a high intensity (600W) UV curing unit (60 sec exposure at 90 % intensity). After UV curing, nanomembrane samples were put into deionized water to complete exchange of solvent and non-solvent and stored in a vial with deionized water overnight to complete extraction of unreacted residuals.
Two main parameters characterize performances of a membrane as the proton conductivity (for application in fuel cells): the proton conductivity alone and the ion-exchange capacity.
The first parameter, proton conductivity was measured as "in plane" of the nanomembrane samples using a 4-point method. The direct current was measured between two gold plated inner electrodes placed 1 cm apart, while keeping the potential difference between two gold plated outer electrodes constant at 100 mV. Under these conditions no significant polarization of the electrodes is expected. Nanomembrane samples, after the extraction of reaction residues for 24 hours in deionized water, were equilibrated in deionized water prior to the measurements. The electrodes were placed onto the skin side of the wet samples. Measured ohmic resistance and nanomembrane thickness (Mitutoyo contact gauge) were then converted into electrical conductivity, which should be equal to the proton conductivity under the measuring conditions.
The ion-exchange capacity (IEC) was determined by titration. Each membrane was placed in 20 ml of 0,05M sodium chloride aqueous solution for 24 h to exchange the protons with sodium ions. The ion-exchanged (hydrogen chloride) solution was titrated to pH 7.0 with 0,01M sodium hydroxide aqueous solution and the end-point was detected using a phenolphthalein indicator.
Values for conductivity obtained from our structures was in the range of 0.0290 - 0.0132 (S/sm), and for the IEC = 0.23 - 0.60 (mEq/g). When compared to standard material, the area specific resistance of nanomembrane samples is favorable to that of the commercial material. It should be emphasized that the area specific resistance of asymmetric nanomembranes includes the resistance of the microporous support of phase separated polymer and polyAMPS; therefore, the area specific resistance of the thin skin layer with biomimetic proton conductive channels is even smaller than measured.
1.2 Delayed onset of liquid-liquid demixing
The second method was coined as "Delayed onset" as such applied for the Austrian patent "Asymmetric interpenetrating proton-conducting membrane".
Briefly described, the method consists of multilayer structure; i.e. of gels and liquids carefully casted on a support. As the layers are mixable, they start to mix, primarily by diffusion. The second important fact is that the layer constituents are photosensitive. In the certain moment, the layer is exposed to the UV radiation and the membrane is polymerized in seconds. In this manner the instantaneous distribution of components are permanently "frozen".
The method is very general. It is possible to select a numerous and different layers, for practically unlimited species separation, as the cross-section of the "frozen" membrane can be symmetrical/unsymmetrical or graduated with multiple gradient.
It has been shown that the method provide proton conductive membranes with superior performances, if compared with other methods examined within the Project. As the method confirms as a novel, simple in manufacturing and enabling a wide range of products for fuel separation, beside the primary application in fuel cells, the method was applied for patent.
The method is based on preparation of asymmetric membranes with an interpenetrating proton-conducting morphology, which consists of cross-linked sulfonic acid functionalized ionomers embedded within a matrix of a thermally-resistant, glassy polymer is presented. This method combines a traditional immersion precipitation process for making membranes with photopolymerization and crosslinking of functional monomers included in the casting solution. The resulting membranes have an integral top skin layer with fine proton-conducting channels on top of a coarser proton-conducting support. In-plane conductivities of some of these membranes measured at ambient temperature were significantly higher than the conductivity of Nafion membranes, while having improved methanol barrier properties. An increase in functionality and molecular weight of crosslinking agents, as well as a selection of materials to promote delayed, rather than instantaneous liquid-liquid demixing were associated with significant improvements in membrane conductivity.
Asymmetric membranes with interpenetrating proton-conducting morphology could be of interest as membranes for direct methanol fuel cells
The second sub-method in manufacturing of biomimetic proton conductive membranes was use of fluorinated polymer nanomembranes with aligned electrical field. The new and innovative process to prepare these materials combines self-assembly of molecules with crosslinking in an electric field. Self assembly, as in biological membranes, is governed by mutual affinity between different chemical groups present in a solution. This self-assembly process create the aligned nanochannels via chaining of the nanodomains.
The preparation of nanomembranes with aligned pores implied two step processes and has for the first time been performed in this Project. Building blocks based on fluorinated monomers and monomers containing sulfonic groups were selected as they are known for their good chemical stability and proton conductivity. Since sulfonic groups can be considered as very polar and fluorinated moieties being very hydrophobic, phase separation and formation of e.g. tubular pores might occur. Furthermore, this difference in dipole moments is expected to allow for orientation of the two phases in the electrical field.
Fluorinated polymer nanomembrane with aligned pores were prepared using a two-step process from 1H,1H,2H,2H-perfluorooctyl acrylate (PFOA), 2-acrylamido-2-methylpropane sulfonic acid (AMPS), glycidyl methacrylate (GMA), photoinitiator Darocur 1173 and N,N-dimethylacetamid (DMAc) as solvent. The components were mixed in an ultrasonic bath. After purging the samples with nitrogen, they were irradiated in an UV chamber. Irradiation inducing radical copolymerization of acrylic groups was carried out to get a formulation with suficient viscosity for coating. The precured formulation was coated on a glass slide, an electrical field of 1.65 kV was applied for pore alignment and the temperature was increased to 150 ?°C to start cationic crosslinking of epoxy groups giving a solid membrane with aligned pores. Afterwards the sample was put in deionized water and electrical conductivity was measured. Subsequently, an electrical field is applied to orient the conductive polymer molecules in the thickness direction. The orientation is fixed by simultaneous crosslinking the membrane at elevated temperature.
The achieved data showed some important facts: the higher concentration of AMPS lead to much higher conductivity. Samples with high AMPS content prepared from the same formulation but with and without applying an electric field show a great difference in conductivity. Therefore we were able to show for the first time that conducting channels out of fluorinated AMPS hybrid formulations are successfully oriented in the electric field.
Final work with respect to the composition and preparation process parameters should further increase proton conductivity. Based on that data a patent should be filed within the next few months.
2. Bottom - Up fabrication of proton nanowires
Up to now we considered the proton conductive structure which in some extent emulates the existing materials, but actually has a different internal structure, more prone to biological cell membranes. The basic difference between existing materials and the developed membrane is the ordered, biomimetic proton-conductive structure of nanochannels. This membranes in not so far future promise to take a dominant position in mid and large scale fuel cells. However, witnessed with a rapid progress in miniaturized devices like smart communication devices, nomadic portable application and similar. Development of a miniature power supplies, more compact, independent from the grid and more efficient than existing ones are anticipated.
Two candidates were characterized: alumina and Si nanopores membranes. Two approaches have been tested for surface functionalization to facilitate proton transport through the membrane: Nafion filling and surface grafting of molecular monolayers. We have discussed confront critical limitations of the nanomembranes for fuel cell application.
The proton exchange membrane is one of the key issues in high performance fuel cells. The basic function of the membrane is to enable proton transport, while being simultaneously impermeable for electron and gas. Up to date, perfluorocarbon-sulfonic acid ionomer membranes are still the best known materials of this class. The main drawback is that the formation of proton conducting channels is highly dependent on humidity (water absorption) and the material therefore has limited working conditions within certain humidity and temperature range. In order to establish structured nanochannels for ion transport, a very recent progress adopted controlled nanoporous Si structures (with finite pore size: 5-7 nm-diameter) for an inorganic-organic proton exchange membrane. This results in a promising proton conductivity that is orders of magnitude higher than that of known materials at low humidity. However, the total film is tens of micrometers thick and the pores are random and branching. We design in our work straight nanopores with nanometer thickness, aiming at high throughput and efficient proton transport structures.
Clearly, within the MultiPlat project the brunt of work belongs to the miniature fuel cells. The well known scaling law intermits at some extent, and the consortium was forced to develop new ultrathin structures for new applications.
The thickness of contemporary membranes is in the range of 100 - 200 ?µm, the dimension not suitable for the devices of few centimeters large. The thickness/area ratio of the contemporary membranes is simply to low. Therefore, a large portion of project was committed to the fundamentally different structures. The basic idea behind was to form the proton-conductive structures strongly adhered to a structure which is inherently straight - the nanotubes. The starting idea was to use carbon nanotubes (CNTs) as a mast. In the time of the Project submission it was a known problem of discrimination of the conductive from non-conductive CNTs (comment: the rapid progress in nanoscience was considered). Clearly only the non-conductive CNTs could be used as a nano-mast.
The problem of CNTs separation on conductive/non-conductive was widely anticipated as one-step before solved and therefore the consortium proposed application of CNTs in the novel membranes. However, no significant progress in CNTs separation from the groups engaged in this work elsewhere (not Mutiplated related) came during the past 4 years. Hence, a redundant strategy for the nano-mast was selected: application of the titanium dioxide nanotubes (TDNs). TDNs are inherently nonconductive material, and therefore can be used as the supporting structure without restriction. Further, the TDNs have hydrophilic structure, which provides strong adhesion of the molecules carrying sulfonic group as the proton conductors.
On the other hand all mentioned considerations were done based on the theoretical work, which was also a fundamental part of the project. Therefore, it was necessary to examine both structures, with carbon and titanium nano-tubes and to compare the results. At first work and results on membranes with oriented CNTs are presented.
Commercially available CNTs were not quite suitable for the application. The form of CNTs should be vertically oriented, or in a carpet form. Hence, custom oriented CNTs were fabricated in EPFL. CNTs were synthesized by catalytic chemical vapor deposition (CCVD) technique from acetylene over a Fe-Co catalyst supported by CaCO3. The supported catalyst powder was placed into a 50 mm diameter CVD reactor. The synthesis was carried out at 640 ?°C in the presence of acetylene and nitrogen. Raw CNTs were purified during 1 day stirring in 1 M hydrochloric acid, washed and filtered with distilled water and freeze dried overnight. The average diameter and length of the CNTs are 11 nm and 10 mm, respectively. As described in literature, the realized carpet was mix of nproton/electron conductivity and therefore was unsuitable for the electrolyte applications. However, the structure filled with a proton conductive resin had enormous potential for the fuel cells electrodes, due to aligned fibers. This issue is presented later in the text.
The protonated TNDs were prepared by a two-step hydrothermal process. In a typical synthesis, 1 g of Titanium (IV) oxide nanopowder (99.7 % anatase, Aldrich) is mixed with 30 ml of 10 M NaOH (97 % Aldrich) solution. The mixture is then transferred to a Teflon-lined stainless steel autoclave (Parr Instrument Company) and heated to 130 ?°C and kept at this temperature for 72 hours. After the treatment, the autoclave is cooled down to room temperature at a rate of 1 degC/min. The obtained Na2Ti3O7 product is then filtered and washed several times with deionized water and neutralized up to pH 6.5 with the appropriate amount of 0.1 M HCl solution (Merck). During this step, sodium exchange proceeds to the formation of Na2Ti3O7 nanotubes. The sample is finally washed with hot (80?°C) deionized water in order to remove any traces of NaCl.
After the fabrication, the next step was TNCs functionalization. The TNCs was successfully functionalized with molecules containing a sulfonic acid at their end. Two different routes to the final functionalised product were evaluated: the first route was to synthesize 2-((2-phosphonoethyl)amino)ethanesulfonic acid first and then functionalize the TiO2 NT with. The second route is to functionalize TiO2 NT with vinylphosphonic acid first and then couple taurine via Michael addition reaction to the functionalized surface. For both methods it has been confirmed by FTIR spectroscopy that the TiO2-NT were functionalised. The second route is to prefer because unreacted can be recovered.
The second pathway was application of nanoporous silicon membranes (pore diameter 30 -60 nm, thickness 200-250 nm). It has been shown that polymerization of proton conductive monomers inside silicon membranes with pores in the micrometer range is possible. However, these nanomembranes were very fragile and first attempts to functionalize them resulted in a collapse of the membrane in the solvent. We solved the problem by deposition of the solvent from the vapor phase into the membrane as described in more detail in the experimental section. The surface reaction was done according to the known procedure. The silicon surface was first oxidized to increase the number of reactive hydroxyl groups. The membrane was then put into an acidic aqueous alcohol solution of the alkoxy silane for 30 min, rinsed with ethanol and cured overnight at 110?°C.
In contrast to previous studies we have chosen phosphonic acid groups for functionalization because they have higher hydrolytic stability than silanes. Similar as TiO2 in earlier studies, our functionalized TiO2 NT can increase water uptake, conductivity and mechanical strength and decrease fuel crossover. Furthermore, TDN can be oriented in an electric field to provide anisotropic behaviour. Using silicon membranes with nanopores instead of micropores increases the surface area for functionalization and improves the barrier properties for larger fuel molecules like methanol.
Next the work on supporting structures for the miniature fuel cells is presented. In reality, during the course of the project, this work flew in parallel with work on the proton conductive structures. In the standard size fuel cells, the proton conductive membrane has a double function: to separate protons from water flow and to provide the necessary mechanical stability. In contrary, in the miniature cells, those functions are separated: the proton conductive structure should enable a maximum conductivity, and the support and stability is deliberated to a separate structure - in our case to arrays of parallel nanochannels within an MEMS membrane. Three materials were processed, accordingly to knowledge base of partners within the Project: ultrathin monocrystalline silicon nanomembranes and ultrathin (50 nm) nanomembrane from ceramics (silicon nitride and aluminum dioxide or alumina).
The silicon nitride supporting membranes were fabricated as follow: first, a thin (100-500 nm thick) low stress silicon nitride (LS SiN) layer is deposited by low pressure chemical vapour deposition on a silicon wafer. Then, electro-beam lithography (EBL) is used to define the nanohole arrays in commercially available ZEP resist. The EBL pattern is then transferred to the silicon nitride layer by using carbon-fluorine based reactive ion etching. Finally, the LS SiN membranes are released by a partial deep reactive ion etching (DRIE) of the silicon followed by KOH wet etching for the final membrane release. KOH membrane release is preferred over dry etching since LS SiN is highly inert to KOH compared to dry Si etching, preventing possible damage to the nanopatterned membranes. Using this approach, membranes with arrays of 50-200 nm wide nanoholes and 50-300 nm interhole spacing have been fabricated on 100 ??m size LS SiN membranes.
In the later phase of work we were more directed toward low cost, self-organized processes. Actually this work belongs to the biomimetic approach. The first biomimetic work was manufacturing of alumina nanochannels.
The alumina nanochannels are fabricated by the anodization of aluminum in an electrochemical cell. No mask is required for anodization process (low cost process).
Up to date, much research has been done on high aspect ratio nanopores by using a pure thick aluminum film or plate. Anodization in proper conditions allows the creation of a self-ordered array of cylindrical pores with diameters ranging from sub 10 nm to 150 nm and depths exceeding 100 ??m. The pore dimensions and periodicity are controlled by the electrolyte composition, temperature, and current-voltage settings. Typically, the aluminum anodization process results in porous alumina (Anodized Aluminum Oxide AAO) with a volume expansion are about 1.5 and porosity of about 10%. Here we combine anodization and micromachining techniques to fabricate nanoporous thin alumina membranes. Compared to commercial bulk AAO membranes (thickness > 10 ??m), they have the advantage of providing low fluid flow resistance and a more flexible membrane geometrical configuration for the studies of ion transport. The fabrication process was adopted as follow: 3 ??m photoresist is spin-coated on to the back of a 4 wafer with double side 500 nm LS SiN. UV photolithography is used to pattern the back of the wafer and to release the LS SiN membranes by Si dry- and wet-etching. Then, a thin Al film (in hundreds of nms) is deposited on the wafer front side. Anodization is performed in 3% (v/v) H2SO4 solution by applying 20 V potential between the Al film and a Pt mesh. This creates a self-ordered array of cylindrical alumina nanopores on the full wafer surface. Alumina nanopore membranes are obtained by the final removal of SiN support layer. The side length of the membrane is determined by the backside mask opening, which is typically 200 - 400 ??m. The membrane thickness can be tuned in a straight-forward way from hundreds of nm to ??m, depending on initial thickness of Al film and final plasma etching of alumina.
Finally experiments were done with porous silicon. As the silicon offers better mechanical and chemical stability than alumina, silicon was choosen as material for the nanochannel fabrication. In addition, the silicon membranes can have well-defined perpendicular channels in the nanometre range offering unique possibilities for selective membranes. By surface functionalization with active molecules they can be modified for different applications. For example, grafting the sulfonic groups would lead to proton conducting membranes. The surface was modified with alkoxysilanes using standard procedure and confirmed the reaction by contact angle measurements. This procedure can be adapted to add different functional groups for other applications.
3. Breadboard model, test of standard, mini and micro fuel cells
The final characterization of the components in the conventional cell assembly has been carried out according to the guidelines put forward by the EC funded FP5 thematic network project FCTESTNET.
The experimental set-up for fuel cell tests and the electrochemical techniques used was then described in detail. Data obtained for commercially available membranes based on perfluorocarbon-sulfonic acid ionomers was used as reference toward the new MultiPlat membranes.
Catalyst Layer: The catalyst inks are made with carbon supported platinum catalyst powders supplied by Tanaka, a 5% perfluorocarbon-sulfonic acid ionomer emulsion, millipore water and ethanol. An ultrasonic pulverisation system was used to spray the catalyst ink onto the membrane (CCM) or the gas diffusion layer (GDE) support. The support that was to be coated was placed on the system vacuum table which was heated to 80?°C. The pulverisation system software allows the operator to control the three-axis motion and the speed of the spraying nozzle, the shape of the spraying pattern and the catalyst ink flow rate. The platinum catalyst loading can thus be easily controlled by the parameters used for pulverisation; the standard platinum loading at PaxiTech is 0.5 mg Pt /cm?².
MEA Assembly using GDE: Since the mechanical and thermal resistances of the new MultiPlat membranes are not known precisely, preliminary tests shall be carried out using MEA where the GDE and the gaskets are simply placed on the membrane. If the new MultiPlat membranes show no sign of gas crossover or electric short circuiting during fuel cell tests, standard MEA assembly procedures shall then be used. Standard MEA assembly involves placing a GDE and a gasket on either side of the membrane and hot-pressing at 140?°C. Hot-pressing was carried out at 140?°C in two steps: no pressure for 3 minutes then 600N/cm?² for 5 minutes. It is highly possible that hot-pressing temperature and pressure conditions may need to be modified for the new MultiPlat membranes.
3.1 Fuel cell Tests
- Open Circuit Voltage Test (OCV): The test provide permeability of the reactant gases through the proton conducting membrane has a very important influence on fuel cell performances. Hydrogen crossover and the subsequent direct recombination of hydrogen and oxygen will result in a decrease in the open circuit voltage (OCV), reduced power density and fuel cell efficiency and an acceleration of membrane degradation and hence reduced MEA lifetime. A measure of the OCV of the MEA as a function of time is carried out at different cell temperatures.
- Electrical Resistance Measurement: Basically characterize the fuel cell toward long term degradation. As with crossover, excessive electrical conduction through the membrane results in degradation of fuel cell performances and reduced MEA lifetime. The ohmic resistance across the MEA is measured before and after fuel cell tests with a multimeter.
- Chronopotentiometry (I pulse): A constant current is applied to the cell and the potential measured as a function of time. The fuel cell is tested under different test conditions which are presented in the Table 1 below. The current applied depends on the size of the test cell and the membrane performances, starting at a minimum of 0.2 A /cm?² and increasing gradually.
3.2 Test results
At total six asymmetric proton conductive nanomembranes samples prepared using a sol-gel process by TUW laboratories were tested in H2 /Air fuel cell. The open circuit voltages obtained for all of these membranes indicate that gas permeation is occurring across the membrane. The fuel cell performances were lower compared to reference data obtained with a NafionG? PEM membrane. Three of the sample membranes could be operated up to cell temperatures of 50?°C resp. 60?°C. Reduction of gas permeation across these nanomembranes is recommended for further improvement of their potential as PEMFC membranes.
The three Nanocyl GDL materials were tested on the cathode side of a 25cm?² H2 /Air fuel cell. There were two types of GDL compositions; one consisted of incorporating the CNT within the carbon non-woven support material and the other of applying a CNT layer onto one side of the carbon non-woven support material. Better fuel cell performances were observed for the GDL with 17wt% CNT content than for the 2 GDL with a CNT layer for the first test. Data obtained for the two GDL with a CNT layer showed that water management needs further optimisation and mass transport overpotentials were observed as the CNT layer thickness increased. MEA flooding was observed for all three GDL samples within 24h of testing. The porosity and thickness of the CNT component of these CNT-GDL materials needs to be optimised so as to improve water removal mechanisms during fuel cell operation.
3.3 Lesson learned
The novel proton conductive membranes developed within the MultiPlat project offers better proton conductivity compared to perfluorocarbon-sulfonic acid ionomer in stand alone test, but lag somewhat behind the known materials in MEA test. This result is encouraging.
The future improvement of the MultiPlat material has to be directed in:
- more and detailed lifespan tests, out of scope of Multiplat Project
- better crossover toward methanol, if the target application is Direct methanol fuel cells. The future application of that fuel cell type is bright, as the methanol storage is more convinced than hydrogen.
- the internal structure of the membrane material is biomimetic (i.e. consist of oriented nanochannels as required within the proposal). Future work scould be directed toward better cannel orientation and particularly to efficient inter-cannel connections, which will provide reduce in the total membrane resistance
4. Advanced devices, fluidic diode, fluidic transistor and fluidic electronic in general
From the practical point of view, transport of hydrogen ion in narrow channels, particularly externally controlled transport have a profound importance in understanding of the vital biological processes, but also for various man-made devices, such as sensors, batteries beside the described application in fuel cells. Protons are very specific ions which generally exist in electrolytes only in the form of hydronium ions (H3O+) associated with clusters of water molecules. Owing to the specific "hopping mechanism" of transport, the protons exhibit abnormally high mobility compared to the related cations as Li+, Na+ or K+. Furthermore, the proton transport in uninterrupted 1-D water chains (water nanowire) is ensured through fast transport of charge defects rather than through the transport of protons themselves. The accurate mechanism of proton transport in water remains not fully understood, despite decades of research. This was the motive for including this work in the MultiPlat project.
Two methodologies are available and applied here for the experimental investigation of proton transport in narrow channels: the examination of proton transport though macromolecules containing a well defined nanochannel filled with water or through the artificial nanochannel which enables us controlled modification of principal device parameters within a single device. For the first method it was necessary to synthesize different organic compounds which contain nanochannels with fixed parameters (i.e. diameter, surface states, etc.). The second approach implies fabrication of an artificial device with integrate nanochannels, fluid reservoirs and control structure - the gate. In this case, the effective cross section of a nanochannel or nanochannels array can be varied solely by variation of an external electric field.
Two groups worked in parallel on research of the ions transport within nanochannels, primarily protons - TU Wien and EPFL. Both groups share theoretical part (modeling) of the ion transport. The idea behind research on this field is based on many common physical effects with electrons in solid state materials. Ions and electrons reassemble each other in many ways. Both can be calculated using Drude model. Moreover, both flow by diffusion and drift mechanism. Additionally, the thermal generation of electrons and holes in semiconductors is analogous to the thermal dissociation of water molecules. As usual more details about this interesting topic is presented in the reports and published papers.
The experimental work was on the fluidic diodes was done at EPFL and on the fluidic transistor at TU Wien.
(Nano)fluidic diode: The basic structure for fluidic diode were hetero-structured nanopore membranes, which strongly emulate the biomimetic directional ion conduction channels. The membrane is composed of Al2O3 and SiO2(Si) layers with dense nanopore arrays. On exposure to a pH-neutral solution, Al2O3 surfaces possess net positive charges, while native dioxide layers of Si possess net negative charges. Due to the asymmetric surface charges, rectification in ionic current can be observed through the nanopore membrane, behaving as arrays of nanofluidic diodes at relatively high electrolyte concentrations.
Thanks to the opposite surface charge states of Al2O3 (positive) and SiO2 (negative), the membrane exhibits clear rectification of ion current in electrolyte solutions with very low aspect ratios compared to previous approaches. Furthermore, the Al2O3/W combination allows the application of an electrical field to further tune the ionic transport through the nanopores with low gate potentials and ultra low gate leakage current. The hetero-structured nanopore arrays provide a valuable platform for high throughput applications such as molecular separation, chemical processors and energy conversion.
The rectification factor is 6 at 1 mM KCl solution. By substituting the Si with a W layer following thermal oxidation to grow WOx dielectric, electrical potential can be applied to further modulate the ion transport across the membrane. The rectification factor can be tuned from 2 to 11. Although the finite rectification ratios for these devices deviate from the ideal diode characteristics, the development of such heterostructured nanopore membranes would enable various high throughput applications in nanofluidic chemical processors and molecular separations.
The nanofluidic diode behavior of the Al2O3/SiO2(Si) membrane is demonstrated in KCl electrolyte at varied concentrations ranging from 0.1 mM to 2 M. The membrane is inserted between two reservoirs with the same ionic concentration and liquid pressure. A pair of Ag/AgCl electrodes is used to apply a potential through the membrane. The electrodes are fabricated by the anodization of Ag wires in HCl solution. For our liquid cell configuration, low porosity samples (i.e. high resistance) are preferred in order to minimize the contribution from cell geometries to ionic current through the membrane.
The method was quite innovative, since the most existing experiments rely on planar geometry where the nanochannels are generally very long and shallow with large aspect ratios. The nanometer limiting dimension, which is normally the channel height, is defined by the thickness of a sacrificial silicon layer while the channel width and length vary from micrometers to millimeters. Based on the "etching followed by bonding" scheme, various types of experiments have been carried out to demonstrate the principle of ionic and molecular transport in one or several nanochannels. However, device minimization and throughput scaling remain significant challenges.
(Nano)fluidic transistor. The transistor, including here developed fluidic MOS field effect transistors are more complex device than diode, but also more versatile. If we use previously described analogy between semiconductors and fluids, the flow of ions can be controlled by an external electrical field, the same mechanism behind semiconductor MOS transistor. Similar to semiconductor MOS equivalent, the fluidic transistor basically consists from (nano)channel, contact electrodes and the gate.
The nanochannels and contact electrodes (water reservoirs) were manufactured in Pyrex wafers. The position and the size of nanochannels was defined using standard photolithography. Nanochannels were plasma etched 4 ?µm wide and 155 nm deep. We made various test structures with different nanochannels lengths (10 ?µm, 60 ?µm and 460 ?µm). Also the number of channels per chip was varied (1, 5, or 10) in order to have larger variety of different test samples for the following up characterization measurements. After processing of the nanochannels, the electrolyte reservoirs were wet etched to a depth of about 18,4 ?µm in highly concentrated HF solution (40%) by using a Cr/Au masking layer (50 nm/400 nm). Reservoirs were made large enough to hold sufficient quantities of fluid during characterization. In this way, the impact of fluid evaporation on measurement results was significantly reduced. Additionally, larger dimensions of the reservoirs enabled good electrical contact to the electrolyte. The same hard mask based on the Cr/Au bi-layer used for the etching of the glass substrate in HF was also used to pattern via a lift-off process the Ta/Pt (10 nm/200 nm) electrodes. The processed glass wafer was subsequently cleaned and cat into the chips with size 10 x10 mm.
The transistor gate was made of dry oxidized silicon. The starting material was a silicon wafer (thickness 350?µm) coated with 250 nm thermal SiO2 and 70 nm of LPCVD Si3N4 The standard processing was used for the deep wet etching of Si in 40% KOH. Next, the hard mask was completely removed from both wafer sides. The wafers were once again plasma cleaned, followed by dry oxidation process for growing of the high quality gate oxide layer (100 nm SiO2).
As mentioned previously, the test structure reassembles the structure of the MOS transistor, where the flow of electrons in the semiconductor channel is replaced with flow of protons/ions in the nanochannel under the applied potential. This fact inspires the authors to use the standard semiconductor parameter analyzer for testing of the device. Prior to any measurement, the device was cleaned with methanol, and then the left reservoir was filled with water. After the delay of 15 min, the time period enough long for capillary diffusion of water from reservoir thought entire nanochannels length, the opposite reservoir was filled with water. Thanks to this procedure, there is no need for application of an external pressure source for filling the nanochannels.
The first measurement sets were verification of the fundamental theoretical consideration of the nanofuidic device - that the current flow within the nanochannel is carried solely by ions flow and not generated inside the device within some electrochemical reaction. A few preclusions to prevent this effect were carried out from the begging of the design process. The S and D electrodes were made symmetrical and from inert Pt metal, next the projected maximal operating voltage of the device was chosen < 1V, below of the potential of the water electrolysis which is -1.23 V at 25 ?°C. It should be considered first, that the nanofluidic transistor is a relatively new device, with few examples previously described. Next, described devices solely relay on the water as fluid with various ions concentration. Pure water is only slightly conducting due the self-ionization of water as
2 H2O H3O+ + OH at app. dissociation rate of 1 event per molecule in every 10 h. As consequence, pure water is very slight electrical conductive with conductivity level of S ~ 0.055 ?µS-cm1 at 25 ?°C. In the same time the water is the best solvent known and even a minute concentration of contaminants within the nanochannel notably increases the conductivity of water. For example, contamination water with NaCl salt at level of 10 - 7 increase the water conductivity for 2 orders of magnitude.
In the next measurement set(s), the nanofluidic devices were connected in the same manner as an ordinary MOS transistor on the semiconductor parameter analyzer. Numerous measurements were done on devices with various combinations of nanochannels length and number. As expected, the signals obtained from devices with single, 460 nm long nanochannels were generally noisy and substantially unstable. However, the plots from devices with nanochannels arrays and particularly with shorter nanochannels were surprisingly stable and reproducible. The fundamental conditions for coexistent results were: the water quality, sufficient delay between the sample reparation and measurement in the aim to reach equilibrium in the ion concentration within the nanochannel and prevention of fluid evaporation.
The transistor transconduntance gm=2?Iout/?Vin were in average in the range of 4 - 8 x10-6 S, a significantly lower value compared to the contemporary 45 nm MOS transistors with gm in the range of 1-30 mS. The transconductance value obtained from early devices may seems to be modest, however the biological nanofluidic transistors, either based on the same principle, have 1-2 order of magnitude higher transconductance than contemporary Si devices.
As conclusion it can be stressed that the results of the experiments confirmed that the confinement of the effective cross-section of nanochannels via external electrical field can modulate the proton/ion flow in the nanochannel. The novel device worked at a low gate voltage of < 1 V and had a higher transconductance (i.e. by a factor of 10) than those reported in the previous publications, due to a thin gate insulation. Also important is the fact that the experimental results are in good agreement with theoretical predictions.
Implementation and evolution of EU policies
EU Fuel Cells and Hydrogen Joint Technology Initiative analyzed the current situation in the R&D efforts in Europe. :
Current situation of hydrogen and fuel cells (FC/H2) research in the EU is:
- The research needed is often so complex that no single fuel cell company or public research institution can perform it alone
- National funds available for FC/H2 are dispersed
- There are still many technical and non technical barriers to overcome before widespread commercial availability is possible.
- Strong R&D competition is coming from other global players (not only US/Japan but increasingly China)
- There is no agreed long-term budget plan and strategic technical and market objectives to encourage industry to commit more of their own resources;
- There is insufficient integration of the EU R&D programme (from fundamental research through to large-scale EU-level demonstrations);
- Technical breakthroughs are needed to improve performance and durability and reduce system costs to meet the expectations of potential customers. (end of citation)
The project was relevant to several other European documents, including EC Energy Package (23 January 2008). MultiPlat is in line with the Fuel cell technology platform and its strategic research agenda.
Improvement of European social and economic cohesion
The strengthening of economic and social cohesion is one of the three basic objectives of the European Union, along with Economic and Monetary Union and the completion of the Single Market. The course of action to reach these objectives, set out at Lisbon and Stockholm, includes an economy based on knowledge, competitiveness, innovation, investment in people, the fight against social exclusion, full employment and the search for sustainable economic growth.
Energy is at the very foundation of a peaceful and prosperous society. Thus the MultiPlat project directly contributed to European and economic cohesion.
Quality of life
Through its research results, MultiPlat has direct impact on the quality of our lives. The applicability of fuel cells in vehicular technology is well known and well appreciated. Besides curbing CO2 emission, which is vital to abating the greenhouse effect, they improve the quality of life in cities by powering electric motors, thereby reducing noise, and by eliminating local pollution caused by nitrogen oxide (NOx), particulates and other emissions.
Other applications influencing the quality of everyday life include mobile devices like cell phones, GPS, portable computers, etc., where fuel cells will bring much longer device autonomy and better performance. In addition to that, the replacement of traditional batteries with a product with longer lifetime can help decrease the amount of toxic waste. Portable medical devices are another highly important field where quality of life can benefit from FCs, since they ensure significantly better and more dignified life to many people with serious and sometimes disabling health problems.
Employment
The results of such a strongly applicative project as MultiPlat bear an intensive impact to the employment through opening new workplaces and offering new opportunities. The most direct impact is through the production of FCs for various applications, but most important ones are in automotive propulsions and mobile devices. Secondary influence is much wider, through industries utilizing these devices, while its indirect impact covers an even larger segment of workforce. Here we consider only a single application of the MultiPlat that of fuel cells in automotive industry.
European automotive industry directly employs about 2.000.000 workers, while the secondary workforce is about ten times larger. The forecasts predict production of 20 millions cars until 2020, and about 1-5 millions of these can be expected to be fuel cell-powered.
Thus a large portion of the automotive industry will be connected with the production of FCs, worldwide and in the EU. Therefore it is not the question if the European workforce will be engaged in the FCs production, but whose licence and know-how will be applied. MultiPlat contributed to the efforts that the novel generation of the fuel cells emerges as a European technology.
Environment
The next twenty years will be critical, both for climate change and for security of supply. In this period, irreversible and catastrophic damage may be done to the climate.
A fuel cell running on pure hydrogen is a zero-emission power source. Some stationary fuel cells use natural gas or hydrocarbons as a hydrogen feedstock, but even those produce far less emissions than conventional power plants.
A direct application of the fuel cell is the transportation. Fuel cell vehicles are the least polluting of all vehicles that consume fuel directly.
- Fuel cell vehicles operating on hydrogen stored on-board the vehicles produce zero pollution in the conventional sense. Neither conventional pollutants nor green house gases are emitted. The only byproducts are water and heat.
- The simple reaction that takes place inside the fuel cell is highly efficient. Even if the hydrogen is produced from fossil fuels, fuel cell vehicles can reduce emissions of carbon dioxide, a global warming concern, by more than half.
- Fuel cells used as auxiliary power units (APUs) to power air conditioners and accessories in over-the-road trucks could reduce emissions by up to 45% from long haul vehicles, and deliver economic benefits to the truck owner in lower fuel use and less wear and tear.
Over this period also, the costs of dependence on very few large producers of energy might also become painfully apparent. Europe can meet these challenges only through the application of technology and the adoption of difficult changes to social values and behaviour. Recent studies suggest that a decentralized energy system (also known as a "distributed" system) would improve overall efficiency through: 1.) reduced losses over power lines since electricity would no longer need to travel hundreds of miles from where it is produced to where it is used, and 2.) reduced investments in the transmission and distribution infrastructure. Furthermore, technologies, such as fuel cells, suited for distributed applications, generally produce less harmful pollution than the large central-station power plants used today. Because fuel cells convert the fuel to electricity through an electrochemical process rather than a combustion process typical of most power plants, the emissions are much cleaner. Compared to burning fossil fuels like coal and oil, which produces emissions of sulfur dioxide, nitrogen oxide, and carbon dioxide, the electrochemical process used in fuel cells only has carbon dioxide and water as byproducts. The low emissions from fuel cells make them an environmentally preferred form of power production. However, it should be emphasized that the first generation of fuel cells will likely operate on natural gas or propane, which are finite fossil fuels whose extraction from the ground and delivery produce negative environmental impacts. In the future, fuel cells will run on gas derived from biomass (plant matter) or pure hydrogen extracted from water using wind or solar energy, thus playing a key role in ushering in a sustainable energy future.
Exploitation plan
After 36 months of MultiPlat's duration 12 exploitation results have been identified and agreed within the consortium:
Production process of nanomembranes as supports for biomimetic proton-conductive structures : new process to make ion selective membranes
- Flexible Sol Gel-based Hybrid and Fluorinated Polymer nanomembranes
- Surface modification with proton conducting macromolecules
- Fabrication process of vertical nanochannel arrays
- Growth of NTs on substrates
- Functionalization of the NTs with nanoparticles
- Nanocomposites CNT/membranes
- Membrane electrode assembly (MEA)
- Portable fuel cells
- Training material (product)
- Monitoring of ion transport in fuel cell membranes
- Combined scanning probe (SPM) and phase contrast acoustic (PSAM) microscope
these results represent the most prominent fields of exploitation to be followed in the future (after the expiration of MultiPlat).
list of Websites:
website: http://www.multiplat.net
contact: Werner Brenner, Vienna University of Technology, Institute of Sensor and Actuator Systems, werner.brenner@tuwien.ac.at
The main scientific objectives of the project have been summarized as: The MultiPlat project will conceptualize, introduce and fabricate novel biomimetic, selective ion-conductive nanometric-thin membranes with highly ordered structure as a multipurpose platform for a range of applications.
After 36 Months duration MultiPlat finished with following main outcomes:
1. The novel proton conductive material was developed. The material has better conductivity than perfluorocarbon-sulfonic acid ionomer under stand-alone tests
2. Patented process for manufacturing of thin proton conductive membrane with robust porous supporting structure (patent application on the method to make asymmetric sol-gel nanomembranes, TUW 2011 - title: Asymmetric interpenetrating proton-conducting membrane.
3. The novel structure was integrated in a Fuel Cell breadboard model
Novel developed :
5. Novel nanofluidic devices, nanofluidic diodes and MOS transistors based on ions instead on electrons were developed and tested, high potential of the novel components for applications in nanomedicine, drug delivery and control circuits can be expected
6. The biomimetic proton conductive material is promising, and the material is considered as worth for follow-up development toward industrial application. These activities will be continued within the consortium via various means (national, industrial, and EU projects).
Project Context and Objectives:
Summary description of MultiPlat context and the main objectives
The main scientific objectives of the project have been summarized as: The MultiPlat project will conceptualize, introduce and fabricate novel biomimetic, selective ion-conductive nanometric-thin membranes with highly ordered structure as a multipurpose platform for range of applications. The functionalisation will be created by integration of selective ion- conducting nanochannels with nanomembranes in a manner analogous to that in biological cells, thus merging artificial and biological. In this way we impart the functionalities of a living structure to our artificial nano-building blocks, ensure smart behavior and enable a quantum leap in many applications of utmost importance for our present and future needs.
As the general concept described above is rather complex and worth decades of R & D, the work concept more specific has been broken down to:
- Development of nanomembranes as a matrix for the biomimetic structures.
- Parallel development of biomimetic channels comprising ion conductive functional groups.
- Integration of nanomembranes and ion conductive functional groups into biomimetic ultrathin membranes with a highly organized structure.
Within these general objectives, the particular steps toward reaching of the primary MultiPlat objectives are:
- Manufacturing of laminated composite nanomembranes samples as precursors for the biomimetic proton conductive structure.
- Experimental validation of proton conductivity in biomimetic structures without net water transport
- Integration/incorporation of ion (proton) conductive nanochannels into nano-membranes
- low proton resistivities
- Novel nanomembrane characterization techniques
- Proof of concept for proton conductive biomimetic nanomembranes at macro level
- Self cleaning of platinum catalyst
- Functional model of a new class of fuel cells
- Proof of concept of the biomimetic nanofluidic transistor
- Contribution to fundamental knowledge and innovation in the nanomembrane philosophy
Work Packages (WPs) of the Project
The essential work flow of the Projects (work tasks and subtasks) has been managed, according to the proposal, through Work Packages as:
WP1 Fabrication of nanomembranes as supports for biomimetic proton-conductive structures. The starting point is single-layer nanomembranes with homogeneous composition. Two procedures were done in parallel: fabrication of inorganic nanomembranes (TUW - ISAS) and organic nanomembranes (TUW - IAS). The next step was integration of nanofillers for improvement of mechanical, thermal and electrical properties. The final process was integration of inorganic and organic layers into a complex nanomembrane. All processes phases were permanently coordinated with characterization procedures (UNL and EPFL).
WP2 Fabrication of highly ordered biomimetic proton wire structures based on Grotthuss mechanism. The program included development of new classes of proton conductive structures: by bottom up nanotechnology and one by top-down approach. Catalyst chemical vapour deposition growth of CNTs as a scalable and easy route for their production is reported. Optimization of the CNT growth is important in order to provide the material suitable for the dispersion in polyelectrolytes in order to reinforce them mechanically and possibly improve their ionic conductivity. In the latter case, an additional surface functionalization with proton conductive groups is demanded. Additionally, the titanate nanotubes (TNTs) growth is described. TNTs were used as an alternative to the CNTs for Nafion based composite membranes preparation. The TNTs are electrically insulating and so the percolation threshold issue could be overcome. The TNTs doping via cation exchange mechanism is proposed as a convenient strategy for the TNTs surface modification. WP2 included the coordination in characterization and measurement with UNL and EPFL.
WP3 focused on fabrication of soft proton-conductive structures and nanochannel arrays. Proton conductive, electron blocking selective nanomembranes can be used.for FCs and nanosensors. Innovative structures were manufactured on nanomembrane support. All free standing and assembled nano-membranes where characterized using a variety of techniques to obtain the properties deemed necessary for their applications in PEM fuel cells. These experiments were carried out as part of project work, but also as a tool to monitor and control the processes involved as well as to verify the products of the different phases. The results where used to assess, control and make the necessary modifications.
Emphasis was laid on investigating the membranes in condition as close to working condition as possible. This resulted in samples which could be readily built into test fuel cells and evaluated. The results obtained therefore indicate a promising potential for biomimetic membranes in fuel cell applications. Further a nanofluidic transistor, diodes and logical circuits based on the ions transport within nanochannels were conceptualized, developed and tested.
WP4 dealed with the assembly of hard NTs as proton wires into nanomembranes, their vertical alignment and opening of their ends to form ion channels. Next, the proton conductive nanomembranes were assembled with electrodes. The functionalisation of the NTs with platinum nanoparticles as catalyst was developed. A Catalyst Covered Membrane (CCM) based on the innovative proton conductors was assembled with Gas Diffusion Layers (GDL) and gaskets. MEAs manufacturing processes was studied within this WP. Final assembling of the new fuel cell in the breadboard model followed by demonstration of the electrochemical performances of MEA prepared with gas diffusion layer materials containing carbon nanotubes developed by Nanocyl.
WP5 The objective of this WP was to evaluate new MEA produced in WP4 into the fuel cell breadboard model. The activities within this WP are: MEA and FC testing procedures, electrochemical measurements and cost evaluation. This evaluation aimed at summarizing the cost of the technologies developed within the project Multiplat and at comparing to the state of the art and to the future requirements. A cost breakdown has been made in order to identify at which cost target are the different elements of the PEMFC stack. This is of interest for assessing the impact of components integration and manufacturing aspects toward different time-scales and/or production rates.
WP6 The MultiPlat is a fully new strategy and has high potential of use in the field of nanomembranes. The idea of nanomembrane biomimetic structures was disseminated throughout the project in order to tailor the developments of the project to the expectations of the market. The work performed within Multiplat has led to a significant number of publications (35 scientific publications) and 43 dissemination activities,
12 prominent fields of exploitation were identified in the latest version of the Plan for use and dissemination of results PUDF-2012. The PUDF was updated regularly.
One patent application on the method to make asymmetric sol-gel nanomembranes was filed in June 2011 by TUW (title: Asymmetric interpenetrating proton-conducting membrane).
Roadmapping activities were performed. This covered the different types of fuel cells, and indicated more specifically for the applications targeted by MultiPlat, the different potential applications like automotive or portable. Among other informations the study tried to predict the future market for PEMFC and other sectors of fuel cells for the near future. Within WP6 a user-group was founded: 11 entities became member and participated in various events like MultiPlat Project workshop, received information etc.
WP7 The management of the project interacted with all the work packages and followed the work performed by the partners to guarantee the achievement of the project and its results.
Throughout the project, WP7 aimed at achieving the following objectives:
- To ensure identification of challenges and stakes to consider project complexity and scale in the organisational structure for efficient decision-making.
- To secure the allocation and coordination of all resources (human and financial) in order to reach project objectives within the pre-defined contractual and time frames.
- To assure all contractual matters and to manage all knowledge and communication issues
This could be done through clearly defined processes that concern: operational management (administrative, financial and contractual coordination) and technical management (progress follow-up, risk management, partnership management).
Project Results:
MultiPlat M36 Science and Technology
In the first part of the Report results are presented which are related to the proton conductive structures solely related to application in fuel cells. This result basically contains principal work and results from entire project Tasks and Subtasks. The second part of the Report presents the important results on the ions transport in ultra-thin channels (nanochannels), the area where the technology meet biological systems. Most of processes within the metabolism of a biological cell are actually based on control of the ions and molecules flow in and out of the cell. Hence, the second part of the Report is devoted to our results in the control of ions in a nanochannel.
The Report has a pyramidal structure. The most important results in a field are presented at the beginning of a Chapter and the partial results are presented beneath, according to its importance. The Report contains two pyramids.
At the begin MultiPlat's significant results are given compared with the claims in the proposal.
- Manufacturing of laminated composite nanomembranes samples
The Project strongly contributes to the novel filed in nanotechnology - the nanomembranes. A nanomembrane may be defined as a nanostructure with a thickness under 100 nm and a high aspect ratio exceeding 1.000,000 with areas reaching several cm2. Within MultiPlat numerous novel nanomembranes structures were developed including complex ones: layer-by-layer carbon nanotubes membranes, polymer/nanotubes multilayered composite membranes, metal oxide fibers/polymer membranes.
- Experimental validation of proton conductivity in biomimetic structures
The proton conductivity in organized nanochannels, similar to the biological channels is positively confirmed in macromolecular selfassembly structures and confined proton conductive nanochannels. Proton conductivity was also confirmed in carbon nanotubes. As the CNTs are electrically conductive they finally were replaced by electrically isolating titanium dioxide nanotubes.
- Integration/incorporation of ion (proton) conductive nanochannels into nano-membranes
The nanochannels in form of water-filled titanium dioxide nanotubes, externally oriented sulfonic groups in methylpropane sulfonic acid AMPS polymers and sulfonic groups of perfluorooctyl acrylate (PFOA) polymer are successfully incorporated in nanomembranes.
-.Low proton resistivities
The proton resistivities of best membranes amounted to 0.03 ohm/cm at 80 degC - 100 degC, the value
60 % lower than for comparable perfluorocarbon-sulfonic acid ionomers. From the practical point of view i.e. applications in fuel cells, the total resistivity of the developed material is lower than for known materials, considering the thickness ratio.
- Self leaning of precious platinum catalyst
An innovative, efficient method of leaning of precious platinum catalyst on the activated L-by-L nanomembranes was developed, tested and published.
- Novel nanomembrane characterization techniques
Characterization techniques were developed and published in Journals.
- Proof of concept for proton conductive biomimetic nanomembranes at macro level. minimal structure area 5 cm2.
The breadboard models are routinely manufactured with membranes of 25 cm2. Moreover, the novel fabrication method shows to be simple and efficient and hence applied to the patent.
- Functional model of a new class of fuel cells
The maximal operating temperature in our fuel cells was 95 - 1000C, The water flow was generally low, the effective thickness of nanomembrane before integration into the support was 1 - 5 ?µm.
- Proof of concept of the biomimetic nanofluidic transistor, exclusively operating on ionic fluids.
A functional biomimetic transistor based on ions could be demonstrated. Moreover an biomimetic array of fluidic diodes was shown. The parameters of those structures exceed the contemporary devices developed elsewhere.
- Breakthrough in the fundamental knowledge
A series of publications in recognized relevant journals evidences the contributions to fundamental knowledge.
1. Biomimetic proton conductive structure
Two concepts of biomimetic membranes were planned within the Project, stated in the Project Proposal. The first one was membrane fabrication by a sol-gel process and the second one was preparation of asymmetric membranes by delayed onset of liquid-liquid demixing using the process of immersion preparation (with and without electrical filed).
1.1 Sol-gel L3 sponge structure
First concept was based on the method of preparing a flexible nanoporous membrane starting from Sol-Gel L3 sponge phases. The necessary flexibility of the membrane was achieved using polyethylene glycol based precursors. It has been confirmed by experiments (AFM and SEM) that the pore diameter can be easily tuned within the 1-100 nm scale by adjusting the ratio of the components.
Poly (sulfone) (Mw 35,000, Mn 16,000) (PSf) and poly (ether imide) (melt index 9 g/10 min at 337o C/6,6 kg) (PEI) were obtained from Sigma-Aldrich. 2-Acrylamido-2-methylpropane sulfonic acid, 99 % purity (AMPS) and poly (ethylene glycol) 700 diacrylate (PEG 700 DA) were obtained from Sigma-Aldrich. Irgacure 2959 was obtained from Ciba. Poly (ethylene glycol) 400 (PEG 400) and n-methyl-pyrrolidone, 99 % purity (NMP) were obtained from Fluka and Sigma-Aldrich, respectively. In all experiments deionized water was used in the second gelation bath to prepare nanomembranes.
Concentrated solutions of PEI (27 wt %) and PSf (30 wt %) in NMP were prepared by mixing on a hot plate overnight. Solutions of AMPS in NMP were prepared fresh by mixing solutions in vials cooled in ice. Specified quantities of AMPS solutions were mixed with the PEI and PSf solutions, respectively. PEG 700 DA and Irgacure 2959 were added to these solutions at 10 mol % and 1 wt % based on AMPS, respectively. Polymer concentration is expressed as weight percentage of polymer per 100 g of polymer and solvent (NMP). AMPS concentration is expressed as mol of AMPS per L of solvent (NMP).
Prepared solutions were casted on a glass plate using a film applicator with a 120 mm gap (Erichsen Quadruple Applicator, Model 360), immersed in a PEG 400 bath, and quickly transferred into a high intensity (600W) UV curing unit (60 sec exposure at 90 % intensity). After UV curing, nanomembrane samples were put into deionized water to complete exchange of solvent and non-solvent and stored in a vial with deionized water overnight to complete extraction of unreacted residuals.
Two main parameters characterize performances of a membrane as the proton conductivity (for application in fuel cells): the proton conductivity alone and the ion-exchange capacity.
The first parameter, proton conductivity was measured as "in plane" of the nanomembrane samples using a 4-point method. The direct current was measured between two gold plated inner electrodes placed 1 cm apart, while keeping the potential difference between two gold plated outer electrodes constant at 100 mV. Under these conditions no significant polarization of the electrodes is expected. Nanomembrane samples, after the extraction of reaction residues for 24 hours in deionized water, were equilibrated in deionized water prior to the measurements. The electrodes were placed onto the skin side of the wet samples. Measured ohmic resistance and nanomembrane thickness (Mitutoyo contact gauge) were then converted into electrical conductivity, which should be equal to the proton conductivity under the measuring conditions.
The ion-exchange capacity (IEC) was determined by titration. Each membrane was placed in 20 ml of 0,05M sodium chloride aqueous solution for 24 h to exchange the protons with sodium ions. The ion-exchanged (hydrogen chloride) solution was titrated to pH 7.0 with 0,01M sodium hydroxide aqueous solution and the end-point was detected using a phenolphthalein indicator.
Values for conductivity obtained from our structures was in the range of 0.0290 - 0.0132 (S/sm), and for the IEC = 0.23 - 0.60 (mEq/g). When compared to standard material, the area specific resistance of nanomembrane samples is favorable to that of the commercial material. It should be emphasized that the area specific resistance of asymmetric nanomembranes includes the resistance of the microporous support of phase separated polymer and polyAMPS; therefore, the area specific resistance of the thin skin layer with biomimetic proton conductive channels is even smaller than measured.
1.2 Delayed onset of liquid-liquid demixing
The second method was coined as "Delayed onset" as such applied for the Austrian patent "Asymmetric interpenetrating proton-conducting membrane".
Briefly described, the method consists of multilayer structure; i.e. of gels and liquids carefully casted on a support. As the layers are mixable, they start to mix, primarily by diffusion. The second important fact is that the layer constituents are photosensitive. In the certain moment, the layer is exposed to the UV radiation and the membrane is polymerized in seconds. In this manner the instantaneous distribution of components are permanently "frozen".
The method is very general. It is possible to select a numerous and different layers, for practically unlimited species separation, as the cross-section of the "frozen" membrane can be symmetrical/unsymmetrical or graduated with multiple gradient.
It has been shown that the method provide proton conductive membranes with superior performances, if compared with other methods examined within the Project. As the method confirms as a novel, simple in manufacturing and enabling a wide range of products for fuel separation, beside the primary application in fuel cells, the method was applied for patent.
The method is based on preparation of asymmetric membranes with an interpenetrating proton-conducting morphology, which consists of cross-linked sulfonic acid functionalized ionomers embedded within a matrix of a thermally-resistant, glassy polymer is presented. This method combines a traditional immersion precipitation process for making membranes with photopolymerization and crosslinking of functional monomers included in the casting solution. The resulting membranes have an integral top skin layer with fine proton-conducting channels on top of a coarser proton-conducting support. In-plane conductivities of some of these membranes measured at ambient temperature were significantly higher than the conductivity of Nafion membranes, while having improved methanol barrier properties. An increase in functionality and molecular weight of crosslinking agents, as well as a selection of materials to promote delayed, rather than instantaneous liquid-liquid demixing were associated with significant improvements in membrane conductivity.
Asymmetric membranes with interpenetrating proton-conducting morphology could be of interest as membranes for direct methanol fuel cells
The second sub-method in manufacturing of biomimetic proton conductive membranes was use of fluorinated polymer nanomembranes with aligned electrical field. The new and innovative process to prepare these materials combines self-assembly of molecules with crosslinking in an electric field. Self assembly, as in biological membranes, is governed by mutual affinity between different chemical groups present in a solution. This self-assembly process create the aligned nanochannels via chaining of the nanodomains.
The preparation of nanomembranes with aligned pores implied two step processes and has for the first time been performed in this Project. Building blocks based on fluorinated monomers and monomers containing sulfonic groups were selected as they are known for their good chemical stability and proton conductivity. Since sulfonic groups can be considered as very polar and fluorinated moieties being very hydrophobic, phase separation and formation of e.g. tubular pores might occur. Furthermore, this difference in dipole moments is expected to allow for orientation of the two phases in the electrical field.
Fluorinated polymer nanomembrane with aligned pores were prepared using a two-step process from 1H,1H,2H,2H-perfluorooctyl acrylate (PFOA), 2-acrylamido-2-methylpropane sulfonic acid (AMPS), glycidyl methacrylate (GMA), photoinitiator Darocur 1173 and N,N-dimethylacetamid (DMAc) as solvent. The components were mixed in an ultrasonic bath. After purging the samples with nitrogen, they were irradiated in an UV chamber. Irradiation inducing radical copolymerization of acrylic groups was carried out to get a formulation with suficient viscosity for coating. The precured formulation was coated on a glass slide, an electrical field of 1.65 kV was applied for pore alignment and the temperature was increased to 150 ?°C to start cationic crosslinking of epoxy groups giving a solid membrane with aligned pores. Afterwards the sample was put in deionized water and electrical conductivity was measured. Subsequently, an electrical field is applied to orient the conductive polymer molecules in the thickness direction. The orientation is fixed by simultaneous crosslinking the membrane at elevated temperature.
The achieved data showed some important facts: the higher concentration of AMPS lead to much higher conductivity. Samples with high AMPS content prepared from the same formulation but with and without applying an electric field show a great difference in conductivity. Therefore we were able to show for the first time that conducting channels out of fluorinated AMPS hybrid formulations are successfully oriented in the electric field.
Final work with respect to the composition and preparation process parameters should further increase proton conductivity. Based on that data a patent should be filed within the next few months.
2. Bottom - Up fabrication of proton nanowires
Up to now we considered the proton conductive structure which in some extent emulates the existing materials, but actually has a different internal structure, more prone to biological cell membranes. The basic difference between existing materials and the developed membrane is the ordered, biomimetic proton-conductive structure of nanochannels. This membranes in not so far future promise to take a dominant position in mid and large scale fuel cells. However, witnessed with a rapid progress in miniaturized devices like smart communication devices, nomadic portable application and similar. Development of a miniature power supplies, more compact, independent from the grid and more efficient than existing ones are anticipated.
Two candidates were characterized: alumina and Si nanopores membranes. Two approaches have been tested for surface functionalization to facilitate proton transport through the membrane: Nafion filling and surface grafting of molecular monolayers. We have discussed confront critical limitations of the nanomembranes for fuel cell application.
The proton exchange membrane is one of the key issues in high performance fuel cells. The basic function of the membrane is to enable proton transport, while being simultaneously impermeable for electron and gas. Up to date, perfluorocarbon-sulfonic acid ionomer membranes are still the best known materials of this class. The main drawback is that the formation of proton conducting channels is highly dependent on humidity (water absorption) and the material therefore has limited working conditions within certain humidity and temperature range. In order to establish structured nanochannels for ion transport, a very recent progress adopted controlled nanoporous Si structures (with finite pore size: 5-7 nm-diameter) for an inorganic-organic proton exchange membrane. This results in a promising proton conductivity that is orders of magnitude higher than that of known materials at low humidity. However, the total film is tens of micrometers thick and the pores are random and branching. We design in our work straight nanopores with nanometer thickness, aiming at high throughput and efficient proton transport structures.
Clearly, within the MultiPlat project the brunt of work belongs to the miniature fuel cells. The well known scaling law intermits at some extent, and the consortium was forced to develop new ultrathin structures for new applications.
The thickness of contemporary membranes is in the range of 100 - 200 ?µm, the dimension not suitable for the devices of few centimeters large. The thickness/area ratio of the contemporary membranes is simply to low. Therefore, a large portion of project was committed to the fundamentally different structures. The basic idea behind was to form the proton-conductive structures strongly adhered to a structure which is inherently straight - the nanotubes. The starting idea was to use carbon nanotubes (CNTs) as a mast. In the time of the Project submission it was a known problem of discrimination of the conductive from non-conductive CNTs (comment: the rapid progress in nanoscience was considered). Clearly only the non-conductive CNTs could be used as a nano-mast.
The problem of CNTs separation on conductive/non-conductive was widely anticipated as one-step before solved and therefore the consortium proposed application of CNTs in the novel membranes. However, no significant progress in CNTs separation from the groups engaged in this work elsewhere (not Mutiplated related) came during the past 4 years. Hence, a redundant strategy for the nano-mast was selected: application of the titanium dioxide nanotubes (TDNs). TDNs are inherently nonconductive material, and therefore can be used as the supporting structure without restriction. Further, the TDNs have hydrophilic structure, which provides strong adhesion of the molecules carrying sulfonic group as the proton conductors.
On the other hand all mentioned considerations were done based on the theoretical work, which was also a fundamental part of the project. Therefore, it was necessary to examine both structures, with carbon and titanium nano-tubes and to compare the results. At first work and results on membranes with oriented CNTs are presented.
Commercially available CNTs were not quite suitable for the application. The form of CNTs should be vertically oriented, or in a carpet form. Hence, custom oriented CNTs were fabricated in EPFL. CNTs were synthesized by catalytic chemical vapor deposition (CCVD) technique from acetylene over a Fe-Co catalyst supported by CaCO3. The supported catalyst powder was placed into a 50 mm diameter CVD reactor. The synthesis was carried out at 640 ?°C in the presence of acetylene and nitrogen. Raw CNTs were purified during 1 day stirring in 1 M hydrochloric acid, washed and filtered with distilled water and freeze dried overnight. The average diameter and length of the CNTs are 11 nm and 10 mm, respectively. As described in literature, the realized carpet was mix of nproton/electron conductivity and therefore was unsuitable for the electrolyte applications. However, the structure filled with a proton conductive resin had enormous potential for the fuel cells electrodes, due to aligned fibers. This issue is presented later in the text.
The protonated TNDs were prepared by a two-step hydrothermal process. In a typical synthesis, 1 g of Titanium (IV) oxide nanopowder (99.7 % anatase, Aldrich) is mixed with 30 ml of 10 M NaOH (97 % Aldrich) solution. The mixture is then transferred to a Teflon-lined stainless steel autoclave (Parr Instrument Company) and heated to 130 ?°C and kept at this temperature for 72 hours. After the treatment, the autoclave is cooled down to room temperature at a rate of 1 degC/min. The obtained Na2Ti3O7 product is then filtered and washed several times with deionized water and neutralized up to pH 6.5 with the appropriate amount of 0.1 M HCl solution (Merck). During this step, sodium exchange proceeds to the formation of Na2Ti3O7 nanotubes. The sample is finally washed with hot (80?°C) deionized water in order to remove any traces of NaCl.
After the fabrication, the next step was TNCs functionalization. The TNCs was successfully functionalized with molecules containing a sulfonic acid at their end. Two different routes to the final functionalised product were evaluated: the first route was to synthesize 2-((2-phosphonoethyl)amino)ethanesulfonic acid first and then functionalize the TiO2 NT with. The second route is to functionalize TiO2 NT with vinylphosphonic acid first and then couple taurine via Michael addition reaction to the functionalized surface. For both methods it has been confirmed by FTIR spectroscopy that the TiO2-NT were functionalised. The second route is to prefer because unreacted can be recovered.
The second pathway was application of nanoporous silicon membranes (pore diameter 30 -60 nm, thickness 200-250 nm). It has been shown that polymerization of proton conductive monomers inside silicon membranes with pores in the micrometer range is possible. However, these nanomembranes were very fragile and first attempts to functionalize them resulted in a collapse of the membrane in the solvent. We solved the problem by deposition of the solvent from the vapor phase into the membrane as described in more detail in the experimental section. The surface reaction was done according to the known procedure. The silicon surface was first oxidized to increase the number of reactive hydroxyl groups. The membrane was then put into an acidic aqueous alcohol solution of the alkoxy silane for 30 min, rinsed with ethanol and cured overnight at 110?°C.
In contrast to previous studies we have chosen phosphonic acid groups for functionalization because they have higher hydrolytic stability than silanes. Similar as TiO2 in earlier studies, our functionalized TiO2 NT can increase water uptake, conductivity and mechanical strength and decrease fuel crossover. Furthermore, TDN can be oriented in an electric field to provide anisotropic behaviour. Using silicon membranes with nanopores instead of micropores increases the surface area for functionalization and improves the barrier properties for larger fuel molecules like methanol.
Next the work on supporting structures for the miniature fuel cells is presented. In reality, during the course of the project, this work flew in parallel with work on the proton conductive structures. In the standard size fuel cells, the proton conductive membrane has a double function: to separate protons from water flow and to provide the necessary mechanical stability. In contrary, in the miniature cells, those functions are separated: the proton conductive structure should enable a maximum conductivity, and the support and stability is deliberated to a separate structure - in our case to arrays of parallel nanochannels within an MEMS membrane. Three materials were processed, accordingly to knowledge base of partners within the Project: ultrathin monocrystalline silicon nanomembranes and ultrathin (50 nm) nanomembrane from ceramics (silicon nitride and aluminum dioxide or alumina).
The silicon nitride supporting membranes were fabricated as follow: first, a thin (100-500 nm thick) low stress silicon nitride (LS SiN) layer is deposited by low pressure chemical vapour deposition on a silicon wafer. Then, electro-beam lithography (EBL) is used to define the nanohole arrays in commercially available ZEP resist. The EBL pattern is then transferred to the silicon nitride layer by using carbon-fluorine based reactive ion etching. Finally, the LS SiN membranes are released by a partial deep reactive ion etching (DRIE) of the silicon followed by KOH wet etching for the final membrane release. KOH membrane release is preferred over dry etching since LS SiN is highly inert to KOH compared to dry Si etching, preventing possible damage to the nanopatterned membranes. Using this approach, membranes with arrays of 50-200 nm wide nanoholes and 50-300 nm interhole spacing have been fabricated on 100 ??m size LS SiN membranes.
In the later phase of work we were more directed toward low cost, self-organized processes. Actually this work belongs to the biomimetic approach. The first biomimetic work was manufacturing of alumina nanochannels.
The alumina nanochannels are fabricated by the anodization of aluminum in an electrochemical cell. No mask is required for anodization process (low cost process).
Up to date, much research has been done on high aspect ratio nanopores by using a pure thick aluminum film or plate. Anodization in proper conditions allows the creation of a self-ordered array of cylindrical pores with diameters ranging from sub 10 nm to 150 nm and depths exceeding 100 ??m. The pore dimensions and periodicity are controlled by the electrolyte composition, temperature, and current-voltage settings. Typically, the aluminum anodization process results in porous alumina (Anodized Aluminum Oxide AAO) with a volume expansion are about 1.5 and porosity of about 10%. Here we combine anodization and micromachining techniques to fabricate nanoporous thin alumina membranes. Compared to commercial bulk AAO membranes (thickness > 10 ??m), they have the advantage of providing low fluid flow resistance and a more flexible membrane geometrical configuration for the studies of ion transport. The fabrication process was adopted as follow: 3 ??m photoresist is spin-coated on to the back of a 4 wafer with double side 500 nm LS SiN. UV photolithography is used to pattern the back of the wafer and to release the LS SiN membranes by Si dry- and wet-etching. Then, a thin Al film (in hundreds of nms) is deposited on the wafer front side. Anodization is performed in 3% (v/v) H2SO4 solution by applying 20 V potential between the Al film and a Pt mesh. This creates a self-ordered array of cylindrical alumina nanopores on the full wafer surface. Alumina nanopore membranes are obtained by the final removal of SiN support layer. The side length of the membrane is determined by the backside mask opening, which is typically 200 - 400 ??m. The membrane thickness can be tuned in a straight-forward way from hundreds of nm to ??m, depending on initial thickness of Al film and final plasma etching of alumina.
Finally experiments were done with porous silicon. As the silicon offers better mechanical and chemical stability than alumina, silicon was choosen as material for the nanochannel fabrication. In addition, the silicon membranes can have well-defined perpendicular channels in the nanometre range offering unique possibilities for selective membranes. By surface functionalization with active molecules they can be modified for different applications. For example, grafting the sulfonic groups would lead to proton conducting membranes. The surface was modified with alkoxysilanes using standard procedure and confirmed the reaction by contact angle measurements. This procedure can be adapted to add different functional groups for other applications.
3. Breadboard model, test of standard, mini and micro fuel cells
The final characterization of the components in the conventional cell assembly has been carried out according to the guidelines put forward by the EC funded FP5 thematic network project FCTESTNET.
The experimental set-up for fuel cell tests and the electrochemical techniques used was then described in detail. Data obtained for commercially available membranes based on perfluorocarbon-sulfonic acid ionomers was used as reference toward the new MultiPlat membranes.
Catalyst Layer: The catalyst inks are made with carbon supported platinum catalyst powders supplied by Tanaka, a 5% perfluorocarbon-sulfonic acid ionomer emulsion, millipore water and ethanol. An ultrasonic pulverisation system was used to spray the catalyst ink onto the membrane (CCM) or the gas diffusion layer (GDE) support. The support that was to be coated was placed on the system vacuum table which was heated to 80?°C. The pulverisation system software allows the operator to control the three-axis motion and the speed of the spraying nozzle, the shape of the spraying pattern and the catalyst ink flow rate. The platinum catalyst loading can thus be easily controlled by the parameters used for pulverisation; the standard platinum loading at PaxiTech is 0.5 mg Pt /cm?².
MEA Assembly using GDE: Since the mechanical and thermal resistances of the new MultiPlat membranes are not known precisely, preliminary tests shall be carried out using MEA where the GDE and the gaskets are simply placed on the membrane. If the new MultiPlat membranes show no sign of gas crossover or electric short circuiting during fuel cell tests, standard MEA assembly procedures shall then be used. Standard MEA assembly involves placing a GDE and a gasket on either side of the membrane and hot-pressing at 140?°C. Hot-pressing was carried out at 140?°C in two steps: no pressure for 3 minutes then 600N/cm?² for 5 minutes. It is highly possible that hot-pressing temperature and pressure conditions may need to be modified for the new MultiPlat membranes.
3.1 Fuel cell Tests
- Open Circuit Voltage Test (OCV): The test provide permeability of the reactant gases through the proton conducting membrane has a very important influence on fuel cell performances. Hydrogen crossover and the subsequent direct recombination of hydrogen and oxygen will result in a decrease in the open circuit voltage (OCV), reduced power density and fuel cell efficiency and an acceleration of membrane degradation and hence reduced MEA lifetime. A measure of the OCV of the MEA as a function of time is carried out at different cell temperatures.
- Electrical Resistance Measurement: Basically characterize the fuel cell toward long term degradation. As with crossover, excessive electrical conduction through the membrane results in degradation of fuel cell performances and reduced MEA lifetime. The ohmic resistance across the MEA is measured before and after fuel cell tests with a multimeter.
- Chronopotentiometry (I pulse): A constant current is applied to the cell and the potential measured as a function of time. The fuel cell is tested under different test conditions which are presented in the Table 1 below. The current applied depends on the size of the test cell and the membrane performances, starting at a minimum of 0.2 A /cm?² and increasing gradually.
3.2 Test results
At total six asymmetric proton conductive nanomembranes samples prepared using a sol-gel process by TUW laboratories were tested in H2 /Air fuel cell. The open circuit voltages obtained for all of these membranes indicate that gas permeation is occurring across the membrane. The fuel cell performances were lower compared to reference data obtained with a NafionG? PEM membrane. Three of the sample membranes could be operated up to cell temperatures of 50?°C resp. 60?°C. Reduction of gas permeation across these nanomembranes is recommended for further improvement of their potential as PEMFC membranes.
The three Nanocyl GDL materials were tested on the cathode side of a 25cm?² H2 /Air fuel cell. There were two types of GDL compositions; one consisted of incorporating the CNT within the carbon non-woven support material and the other of applying a CNT layer onto one side of the carbon non-woven support material. Better fuel cell performances were observed for the GDL with 17wt% CNT content than for the 2 GDL with a CNT layer for the first test. Data obtained for the two GDL with a CNT layer showed that water management needs further optimisation and mass transport overpotentials were observed as the CNT layer thickness increased. MEA flooding was observed for all three GDL samples within 24h of testing. The porosity and thickness of the CNT component of these CNT-GDL materials needs to be optimised so as to improve water removal mechanisms during fuel cell operation.
3.3 Lesson learned
The novel proton conductive membranes developed within the MultiPlat project offers better proton conductivity compared to perfluorocarbon-sulfonic acid ionomer in stand alone test, but lag somewhat behind the known materials in MEA test. This result is encouraging.
The future improvement of the MultiPlat material has to be directed in:
- more and detailed lifespan tests, out of scope of Multiplat Project
- better crossover toward methanol, if the target application is Direct methanol fuel cells. The future application of that fuel cell type is bright, as the methanol storage is more convinced than hydrogen.
- the internal structure of the membrane material is biomimetic (i.e. consist of oriented nanochannels as required within the proposal). Future work scould be directed toward better cannel orientation and particularly to efficient inter-cannel connections, which will provide reduce in the total membrane resistance
4. Advanced devices, fluidic diode, fluidic transistor and fluidic electronic in general
From the practical point of view, transport of hydrogen ion in narrow channels, particularly externally controlled transport have a profound importance in understanding of the vital biological processes, but also for various man-made devices, such as sensors, batteries beside the described application in fuel cells. Protons are very specific ions which generally exist in electrolytes only in the form of hydronium ions (H3O+) associated with clusters of water molecules. Owing to the specific "hopping mechanism" of transport, the protons exhibit abnormally high mobility compared to the related cations as Li+, Na+ or K+. Furthermore, the proton transport in uninterrupted 1-D water chains (water nanowire) is ensured through fast transport of charge defects rather than through the transport of protons themselves. The accurate mechanism of proton transport in water remains not fully understood, despite decades of research. This was the motive for including this work in the MultiPlat project.
Two methodologies are available and applied here for the experimental investigation of proton transport in narrow channels: the examination of proton transport though macromolecules containing a well defined nanochannel filled with water or through the artificial nanochannel which enables us controlled modification of principal device parameters within a single device. For the first method it was necessary to synthesize different organic compounds which contain nanochannels with fixed parameters (i.e. diameter, surface states, etc.). The second approach implies fabrication of an artificial device with integrate nanochannels, fluid reservoirs and control structure - the gate. In this case, the effective cross section of a nanochannel or nanochannels array can be varied solely by variation of an external electric field.
Two groups worked in parallel on research of the ions transport within nanochannels, primarily protons - TU Wien and EPFL. Both groups share theoretical part (modeling) of the ion transport. The idea behind research on this field is based on many common physical effects with electrons in solid state materials. Ions and electrons reassemble each other in many ways. Both can be calculated using Drude model. Moreover, both flow by diffusion and drift mechanism. Additionally, the thermal generation of electrons and holes in semiconductors is analogous to the thermal dissociation of water molecules. As usual more details about this interesting topic is presented in the reports and published papers.
The experimental work was on the fluidic diodes was done at EPFL and on the fluidic transistor at TU Wien.
(Nano)fluidic diode: The basic structure for fluidic diode were hetero-structured nanopore membranes, which strongly emulate the biomimetic directional ion conduction channels. The membrane is composed of Al2O3 and SiO2(Si) layers with dense nanopore arrays. On exposure to a pH-neutral solution, Al2O3 surfaces possess net positive charges, while native dioxide layers of Si possess net negative charges. Due to the asymmetric surface charges, rectification in ionic current can be observed through the nanopore membrane, behaving as arrays of nanofluidic diodes at relatively high electrolyte concentrations.
Thanks to the opposite surface charge states of Al2O3 (positive) and SiO2 (negative), the membrane exhibits clear rectification of ion current in electrolyte solutions with very low aspect ratios compared to previous approaches. Furthermore, the Al2O3/W combination allows the application of an electrical field to further tune the ionic transport through the nanopores with low gate potentials and ultra low gate leakage current. The hetero-structured nanopore arrays provide a valuable platform for high throughput applications such as molecular separation, chemical processors and energy conversion.
The rectification factor is 6 at 1 mM KCl solution. By substituting the Si with a W layer following thermal oxidation to grow WOx dielectric, electrical potential can be applied to further modulate the ion transport across the membrane. The rectification factor can be tuned from 2 to 11. Although the finite rectification ratios for these devices deviate from the ideal diode characteristics, the development of such heterostructured nanopore membranes would enable various high throughput applications in nanofluidic chemical processors and molecular separations.
The nanofluidic diode behavior of the Al2O3/SiO2(Si) membrane is demonstrated in KCl electrolyte at varied concentrations ranging from 0.1 mM to 2 M. The membrane is inserted between two reservoirs with the same ionic concentration and liquid pressure. A pair of Ag/AgCl electrodes is used to apply a potential through the membrane. The electrodes are fabricated by the anodization of Ag wires in HCl solution. For our liquid cell configuration, low porosity samples (i.e. high resistance) are preferred in order to minimize the contribution from cell geometries to ionic current through the membrane.
The method was quite innovative, since the most existing experiments rely on planar geometry where the nanochannels are generally very long and shallow with large aspect ratios. The nanometer limiting dimension, which is normally the channel height, is defined by the thickness of a sacrificial silicon layer while the channel width and length vary from micrometers to millimeters. Based on the "etching followed by bonding" scheme, various types of experiments have been carried out to demonstrate the principle of ionic and molecular transport in one or several nanochannels. However, device minimization and throughput scaling remain significant challenges.
(Nano)fluidic transistor. The transistor, including here developed fluidic MOS field effect transistors are more complex device than diode, but also more versatile. If we use previously described analogy between semiconductors and fluids, the flow of ions can be controlled by an external electrical field, the same mechanism behind semiconductor MOS transistor. Similar to semiconductor MOS equivalent, the fluidic transistor basically consists from (nano)channel, contact electrodes and the gate.
The nanochannels and contact electrodes (water reservoirs) were manufactured in Pyrex wafers. The position and the size of nanochannels was defined using standard photolithography. Nanochannels were plasma etched 4 ?µm wide and 155 nm deep. We made various test structures with different nanochannels lengths (10 ?µm, 60 ?µm and 460 ?µm). Also the number of channels per chip was varied (1, 5, or 10) in order to have larger variety of different test samples for the following up characterization measurements. After processing of the nanochannels, the electrolyte reservoirs were wet etched to a depth of about 18,4 ?µm in highly concentrated HF solution (40%) by using a Cr/Au masking layer (50 nm/400 nm). Reservoirs were made large enough to hold sufficient quantities of fluid during characterization. In this way, the impact of fluid evaporation on measurement results was significantly reduced. Additionally, larger dimensions of the reservoirs enabled good electrical contact to the electrolyte. The same hard mask based on the Cr/Au bi-layer used for the etching of the glass substrate in HF was also used to pattern via a lift-off process the Ta/Pt (10 nm/200 nm) electrodes. The processed glass wafer was subsequently cleaned and cat into the chips with size 10 x10 mm.
The transistor gate was made of dry oxidized silicon. The starting material was a silicon wafer (thickness 350?µm) coated with 250 nm thermal SiO2 and 70 nm of LPCVD Si3N4 The standard processing was used for the deep wet etching of Si in 40% KOH. Next, the hard mask was completely removed from both wafer sides. The wafers were once again plasma cleaned, followed by dry oxidation process for growing of the high quality gate oxide layer (100 nm SiO2).
As mentioned previously, the test structure reassembles the structure of the MOS transistor, where the flow of electrons in the semiconductor channel is replaced with flow of protons/ions in the nanochannel under the applied potential. This fact inspires the authors to use the standard semiconductor parameter analyzer for testing of the device. Prior to any measurement, the device was cleaned with methanol, and then the left reservoir was filled with water. After the delay of 15 min, the time period enough long for capillary diffusion of water from reservoir thought entire nanochannels length, the opposite reservoir was filled with water. Thanks to this procedure, there is no need for application of an external pressure source for filling the nanochannels.
The first measurement sets were verification of the fundamental theoretical consideration of the nanofuidic device - that the current flow within the nanochannel is carried solely by ions flow and not generated inside the device within some electrochemical reaction. A few preclusions to prevent this effect were carried out from the begging of the design process. The S and D electrodes were made symmetrical and from inert Pt metal, next the projected maximal operating voltage of the device was chosen < 1V, below of the potential of the water electrolysis which is -1.23 V at 25 ?°C. It should be considered first, that the nanofluidic transistor is a relatively new device, with few examples previously described. Next, described devices solely relay on the water as fluid with various ions concentration. Pure water is only slightly conducting due the self-ionization of water as
2 H2O H3O+ + OH at app. dissociation rate of 1 event per molecule in every 10 h. As consequence, pure water is very slight electrical conductive with conductivity level of S ~ 0.055 ?µS-cm1 at 25 ?°C. In the same time the water is the best solvent known and even a minute concentration of contaminants within the nanochannel notably increases the conductivity of water. For example, contamination water with NaCl salt at level of 10 - 7 increase the water conductivity for 2 orders of magnitude.
In the next measurement set(s), the nanofluidic devices were connected in the same manner as an ordinary MOS transistor on the semiconductor parameter analyzer. Numerous measurements were done on devices with various combinations of nanochannels length and number. As expected, the signals obtained from devices with single, 460 nm long nanochannels were generally noisy and substantially unstable. However, the plots from devices with nanochannels arrays and particularly with shorter nanochannels were surprisingly stable and reproducible. The fundamental conditions for coexistent results were: the water quality, sufficient delay between the sample reparation and measurement in the aim to reach equilibrium in the ion concentration within the nanochannel and prevention of fluid evaporation.
The transistor transconduntance gm=2?Iout/?Vin were in average in the range of 4 - 8 x10-6 S, a significantly lower value compared to the contemporary 45 nm MOS transistors with gm in the range of 1-30 mS. The transconductance value obtained from early devices may seems to be modest, however the biological nanofluidic transistors, either based on the same principle, have 1-2 order of magnitude higher transconductance than contemporary Si devices.
As conclusion it can be stressed that the results of the experiments confirmed that the confinement of the effective cross-section of nanochannels via external electrical field can modulate the proton/ion flow in the nanochannel. The novel device worked at a low gate voltage of < 1 V and had a higher transconductance (i.e. by a factor of 10) than those reported in the previous publications, due to a thin gate insulation. Also important is the fact that the experimental results are in good agreement with theoretical predictions.
Implementation and evolution of EU policies
EU Fuel Cells and Hydrogen Joint Technology Initiative analyzed the current situation in the R&D efforts in Europe. :
Current situation of hydrogen and fuel cells (FC/H2) research in the EU is:
- The research needed is often so complex that no single fuel cell company or public research institution can perform it alone
- National funds available for FC/H2 are dispersed
- There are still many technical and non technical barriers to overcome before widespread commercial availability is possible.
- Strong R&D competition is coming from other global players (not only US/Japan but increasingly China)
- There is no agreed long-term budget plan and strategic technical and market objectives to encourage industry to commit more of their own resources;
- There is insufficient integration of the EU R&D programme (from fundamental research through to large-scale EU-level demonstrations);
- Technical breakthroughs are needed to improve performance and durability and reduce system costs to meet the expectations of potential customers. (end of citation)
The project was relevant to several other European documents, including EC Energy Package (23 January 2008). MultiPlat is in line with the Fuel cell technology platform and its strategic research agenda.
Improvement of European social and economic cohesion
The strengthening of economic and social cohesion is one of the three basic objectives of the European Union, along with Economic and Monetary Union and the completion of the Single Market. The course of action to reach these objectives, set out at Lisbon and Stockholm, includes an economy based on knowledge, competitiveness, innovation, investment in people, the fight against social exclusion, full employment and the search for sustainable economic growth.
Energy is at the very foundation of a peaceful and prosperous society. Thus the MultiPlat project directly contributed to European and economic cohesion.
Quality of life
Through its research results, MultiPlat has direct impact on the quality of our lives. The applicability of fuel cells in vehicular technology is well known and well appreciated. Besides curbing CO2 emission, which is vital to abating the greenhouse effect, they improve the quality of life in cities by powering electric motors, thereby reducing noise, and by eliminating local pollution caused by nitrogen oxide (NOx), particulates and other emissions.
Other applications influencing the quality of everyday life include mobile devices like cell phones, GPS, portable computers, etc., where fuel cells will bring much longer device autonomy and better performance. In addition to that, the replacement of traditional batteries with a product with longer lifetime can help decrease the amount of toxic waste. Portable medical devices are another highly important field where quality of life can benefit from FCs, since they ensure significantly better and more dignified life to many people with serious and sometimes disabling health problems.
Employment
The results of such a strongly applicative project as MultiPlat bear an intensive impact to the employment through opening new workplaces and offering new opportunities. The most direct impact is through the production of FCs for various applications, but most important ones are in automotive propulsions and mobile devices. Secondary influence is much wider, through industries utilizing these devices, while its indirect impact covers an even larger segment of workforce. Here we consider only a single application of the MultiPlat that of fuel cells in automotive industry.
European automotive industry directly employs about 2.000.000 workers, while the secondary workforce is about ten times larger. The forecasts predict production of 20 millions cars until 2020, and about 1-5 millions of these can be expected to be fuel cell-powered.
Thus a large portion of the automotive industry will be connected with the production of FCs, worldwide and in the EU. Therefore it is not the question if the European workforce will be engaged in the FCs production, but whose licence and know-how will be applied. MultiPlat contributed to the efforts that the novel generation of the fuel cells emerges as a European technology.
Environment
The next twenty years will be critical, both for climate change and for security of supply. In this period, irreversible and catastrophic damage may be done to the climate.
A fuel cell running on pure hydrogen is a zero-emission power source. Some stationary fuel cells use natural gas or hydrocarbons as a hydrogen feedstock, but even those produce far less emissions than conventional power plants.
A direct application of the fuel cell is the transportation. Fuel cell vehicles are the least polluting of all vehicles that consume fuel directly.
- Fuel cell vehicles operating on hydrogen stored on-board the vehicles produce zero pollution in the conventional sense. Neither conventional pollutants nor green house gases are emitted. The only byproducts are water and heat.
- The simple reaction that takes place inside the fuel cell is highly efficient. Even if the hydrogen is produced from fossil fuels, fuel cell vehicles can reduce emissions of carbon dioxide, a global warming concern, by more than half.
- Fuel cells used as auxiliary power units (APUs) to power air conditioners and accessories in over-the-road trucks could reduce emissions by up to 45% from long haul vehicles, and deliver economic benefits to the truck owner in lower fuel use and less wear and tear.
Over this period also, the costs of dependence on very few large producers of energy might also become painfully apparent. Europe can meet these challenges only through the application of technology and the adoption of difficult changes to social values and behaviour. Recent studies suggest that a decentralized energy system (also known as a "distributed" system) would improve overall efficiency through: 1.) reduced losses over power lines since electricity would no longer need to travel hundreds of miles from where it is produced to where it is used, and 2.) reduced investments in the transmission and distribution infrastructure. Furthermore, technologies, such as fuel cells, suited for distributed applications, generally produce less harmful pollution than the large central-station power plants used today. Because fuel cells convert the fuel to electricity through an electrochemical process rather than a combustion process typical of most power plants, the emissions are much cleaner. Compared to burning fossil fuels like coal and oil, which produces emissions of sulfur dioxide, nitrogen oxide, and carbon dioxide, the electrochemical process used in fuel cells only has carbon dioxide and water as byproducts. The low emissions from fuel cells make them an environmentally preferred form of power production. However, it should be emphasized that the first generation of fuel cells will likely operate on natural gas or propane, which are finite fossil fuels whose extraction from the ground and delivery produce negative environmental impacts. In the future, fuel cells will run on gas derived from biomass (plant matter) or pure hydrogen extracted from water using wind or solar energy, thus playing a key role in ushering in a sustainable energy future.
Exploitation plan
After 36 months of MultiPlat's duration 12 exploitation results have been identified and agreed within the consortium:
Production process of nanomembranes as supports for biomimetic proton-conductive structures : new process to make ion selective membranes
- Flexible Sol Gel-based Hybrid and Fluorinated Polymer nanomembranes
- Surface modification with proton conducting macromolecules
- Fabrication process of vertical nanochannel arrays
- Growth of NTs on substrates
- Functionalization of the NTs with nanoparticles
- Nanocomposites CNT/membranes
- Membrane electrode assembly (MEA)
- Portable fuel cells
- Training material (product)
- Monitoring of ion transport in fuel cell membranes
- Combined scanning probe (SPM) and phase contrast acoustic (PSAM) microscope
these results represent the most prominent fields of exploitation to be followed in the future (after the expiration of MultiPlat).
list of Websites:
website: http://www.multiplat.net
contact: Werner Brenner, Vienna University of Technology, Institute of Sensor and Actuator Systems, werner.brenner@tuwien.ac.at