Skip to main content

Development of a Portable Internal Reforming Methanol High Temperature PEM Fuel Cell System

Final Report Summary - IRMFC (Development of a Portable Internal Reforming Methanol High Temperature PEM Fuel Cell System)

Executive Summary:
The complexity of the balance of plant of a fuel cell-fuel processor unit challenges the design/development/demonstration of compact and user friendly fuel cell power systems for portable applications. An Internal Reforming Methanol Fuel Cell (IRMFC) stack poses a highly potential technological challenge for High Temperature Polymer Electrolyte Membrane Fuel Cells (HT-PEMFCs) in portable applications. It aims at opening new scientific and engineering prospects, which may allow easier market penetration of the fuel cells. The core of innovation of IRMFC is the incorporation of a methanol reforming catalyst in the anode compartment or in between the bipolar plates of a High Temperature Polymer Electrolyte Membrane Fuel Cell (HT-PEMFC). In order to obtain an economically technologically viable solution, low-cost materials with certain functional specifications within 200-220oC (electrolytes, catalysts and bipolar plates) and production techniques, with easy maintenance and high durability will be employed. Taking advantage of the innovative outcomes of the previous FCH-JU IRAFC 245202 project, the functionality of MeOH-fuelled integrated 100 W module was demonstrated. IRMFC partnership brought together specialists in catalysis (FORTH, UMCS, ZBT, IMM), HT polymer electrolytes (UPAT, ADVENT, FORTH), as well as the technological know-how to design, construct and test balance-of-plant components and HT-PEMFC stacks (IMM, ZBT, JRC-IET, ADVENT). Special role was adapted throughout the project for end-user (ARPEDON) with respect to regulations-standards-codes requirements for portable applications.
The main objective of the project was the development/demonstration of an Internal Reforming Methanol Fuel Cell at 100 W net power output for portable applications. This power device should be able to operate for 1000 h (including startup/shutdown cycles) on MeOH/H2O fuel at a current density of 0.2 A/cm2 at 210oC. Towards this end, the following goals were accomplished:
• Scale-up of the synthesis of the robust high temperature (210-220oC) polymer electrolyte membranes for HT-PEMFCs, developed within the framework of FCH-JU IRAFC 245202 project.
• Scale-up of the synthesis of highly active and stable methanol reforming catalysts for the production of CO-free hydrogen in the temperature range of HT-PEMFCs, i.e. at 210-220oC.
• Double arrangement of ultrathin methanol reformers integrated in the anode side of fuel cell, able to maintain the methanol conversion at values higher than 95% for more than 1000 h including on/off cycles.
• Integration of reforming catalyst and high temperature MEA in a compact Internal Reforming Methanol Fuel Cell aiming to specific and volumetric power densities of less than 35 kg/kW and 50 l/kW (fuel amount excluded).
• Design and construction of the BoP to support the fuel cell operation.
• Design, construction and successful testing of two types of IRMFC modules (up to 100 Wel output): graphite- and metal- bipolar plates based short and 100 W modules.
• Operation at 210oC and 0.2 A/cm2 with a voltage output of 19-20 V and power output of 90-100 Wel.
Specific targets for improvement of the efficiency have been identified. Heavy and voluminous compression plates and insulation casing can be further optimized and result in much lower specific weight and size (lower than 35 kg/kW and 50 lt/kW). Moreover, the perimeter area of the metallic BPPs can be significantly decreased and contribute in the decrease of weight and volume. Nevertheless, the most important deviation identified, had to do with the thermal and mechanical stability of the membrane for operation above 200ºC and especially under on/off cycles. The latter drawback of the polymer electrolyte membrane is currently under investigation by partner UPAT and promising materials have been already developed and soon will be tested under realistic conditions.
Project Context and Objectives:
The main objective of the project was the development/demonstration of 100 W internal reforming methanol high temperature PEM fuel cell system for portable applications. Towards this end, the following goals were accomplished:
• Scale-up of the synthesis of the robust high temperature (210-220oC) polymer electrolyte membranes for HT-PEMFCs, developed within the framework of FCH-JU IRAFC 245202 project.
• Scale-up of the synthesis of highly active and stable methanol reforming catalysts for the production of CO-free hydrogen in the temperature range of HT-PEMFCs, i.e. at 210-220oC. The reforming catalysts have been also developed within the FCH-JU IRAFC 245202 project.
• Integration of reforming catalyst and high temperature MEA in a compact Internal Reforming Methanol Fuel Cell aiming to certain power density and dimensional specifications. Integration was achieved via two different configurations (one based on new composite graphite bipolar paltes and the second one based on metal bipolar plates). Single and double reformer arrangement of ultrathin catalytic layers were employed.
• Design and construction of the BoP to support the fuel cell operation.

The major components of the integrated fuel cell/fuel processor system were:
• Methanol reformer for the production of hydrogen-rich gas at the operating temperature of the fuel cell (200-210oC) . It contained the most efficient Cu-based methanol reforming catalyst developed in the present project, taking into account the corresponding outcome of the previous IRAFC project. The catalysts was supported on a conductive carrier such as carbon paper and was incorporated in the anode side of the stack, separated from the anode electrocatalyst with an ultrathin carbon-based gasket.
• A high temperature membrane electrode assembly (HT-MEA), being able to operate at temperatures of 210-220oC. This was based on the high temperature cross-linked aromatic polyether with pyridine H3PO4-imbibed HT-MEAs developed in previous IRAFC project.
• Electronic balance of system unit (fuel feeding system, power and thermal management and control of operational parameters of the whole system).
There are a number of reasons which require the design and development of compact and friendly to the end user, fuel processor-fuel cell energy systems for portable applications. These reasons are briefly the following:
• Simplification of fuel cell-fuel processor unit via (i) high temperature (>200οC) operation of fuel cell and (ii) incorporation of fuel processor in the fuel cell stack. These are also related to additional advantages concerning steam, heat (the “waste” heat produced by the fuel cell is in-situ utilized to drive the endothermic reforming reaction) and air management and volume reduction.
• The increasing demands for high energy and high power sources in a variety of portable applications with respect to cost effective long term and safe operation at wide environmental temperatures and any orientation.
• Utilization of methanol as primary hydrogen carrier offers higher energy densities and overcomes technical obstacles associated with the use, storage and transportation of pure hydrogen, while it may be derived from renewable energy sources. In such a case, a full environmental benefit is achieved, because no CO2 emissions are added to the atmosphere, and fossil energy resources are preserved.
Aforementioned issues may be technologically approached and satisfied via the proposed IRMFC system of 100 W for various portable applications. The proposed system does away with conventional fuel processor-fuel cell systems in terms of (i) operation temperature, (ii) reformer utilization, (iii) heat management, (iv) cost-volume-weight effectiveness. The latter factor with respect to the corresponding power density makes IRMFC very attractive for emerging commercial/military portable applications.
The main objectives of the project with respect to the main research directions were the following (the current status is shown in parenthesis):
➢ High temperature polymer electrolyte membranes - electrodes – assemblies
A main prerequisite for the integrated system of an Internal Reforming Alcohol High Temperature PEM Fuel Cell Stack operating at temperatures higher than 210oC, is polymer electrolytes of exceptional mechanical, thermal and oxidative stability in combination with decreased requirements of H3PO4 doping levels in order to acquire proton conductivity values up to 10-1 S/cm. The IRMFC project will focus on the optimization of developed functionalized copolymers and their cross-linked counterparts for HT PEMFCs operating above 210oC and up to 220oC.
DONE √
➢ Methanol reforming catalysts
Reforming catalysts operating at 210-220oC with high stability, activity and selectivity towards hydrogen
Production. Important improvements of reforming catalysts are also required regarding H3PO4 tolerance and their ability to operate without preactivation steps. Special attention will be given on the scale-up synthesis.
DONE √
➢ Design, construction and testing of bipolar plates, short IRMFC modules, 100 W stacks
Development of high temperature stable (ca. 220°C) graphite compound based bipolar plates and ultrathin metallic plates, small IRMFC modules (containing one and five cells), and a IRMFC module delivering 100 Wel
DONE √
➢ Design and construction of BoP
The design and construction of all necessary BoP components such as heat exchangers, evaporators, pumps
and blowers so as to complete the component development for a fully automated and self-sustaining system. All developed components shall then be delivered to WP6 for system integration and testing.
DONE √
➢ System integration and testing
To completely integrate systems with most promising stack configurations; To achieve 1000 h of operation including start-up, shut-down and load changes
Comment: the low thermal and mechanical stability of polymer electrolyte membrane against on/off cycling conditions limited the efficiency of the final modules for long periods, while technical difficulties with the thermal efficiency of the BoP to support the operation of the stack up to 210oC, forced us to activate an alternative simulated BoP and demonstrate the promising performance of an 100 W IRMFC module containing 32 MEAs and 32 double reformers.
➢ Compliance to regulations, standards and codes
Synchronized codes and standards development and adoption with technology deployment and
commercialization needs.
Coordination of IRMFC with codes and standards.
Adoption of the latest developed codes and standards.
Methanol safety information.
Readily available RSC material for first responders and other key stakeholders.
Comment: The CE-Conformity Assessment Roadmap for the IRMFC system was developed and the necessary steps to perform a CE-certification were described as well as forms for most steps were developed. Some check fails were observed due to technical limitations and range of availability of the final system, but they were considered of minor importance compared to the magnitude of the project.
Project Results:
High temperature polymer electrolyte membranes - electrodes - assemblies
1st period
Copolymer synthesis
The development of crosslinkable polymers and membranes and their scale up preparation was the main task of this WP. Based on the results obtained in the previous IRAFC project and the work that was initially performed to finalize the crosslinking methodology that is more proper for large scale synthesis, we came out with the selection of the double bond side functionalized copolymers that are able to be crosslinked after the membrane formation and during the necessary treatment with the phosphoric acid. The first copolymer structure that was used for this reason were copolymers 1 having the pyridine units in the main chain that enable their interaction and uptake of the strong acids and the side double bond functionality that was introduced by the copolymerization of the diallyl-bisphenol monomer.
Despite the many attempts that were devoted to this synthesis and the fact that these copolymers with low molecular weights were effectively synthesized we faced problems in scaling up the copolymer preparation as well as to increase the molecular weights.
Thus terpolymers resembling the Advent TPS structure (copolymers 2) were selected by introducing the tetramethyl bisphenol monomer that helped in solubility of the final copolymer along with the crosslinkable diallyl-bisphenol monomer. For these copolymers 2 synthetic efforts were devoted for their optimization in respect to the copolymers composition and polymerization conditions. Besides the effective monomers copolymerization the optimization of the reaction conditions was necessary since a significant discrepancy between the monomers’ feed ratio and the final copolymer’s composition mainly in terms of the double bond content was noticed in several cases. This discrepancy was studied using different copolymers obtained under different polymerization conditions and their characterization was performed by combined FTIR and HNMR spectroscopies. Through the above exhaustive conditions and monomers ratio examination we found the optimum conditions in order to keep alive the double bond functionality that will enable the final crosslinking of the membranes.
Additionally and in order to explore alternative crosslinkable copolymers we explored copolymers bearing the 2,6 pyridine units in the main chain and the crosslinkable side double bonds (copolymers 3). Copolymers 3 were synthesized in different composition that in some cases showed high molecular weights while in all cases they present high phosphoric acid doping levels and excellent solubility.
Another alternative polymeric structure was also proposed and examined namely quinoline based homopolymers. These novel polymer electrolyte materials were synthesized using an AB monomer bearing both required functionalities for high temperature polycondansations in the same molecule, a hydroxyl group and a fluoro-phenyl moiety (Homopolymer 1).
Different polymerization conditions were used giving in some cases reasonably high molecular weight polymers. Also a solventless polymerization has been attempted giving promising results. The advantages that were introduced by this homopolymer are that we avoid the stoichiometry problems that in many cases leads to the synthesis of low MW copolymers by using the AB monomer and also due to the chemical structure and the presence of the quinoline groups this homopolymer shows high acid doping levels while keeping the membrane integrity because of the presence of the hydrophobic fluorinated part. Additionally, due to the robust chemical structure, high thermal stability was observed both for the doped and undoped membranes.
Polymer blends
Finally the blending concept was introduced in order to overcome the problems of high MW copolymers combined with high doping level and doped membrane stability. In this case we used a high MW polymer electrolyte like the Advent TPS that is obtained in Mn values, above 100.000 and up to 150.000 but shows limited acid doping uptake, with the new synthesized crosslinkable copolymers 2 and 3 or the new quinoline based homopolymer 1 since in these cases we have higher acid doping ability.
Especially in the case where crosslinkable copolymers were used in blends with TPS, we achieved to further stabilize the membranes during doping where the crossilinking of the double bonds takes place. This was proven by the reduced solubility of the final crosslinked blend membranes.
Additionally and based on the crosslinkable copolymer 3 molecular weight as well as the blend composition, a significant improvement of the Advent TPS doping ability was noticed.
Research efforts are currently devoted to the blend direction in order to select the most promising copolymer composition, molecular weight as well as blend compositions that will enable the high temperature operation of these highly doped membranes with an appropriate and reasonable stability.
Scale up synthesis of MEAs
The idea was to investigate and upscale new polymer electrolyte aromatic polyethers bearing side chain functionality that would lead to covalently cross-linked polymers by cationic curing in H3PO4. This was based on work previously carried out in the previous IRAFC project and has also been published by K.D. Papadimitriou et al.1 (J. Membr. Sci. 433 (2013) 1-9). The cross linking principle behind this type of monomers (bearing side double bonds) is that once incorporated into a polymeric structure, the polymer can be covalently cross linked through a cationic polymerization of the double bonds by the aid of phosphoric acid (H3PO4), a standard procedure followed to introduce ionic conductivity in the polymeric membrane during MEA (Membrane Electrode Assembly) fabrication. In general, cross linking is used to increase the glass transition temperature (Tg) of the membranes in order to be able to perform at higher temperature fuel cell systems (210 oC) with improved long term stability (>1000 h).
Several experiments with this cross-linking concept were carried out during the previous IRAFC project. This kind of polymers were chosen as the “hit” compound and MEA’s were fabricated in order to carry out I-V and long term experiments.
A particular batch (NP-203) was able to run continuously for 650 h at 210 oC and at 0.2 A/cm2 with a degradation rate of 21 μV/h.
To take this further to the IRMFC project, we decided to start scaling up these polymers since the scale that they were synthesized by that time was in the order of 4-5 g. We attempted to scale up the polymers in the range of 10-25 g while slightly modifying the A/B ratio, temperature, concentration and reaction time. Unfortunately, all the experiments led to fragile polymer films, possibly because of low average molecular weights (<25 kDa) compared to the previously synthesized polymers with higher Mn values (35-40 kDa).

Since we could not obtain high Mn values (and thus good films) we decided to follow a slightly different approach: combine the properties of an already known polymer (with excellent film properties and fuel cell performance/durability) such as TPS® (Advent), and the cross-linking property of the 2, 2’-diallyl bisphenol A and synthesize different terpolymers (4 monomer polymers).
We started experimenting by lowering the percentage of double bonds in the final polymer and incorporating the methyl diols, while keeping the main chain pyridine diol at specific percentage to keep the phosphoric acid uptake at high levels. The excellent film properties obtained, led us to the synthesis of some more polymers with similar stoichiometry and average molecular weights in the range of 33-38kDa.
Promising polymer films were fabricated into MEA’s, using Advent’s standard MEA manufacturing procedure, and were subjected to performance and long term durability tests. The MEAs prepared had an active area of 25 cm2.
The MEAs were tested in single cells under H2/air atmosphere and ambient pressure. The effect of temperature on performance was recorded by taking polarization curves from 180 oC to 220 oC. For constant current density of 0.2 A/cm2, a voltage increase is observed from 649 mV at 180oC to 690 mV at 220oC. The ohmic resistance as measured from EIS (electrochemical impedance spectroscopy) decreased from 100mOhm*cm2 to 90 mOhm*cm2.
Following the above promising single cell results of MP326 in addition to the excellent film forming quality, a scale up was attempted. This was a successful scale up from 5g of MP326 to 50g of MP330. Both films presented identical behavior in single cell testing conditions.
MEAs from both MP330 and MP343 were prepared with the standard active area of 25 cm2. The obtained voltage at 0.2 A/cm2 (210oC) was 680mV for MP330 and 665mV for MP343. They both ran under constant current showing a performance degradation after 300 h of continuous operation. In more detail, long term stability testing of MP330 showed no degradation up to 290 h, however for the next 720 h a degradation of 130 μV/h was measured accompanied by an open circuit voltage decrease.
MP343 long term test profile was slightly improved, showing though the same trend. For the first 340 hrs of operation the voltage remained stable, however a degradation of 80μV/hr was observed for next 430 hrs of operation.
The trends/drops in long term stability of these MEAs, led us in further investigations in the polymeric structure. We were not sure if the degradation observed had to do with the polymeric structure, the acid uptake or other parts of the MEA such as the catalyst layer etc. The observed oxidative stability and degradation under long term ked us to synthesize polymers with lower cross linkable monomer composition. We decided to go as low as 5% and 3 new polymers were synthesized. Lower degree of cross linking would mean that a lower Tg value would be obtained. With these compositions we were able to make excellent polymer films (high Mn) and also increase the amount of monomer A and thus achieve higher doping levels, higher ionic conductivity, and improved performance.
MEAs of active area 25cm2 were first prepared using polymer film MP354. They were tested in single cells under H2/air atmosphere and ambient pressure. For constant current density of 0.2 A/cm2, a voltage increase is observed from 660 mV at 180 oC to 681 mV at 210 oC. This performance was improved compared to the MP343.
Next step was the long term stability run under constant current for MP354. There was a continuous degradation of 80 μV/h. So, even though the performance was better compared with previous prepared polymers, degradation was still observed.
A new polymer MP355 was synthesized by slightly changing the monomer ratio (pyridine diol/diallyl diol) in order to increase the molecular weight of the polymer as well as the acid doping ability of the membrane. We assumed that the low average molecular weight of MP354 was a possible reason of degradation at this point. An MEA was prepared and the performance of MP355 was similar to MP354 as shown below. The MEA prepared using MP355 showed the best stability over time so far, when ran under constant current 0.2 A/cm2 at 210 oC. No degradation was observed up to 270 h of operation, while for next 500hrs the degradation was in the order of 36 μV/h.
After assessing all the gathered results, we made a conclusion that MP-355 was the best material so far and a decision was made to scale it up. The polymer was successfully synthesized in a 150g scale, with high purity and good average molecular weight.
erformance of scaled up MP360 based MEA in single cell was identical to MP355 as expected, and the long term stability (Figure 1) under constant current showed excellent results. Specifically, the first sample (MP360) stopped running at 400hrs due to hydrogen feed shortage, with no degradation until then. This was replaced by a second sample (MP360b) which ran with no degradation after 800hrs of operation.
In addition to the above assumption for degradation, we also wanted to see what the effect of acid evaporation on performance is and how this is connected to degradation issues. We know that phosphoric acid evaporation rate dramatically increases as temperature increases above 180 oC.
The acid content of “fresh” MEA’s (pressed only, not used in a cell) and “used” MEA’s (pressed and then long term run in a fuel cell) by titration was measured. Specifically, the acid was extracted from a sample of 2.4*2.4cm2 area by stirring with acetone/water solution. This solution was further titrated with NaOH 0.1N while monitoring the pH change. Two samples were tested for each MEA. For MEA MP330 the decrease is in the range of 82% loss while for MP360 a smaller loss is observed (16%). It is worth mentioning that the MP360 MEA ran for 800hrs with no degradation and was stopped on purpose in order to see the acid content of the MEA without any evident degradation. As it can be seen from the results, most of the acid is retained within the MEA. Samples MP354 and MP355 show a similar loss of acid content; however the degradation rate for MP354 is higher.
After carefully observing all the obtained results that have been described so far, we can conclude that MP360 is the polymer of choice for IRMFC. MEA’s made from MP360 was sent to partners (ZBT, FORTH/ICEHT). These MEA’s were prepared from the same membrane batch (a polymer film casting on a 30cm X 70 cm (0.21m2) glass surface) in order to obtain consistent and easy to analyze results.

Our next goal was to test MEA’s from MP360 under on-off cycling operation (required testing for final auxiliary power application). This type of testing would have to be performed using ZBT plates that can withstand continuous operation at 210 oC. The plates were sent from ZBT and were further machined by Advent using the standard triple serpentine design. In order to test the ZBT plates behavior under standard conditions, two different experiments were carried out involving IV curves and impedance curves. The plates were placed in a standard Advent single cell and an initial test was performed using Advent’s standard TPS® MEA in a single cell using Poco graphite commercial plates. The same MEA (TPS) was then tested using the ZBT plates. The results were identical for both plates. The on-off cycling (Figure 2) was performed using a simulated reformate gas (66.2%H2, 21.6% CO2, 11.6% H2O, 0.5% CO). An MEA was fabricated using polymeric membrane MP360. The on-off cycling was as follows: The current (0.2 A/cm2) was cut at 210 oC, the temperature was decreased to 140 oC under OCV conditions and then increased back to 210 oC under the same open circuit conditions. At 210 oC, current was applied again and successive on-off cycles took place until failure.
By comparing the IV graphs of the MEA at the beginning of life and end of life, it is obvious that ohmic resistance is increased (change in slope in the ohmic region increases) while by comparing the EIS data, polarization resistance is also increased by 10 mOhm Both results imply that acid evaporation is the main reason for this, however further investigation is needed to understand the phenomenon and provide a solution. A first approach, and since we cannot change the evaporation rate of acid, is to work on the electrode and change its composition in order to prevent the acid to evaporate. This can be achieved by increasing the hydrophobicity of the electrode, in other words to create smaller pores on the catalytic layer and prevent acid to move, or add a pyridine based binder in the catalytic layer that will hold the acid.
2nd period
Efforts during the second period were concentrated on the scale up of the cross-linked polymers of choice and investigation of alternative structures/monomers that were suggested by UPAT (2,6-pyridine diol). In addition, long term testing of MEAs was continued and on/off cycles of the MEAs were performed using reformate gas. The MEAs were characterized after testing in order to investigate the degradation paths.
As far as polymer structure is concerned, the cross-linked polymer shown in Scheme 1 with x=70%, y=5% and z=25% is the material of choice for IRMFC because of the lowest degradation over time and high phosphoric acid retention.
The path used to conclude on monomer percentage was by investigating the degradation parameters related to chemical structure. The tools used were the long term stability data and phosphoric acid titration data before and after testing. One of the most critical outcome of this study was that decrease of diallyl content enhances long term durability. This can be observed from the comparative IV graphs of Figure 3 where the diallyl content of the polymer used, was decreasing as follows: MP330>MP343>MP360. MP360 with the 5% diallyl shows the best stability over time.
In addition to the long term stability test with pure hydrogen, the MEA had also ran under synthetic reformate which stimulated the composition of the methanol reformate stream. The results are shown on Figure 4.

The IRMFC polymer has been successfully scaled up in single batches of 100g, 250g, 350g and 450g. Large area membranes (0.21m2) have been prepared via standard film casting procedures while MEAs have been manufactured out of these batches using Pt based electrodes and tested accordingly to pass Advent QC testing. According to partners' drawings and specifications, 8+24 MEAs have been prepared for ZBT and 6+34 MEAs for FORTH for the short-stack modules and final fuel cell stack modules.

An alternative monomer structure to that of the 2,5-dibromopyridine was suggested by UPAT and was also investigated and evaluated for scale-up. The general structure of these new polymers is also shown in Scheme 2.
After concluding on the x, y ratios Advent proceeded to the scale up of the 2,6-pyridine diol polymer on a 10g and 30g scale. Representative IV curves are shown in Figure 5 with the performance to be similar to the IRFMC polymer of choice (Scheme 1).

These initial results were very encouraging, however further investigation is needed in order to optimize the durability over time and make it an alternative option to the first material of choice of IRMFC project.

UPAT progress with respect to the improvement of cycling behavior of MEAs
In the second period of the project, UPAT efforts focused on the stabilization of the membranes during doping in H3PO4 85%, in order to improve the MEAs’ ability to operate under cycling conditions.
Thus, a reinforcement technique was employed for the stabilization of membranes prepared either of linear or cross-linkable copolymers bearing pyridine units. Moreover, the cross-linking of the membranes carrying side double bonds proceeded as effectively for the reinforced membranes as for the primary membranes case, while doping in H3PO4 85%. These facts demonstrate the applicability and potentiality of the reinforcement technique as an extra stabilization step of the membranes used in MEAs for HT PEMFCs.
For that reason new high molecular weight cross-linkable copolymers with high doping ability were synthesized. Optimization in terms of MW and cross-linking units content enabled the preparation of high quality membranes with extremely high phosphoric acid uptake. These copolymers were used as polymeric matrixes for the formation of reinforced membranes using PTFE fabrics. The combination of the cross-linkable polymers with the reinforcement fabrics resulted in a new generation of membranes with high and controllable phosphoric acid uptake and control of their swelling bahaviour e.g. negligible expansion of the membranes in the x, y axis. A representative example of a reinforced membrane prepared during this study is shown in Figure 6.
Regarding the novel quinoline bearing polymer electrolytes P(OPh5FQ) and after optimization of the polymerization conditions covalent cross-linking through aromatic polyether bond formation enabled the creation of wholly aromatic cross-linked polymeric electrolyte membranes (Figure 7). More specifically, the perfluorophenyl units were cross-linked with other hydroxyl end-functionalized moieties, providing membranes with enhanced chemical and mechanical properties that were moreover easily doped with phosphoric acid even at ambient temperatures. All membranes were evaluated for their structural and thermal characteristics and their and doping ability with phosphoric acid. Notably, all cross-linked membranes presented high doping levels and at the same time maintained their dimensional stability along the x,y axes after doping. Especially in the cases on non-dopable cross-linkers the membranes presented negligible changes along their x,y dimensions. Selected cross-linked membranes were further tested in terms of their single cell performance at the temperature range 160°C - 200°C showing promising performance and very high conductivity values up to 0.2 Scm-1.

Methanol reforming catalysts
1st period
Two different types of methanol reforming catalyst were investigated:
Copper-based catalysts
FORTH/ICE-HT focused on the optimization of Cu-based methanol reforming catalyst synthesis method (scale-up) with respect to (i) activity at 210-220oC, (ii) long-term stability, (iii) pyrophoricity and (iv) conductivity. A new methodology was followed in order to avoid voluminous and heavy foam based catalysts of previous IRAFC project, by employing a CuZnAlOx commercial catalyst (HiFUEL® R120) supported on the gas diffusion layer of a carbon paper (GDL-CP), yielding an ultra-thin reformer of 400-600 μm (compressed thickness, depending on the catalyst loading). This catalyst has the following advantages as compared with CuMnAlOx catalysts developed during previous IRAFC project: (i) readily available in large quantities, (ii) low cost (0.15 €/g), (iii) higher reducibility at temperatures lower than 210oC, thus not requiring ex-situ preactivation step, (iv) higher activity. Taking into account the negative effect of (i) unreacted methanol on the electrochemical performance of MEA and (ii) phosphoric acid on the reformer catalytic properties, a special IRMFC assembly was constructed. Two different anode flowfield designs were applied and different types of separation plate were studied. Taking into account the engineering and catalytic aspects of the present work and the operational requirements of an IRMFC, the in-plane configuration was chosen for further studies. Regardless of the H2 inhibition effect, impressive behavior appears in this low temperature region and conversions higher than 90% can be achieved under concentrated methanol/water feeds at high contact times. This catalytic performance is sufficient to support the operation of a high temperature MEA at 0.2 A cm-2 with a stoichiometric ratio of 1.2. Using a 30% MeOH/45% H2O/He feed stream, the methanol reforming catalyst performance was quite stable and ca. 15% decline in methanol conversion was observed during the whole period of the catalytic run (300 h).

Palladium-based catalysts
The main conclusions on the palladium-based catalyst, potentially applicable to the high-temperature PEM fuel cells with the internal reforming of methanol, resulted from above described studies, are as follow:
• Preparation of palladium-zinc oxide-based catalyst by impregnation from aqueous salt of palladium (used in the previous IRAFC project) does not successfully lead to the catalyst with much lower palladium loading. 10 wt.% of palladium is still required to achieve simultaneously sufficient activity and low carbon monoxide selectivity in the steam reforming of methanol.
• Preparation method by chromium-modified zinc oxide support impregnation from a non-aqueous palladium salt solution successfully leads to the catalyst with lower palladium loading. The 5 wt.%-palladium containing catalyst with the support modified with 30 wt.% of chromium, synthesised by impregnation from non-aqueous solution of Pd(acac)2, makes possible to obtain:
• higher methanol conversion
• lower selectivity of carbon monoxide formation
• higher yield of hydrogen
than those obtained over the IRAFC catalyst, i.e. 15wt.%Pd/ZnOCr15
• Decreasing of palladium loading below 5wt.% leads to worse properties of the palladium-based catalyst in the steam reforming of methanol.
• Co-impregnation of the zinc oxide-chromium modified support with palladium and zinc acetylacetonates solution leads to a better catalyst regarding its initial properties in the SRM and also regarding its stability and resistance to deactivation during shutdown-start up cycles than it was achieved with the catalyst synthesized by single-impregnation of the support with palladium salt only.
• The increase in the chromium modifier amount in the support makes reduction of palladium oxide more difficult (however, in all palladium-based catalysts the reduction is complete below the temperature of the SRM).
• Among palladium-chromium modified zinc oxide support co-impregnated catalysts, 10 wt.% of that modifier seem to be the optimum for the best stability and resistance of the catalyst to the SRM shutdown-start up cycles.
• Pre-reduction of palladium-based catalyst at 350oC before supporting it inside
a fuel cell guarantee better catalytic properties in the steam reforming of methanol. The pre-reduced catalyst is also more resistant to oxidation and more easily re-activated in the MeOH+H2O reaction mixture at reaction conditions during the SRM re-starting after previous process shutdown and catalyst cooling in air stream.
• The shutdown of the SRM carried out over 5 wt.%-palladium loaded catalyst and the catalyst cooling in air should be started at the temperature of 120-130oC.
• The higher partial pressure of the MeOH +H2O vapours (equivalent to the lack of any inert gas in the reaction mixture) enables the SRM over palladium-based catalyst to be more selective (with a little lower conversion of methanol only)
• The decreasing of the H2O/MeOH molar ratio in the reaction mixture to the value of 1.1/1 enables the steam reforming of methanol to go more selectively
• The CuZnAlOx (HiFUEL R120) catalyst enables us to achieve higher conversions of methanol and water in the steam reforming of methanol than attainable over the CuMnAlOx (FORTH) and PdZnCrOx (UMCS) catalysts. In the range of low 200-220oC temperatures the activity of the 5wt.%Pd/ZnOCr (UMCS) catalyst is higher (with acceptable selectivity towards carbon monoxide) than that of the CuMnAlOx (FORTH). The palladium-based catalyst is less pyrophoric than the copper-based catalysts are.
• The estimated cost of the 5 wt.% palladium loaded catalyst is acceptable for its application in the high-temperature PEM fuel cells with the internal reforming of methanol.

2nd period
Successful scale-up synthesis of new ultra thin reformers (50 cm2, 400-600 μm compressed thickness) was carried out. The copper-based catalytic layer (HiFUEL® R120) was supported on the gas-diffusion-layer (GDL) of a carbon paper and 300 h testing under methanol reforming reaction conditions (including startup/shutdown cycles) demonstrated the functionality of the unit. CuZnAlOx (HiFUEL® R120) catalyst resulted in higher conversions of methanol and water in the steam reforming of methanol than attainable over the CuMnAlOx (FORTH) and PdZnCrOx (UMCS) catalysts. Therefore, copper-based reformer was chosen for IRMFC measurements in the present of MEA. Methanol conversions higher than 85% were obtained at 210oC with ca. 15% deactivation. Optimization of the flow distribution along the catalytic bed employed an in-plane cell configuration. The reformers were tested in an IRMFC type single cell (in the absence and in the presence of MEA) with flowfields with active areas from 25 up to 50 cm2.
In the second period, UMCS continued the work (i) on optimum palladium loading in palladium-based catalyst since these materials showed in the previous work good activity, selectivity and resistance to shutdown-startup cycles when the cooling in air stream takes place at 120oC, and (ii) on microscopic characterization of copper-based catalyst used in tests of short IRMFC modules as well as the depositions found inside the flow-field of the MEA anode compartment. Scale-up synthesis of copper-based methanol reformers was successfully carried out in the previous period of the project.
A double reformer arrangement was applied in an IRMFC type cell in order to ensure methanol conversion higher than 95% and sufficient hydrogen flowrates (supporting MEA operation at 0.2A/cm2) for more than 1000 h (including startup/shutdown cycles) (Figure 8). 130 reformers were prepared and provided to FORTH and ZBT for short and final modules testing.
Design, construction and testing of bipolar plates, short IRMFC modules and 100 W stacks
1st period
220°C stable graphite compound based bipolar plates were successfully manufactured at ZBT and the obtained cells were successfully integrated with MEA and reformer provided from Advent and FORTH, respectively. Plain plate material was also provided to FORTH and a single cell was constructed. The performance of the obtained cell was the same as the one with Poco® graphite plates. The IRMFC operated efficiently for more than 72 h at 210°C with a cell voltage of 642 mV at 0.2 A cm-2 (low contact time 2.41 g s cm-3) and 789 mV at 0.1 A cm-2 (high contact time 4.81 g s cm-3). 18 graphite composite plate samples of 140 X 140 cm² were manufactured by ZBT, which would be sufficient for the planned 100 Wel IRMFC module, as well as single and short modules of 50 cm² of active area (ca. 70 cm² of cell area).
A short stack with metallic bipolar plates was also developed by IMM and FORTH, taking into account the specific weight and volume requirements of the project (100 W system with less than 3.5 kg and 5 l). Two cell measurements with new metallic plates (in-plane flow distribution), ultrathin reformer and Advent’s crosslinked MEAs demonstrated the functionality of the unit (1518 mV at 0.1 A/cm2 at 210oC; anode feed: 30% MeOH/45% H2O/He; cathode feed: pure O2).
The most efficient methanol reforming catalyst, with respect to methanol conversion and utilization at low reforming temperatures of around 210-220°C was adapted from FORTH and ZBT to the IRMFC designs according to a configuration where there is a separator between the reformer and the anode electrode (indirect configuration). ZBT developed a pocket type design taking into account the temperatures, vaporisation and super heating, liquid and gaseous media distribution, startup and shutdown as well as nominal operation of the reformer. The location of i) in and outlets for the fuel methanol and water, as well as ii) the outlet of the reformate gases were optimized by performing necessary simulations. On the other hand, FORTH’s design of the reformer compartment followed the typical arrangement of an anode electrocatalyst, since the methanol reforming catalyst was supported on a carbon paper. Both partners performed tests to study fuel conversion, space velocities and reformate gas compositions at the desired operating conditions to further optimize the reformer unit. It can be concluded that ZBT and FORTH has successfully integrated the methanol reforming catalyst into the IRMFC unit. Analysis of the geometrical parameters related to the reformer compartment showed that an additional 15% reforming catalyst area would be required to produce H2 required for target fuel cell operating current. For ZBT this can be easily done by adding more reformer compartments in the final stack, where the reformate gas from all the reformers will be collected in a single manifold and then will be distributed in the anode electrodes flowfields. On the other hand, FORTH developed a double reformer arrangement, ensuring methanol conversion higher than 95% for long operation periods, thus generating the required hydrogen for the target load of 200 mA/cm² (650 mV). Preliminary tests demonstrated the efficiency of the modified configuration and construction of 25 W short stack modules was completed.
2nd period
Taking into account the outcome of the first period with respect to the requirement for additional reformer loading in order to support high methanol conversions for long periods at IRMFC operation conditions, various IRMFC modules consisting graphite-compound based BPPs were successfully constructed, tested and evaluated at ZBT. A 20C40R (20 MEAs-40 reformers) module (Figure 9) produced 115 Wel on a load current of 219 mA/cm² when fed with H2/Air and 112 Wel on the same load current when fed with Reformate/Air at 190°C.
Figure 10 depicts performance comparison when the 20C40R module was operated with H2/Air as well as with Reformate/Air. As can be seen from Figure 10, there was a 3 Wel loss in stack power when operated with methanol reformate compared to pure H2/Air feed, which translates to a loss of 150 mW per cell at a load current of 219 mA/cm². It would be necessary to operate the module at more than 300 mA/cm² to generate enough heat to sustain the module at the required temperature of 190°C without any external heat supply. At lower load currents of 9,1-12 A, additional heat needed to be supplied to maintain a module temperature of 190°C.
It was intended to study the long term performance behaviour of the 20C40R module. As can be seen from Figure 11, the load current was kept at 10A (219 mA/cm²) while maintaining a premix feed rate of 200 g/h into the 40 reformers of the module. During the first one hour the module voltage went up to to 11,65 V and after that the same began to fall to 10V. On a load current of 9,1A (200 mA/cm²), the module voltage was initially 10 V and during a 2 hour operation, the same fell to 8,8 V. This corresponds to a 1,2 V fall in module voltage or 0,06 V per cell or 30 mV/cell/hour. This observed higher voltage loss could be attributed to many factors including a) lower operating temperature of 190°C instead of the planned 210°C operation, b) FC anode catalyst’s performance at lower temperatures when fed with CO and CO2 containing reformate species as well as trace amounts of methanol and CH4, c) H2 partial pressure distribution across the active area of each cell. In the end, to avoid irreversible cell degradation, the load current was reduced and an additional test was performed with H2/Air feed. The second test with H2/Air feeds has shown that the stack had experienced a 100 mV permanent voltage loss.
It may be concluded that the 20C40R IRMFC module consisting graphite-compound based BPPs was successfully constructed, tested and evaluated. The same has produced 115 Wel on a load current of 219 mA/cm² when fed with H2/Air and 112 Wel on a load current of 219 mA/cm² when fed with Reformate/Air at 190°C. These values correspond to:
Specific size (a) = 3,19 lt/112 W = 28,5 l/kW (under MeOH-water/air, 190oC)
Specific size (b) = 3,19 lt/115 W = 27,7 l/kW (under hydrogen/air, 190oC)
And
Specific weight (a) = 5,5 kg/112W = 49 kg/kW (under MeOH-water/air, 190oC)
Specific weight (b) = 5,5 kg/115 W = 49,1 kg/kW (under hydrogen/air, 190oC)

Due to gasket limitations and continuous degradation of the stack, especially under cycling conditions (low MEA stability), and in order to avoid irreversible stack degradation, the stack wasn’t tested at the intended temperature of 210°C and for longer periods.

Modules with metal-based bipolar plates. FORTH and IMM designed and constructed 4MEAs-4DRef and 32MEAs-32DRef stacks based on an innovative double reformer arrangement and metallic bipolar plates with unique flowfields arrangement. The latter stack was tested by FORTH and resulted after several on/off cycles in a power output of 91 Wel (2,84 W per MEA; 0,103 W/cm2) at 200oC and 180 mA/cm2 stack.
The short stack’ s steady state performance was examined by means of I–V steady state polarization measurements, while electrochemical impedance spectroscopy was also employed as a powerful diagnostic tool for the determination of total ohmic resistance. Typical polarization curves obtained at 210oC, under various feeds and for several on-off cycles are shown in Figure 12. The following main results can be commented:
Operation under H2/He:
• Failure of MEA No3 after 10 cycles
• Degradation rate: 14 mV/cycle or 70 mW/cycle
Operation under MeOH/H2O:
• 2,720 V and 15 W (mean value of 10 cycles) were obtained at 0.2 A/cm2
• Failure of MEA No3 after 10 cycles
• Degradation rate: 22 mV/cycle or 122 mW/cycle
• Higher degradation rate in the presence of methanol
• Fuel cell ohmic resistance = 633 mΩ.cm2 (= 158 mΩ.cm2/cell)
• Other byproducts: CO (1.7%), acetic acid, methyl formate in low amounts
The duration of all these tests, including preliminary sealing, heating and pressure drop tests, was 3 months and most of the main components showed a very promising behaviour, especially the metallic plates and the reformers. Post mortem photos of the metallic plates showed neither cracks nor corrosion of the protective coating, in agreement with the obtained ohmic resistance after each cycle. Methanol reformers showed very stable behaviour and conversion higher than 95% were obtained during these tests. However, the fact that some deformation of MEAs subgasket took place and might caused a variation in the pressure drop per cell, resulting in non uniform distribution of the reactants/products, anode was operated with λΗ2 = 1.5. Next generation components employed in the 100 W stack, employed different subgasket material in order to avoid this kind of deformation phenomena. Moreover, slight modifications in the inlet/outlet channels were applied in order to avoid blocking of the channels, pressure drop phenomena and crossover phenomena.
100 W metallic stack
Based on the results of metallic short modules testing, the most efficient methanol reforming catalyst (developed by FORTH) and high temperature MEA (developed by Advent and UPAT), together with the 4th generation of metallic plates, were employed on the construction of the 100 W stack. Figure 13 shows the integrated 32 MEA-32DRef stack.
It has to be noted that prior activation, the stack was heated up to 210oC several times (10 cycles) and cooled down to room temperature, followed be opening of various compartments in order to make appropriate modifications in the position of gaskets. Thus, the employed materials, and especially reformers and MEAs have already been incurred with a heating/cooling cycling test. However, the polymer electrolyte membrane cannot tolerate a cycling operation with heating at such high temperatures (i.e. 210o) followed by cooling at room temperature. Moreover, the copper-based reformers are very sensitive in exposure in air after being activated/operated at the fuel cell temperature level of 200-210oC. If the catalytic surface covered by dispersed metallic copper species is directly exposed in air atmosphere, without being passivated, then severe sintering might take place resulting in irreversible progressive deactivation of the reformers.
Taking into account the aforementioned drawbacks, and especially the limited tolerance of MEAs with respect to the cycling operation, as this was evidenced in the short stack testing (Figure 12), where less than 10 cycles could be achieved, the present 32MEAs-32 DRef was tested under H2/air for 3 on/off cycles prior switching to MeOH-H2O/air in order to confirm the performance level of both MEAs and reformers, and then send the stack to IMM for integration with the BoP. More cycles could have been done, however due to the limited cycling behaviour of the MEAs, we wanted to ensure that we will a functional and active stack to be tested with the BoP.
Electrochemical behaviour of 32MEAs-32 DRef IRMFC stack
The stack was heated up to 120oC without any gas flows. At this temperature, anode and cathode were fed with 6,8 lt/min air (λΟ2 = 2.5) and 2 lt/min hydrogen (λH2 = 1.6) respectively. These lamda values correspond to the values that will be applied at 0.2 A/cm2 in order to avoid limiting current phenomena. I-V curves obtained at 200oC for the 32MEAs-32 DRef are depicted in Figure 14. No deactivation was observed, and a voltage of 20,32 V was obtained at 0,2 A/cm2, corresponding to a power output of 111,8 W (3,5 W per MEA; 0,127 W/cm2). After 3 runs, the stack was switched to MeOH-H2O/air and checked the performance of the stack at 200oC. A compact 40 X 60 mm² evaporator provided by partner IMM (the same will be used in the final system) was connected with anode inlet of the stack stack, while an HPLC pump was feeding the evaporator with the appropriate liquid flow of MeOH/H2O. MeOH and water were pre-mixed in a molar ratio of H2O/MeOH = 1,08. This ratio was applied on the basis of the stability test outcome of the reformer. Since the methanol reforming catalyst is stable for long periods under H2O/MeOH ratio lower than 1.5 and closer to the stoichiometric value, this is in advantage of the stack operation, since higher amounts of methanol are placed in the fuel tank and also lower MeOH-H2O flowrates can be applied, thus resulting in higher contact times in the catalytic bed. In this way, higher methanol conversions can be obtained. The first run at 200oC was followed by cooling at 120oC and re-heating at 200oC. It has to be noted that during cooling and heating all the flows were turned off. Restart took place at 200oC applying the previous conditions. The corresponding polarization curves are shown in Figure 14 in comparison with the operation under hydrogen/air. Taking into account the limited tolerance of the under on-off cycles and in order to ensure that the stack will be operative during testing with the BoP at IMM, no more cycles were run at this stage of the project.
Lower performance was obtained in the case of operation under MeOH/H2O. The fact that the reformers were operated at 200oC, resulted in methanol conversion lower than 90% (80% methanol conversion resulted in a reformate gas of 5,4% MeOH, 21,7% CO2, 65,1% H2, 7,69% H2O, <0,5% CO) therefore significant concentrations of unreacted methanol are introduced in the anode electrode affecting negatively its performance. However, it’s quite impressive that this generation of MEAs can tolerate methanol concentrations such high as 5%. Moreover, the power output of the stack remained stable after a heating/cooling/heating cycle and 91 W (2,84 W per MEA; 0,103 W/cm2) were obtained at 180 mA/cm2 stack, under MeOH-H2O/air.
Regarding the targets of specific size and weight of less than 35 kg/kW and 50 l/kW (fuel amount excluded) for the final system (stack plus BoP) the 32MEAs-32 DRef IRMFC stack represents the following characteristics (including heavy and voluminous compression plates and bars):
Specific size (a) = 3,024 lt/91 W = 33 l/kW (under MeOH-water/air, 200oC)
Specific size (b) = 3,024 lt/111,8 W = 27 l/kW (under hydrogen/air, 200oC)
Without the insulation box and heating module:
W x H x L = 8 X 19 X 13 = 1,976 lt
Specific size (a) = 1,976 lt/91 W = 21,7 l/kW (under MeOH-water/air, 200oC)
Specific size (b) = 1,976 lt/111,8 W = 17,7 l/kW (under hydrogen/air, 200oC)
Specific weight (a) = 4,2 kg/91W = 46 kg/kW (under MeOH-water/air, 200oC)
Specific weight (b) = 4,2 kg/111,8 W = 37,6 kg/kW (under hydrogen/air, 200oC)
➢ 32MEAs-32 DRef stack (FORTH and Fraunhofer) was chosen for further testing with the BoP (Fraunhofer), since it offers more attractive characteristics regarding the specific size and weight, while it could be operated up to 210oC for longer periods, as compared with the graphite BPPs-based stack (ZBT), which faced technical difficulties limiting its operation up to 190oC.

Design and construction of BoP
All relevant Bop components such as pumps, blowers, heaters, catalytic reactors and control boards were either purchased or developed and then characterized. They were integrated into subsystems, then tested again and finally integrated into a complete system.
An ASPEN model was set up to describe the system and to gain the specifications for the BoP components. A flow scheme of the system was made and contained information about some of the BoP components finally chosen.
The system has two different modes of operation:
1) During start-up, about 2.7 ml/min Methanol are dosed by Pump P-01 into the evaporator H-01/H-02 and mixed with air from blower B-01 in Mixer M-01. The methanol is combusted in the burner H-03 (metallic monolith) and heats the fuel cell stack (along with the evaporator (H-04) which is attached to it). The flow of combustion gases is controlled by Iris blends V-04 and V-05.
2) During normal operation methanol and water are dosed into the system by pumps P-02 and P-03 respectively, the mixture is evaporated in the evaporator H-04 and the reforming and hydrogen conversion to water and electricity takes place in the fuel cell. The fuel cell cathode is supplied with air by a diaphragm pump C-01, while an axial blower (B-02) supplies the afterburner (R-02) with air. The additional air from the small axial blower B-03 cools the hot exhaust gas from the afterburner.
Pumping units
The liquid dosing is performed by a combination of purchased pump, 3-way valve and pressure sensor.
Air blowers
A purchased radial blower was chosen as start-up blower which generates a static pressure of 56 mbar and create a flow of 250 l/min at a pressure of 25 mbar. For the purchased afterburner a purchased axial blower was chosen). Owing to the huge pressure drop of the fuel cell cathode, a diaphragm pump was required to supply it with air.
Heaters for heating the evaporators
Purchased ceramic resistance heaters were chosen for heating the evaporators for methanol.
Catalytic start-up and afterburners
Purchased metallic monoliths coated with platinum containing catalyst were prepared and used as burners. The reactors were pre-heated by electric resistance heaters surrounding them.
Set-up of the afterburner subsystem
Figure 15 shows the CAD model of the afterburner subsystem. The inlets carrying red crosses were closed for the tests.
Development of control boards
Control boards were designed at ICT-IMM and fabricated with external suppliers: micro diaphragm pump controller board with integrated µ-pump; thermocouple board stack where the measurement of 8 temperatures is possible on a single board; analog-digital-converter board where the measurement of 8 single cell voltages is possible on a single board; stepping motor controller board where the flow area is adjusted. The electronic mainboard contains:
• The fuel-cell input and load output (Measurement of current and voltage)
• The external power connection (measurement of current and voltage)
• Three push buttons (individually programmable)
• A raspberry Pi Computing module
• Twelve thermocouple connections
• Three UART connection to control the ADC-, thermocouple-, stepping motor- and pump-unit-boards
• Eight 24 Volt outputs via Darlington Array to switch the valves
• Four controllers for the heater for the methanol and water evaporators
• One controller for the heater for the monolith
• The controller for the blower
• The controller for the axial blower
• The controller for the radial-blower
• Two USB connection
• One Ethernet connection
Firstly, the main board was tested with laboratory power supply, voltage and current measurement was performed to calculate the battery parameters (energy density, number of battery packs, voltage, current). Then an external battery management board with connection to the main board was designed (external power connection and load to charge the battery pack).
Start-up system testing
A dummy cell was designed in order to simulate the fuel cell stack body and its pre-heating behavior was tested. It was made from AlMg3, has a weight of 3.75 kg, the same size as the stack and 14 temperature measuring points. The metallic monoliths were on operating temperature, while it took about 80 min to heat the fuel cell stack to a temperature of 180°C.

System integration and testing
Figure 16 shows the CAD model of the fully integrated system. The control boards are mounted at the outer surface (see Figure 16b). The system is foreseen to be operated in On/Off mode, has a design power output of 100 W, a volume of 27.2cm x 26.2cm x 12cm (8.55 liter) a weight of 4.5 kg without battery, the battery can be either Li-Ion or Li-Polymer. The capacity of the system amounts to 360 Wh. The following Table provides a comparison to the ULTRACELL system, which shows size and weight advantages of our system, namely 25% volume and 40% weight savings.
FORTH assembled the 32 MEAs-32 DRef IRMFC stack (Figure 17) based on last generation metallic bipolar plates and sent it to ICT-IMM for integration with the BoP. BoP metallic box is shown in Figure 18, while Figure 19 shows the stack placed inside the insulation casing at the test stand at IMM lab. Despite the fact that BoP unit was successfully tested in previous months the applied insulation casing could not successfully support the efficient heating of the stack at 200-210oC and in reasonable time. The main reason for such a low heating performance had to with the huge volume of the space between the stack and the casing which resulted in high thermal demands which could not be supported by the catalytic burner. Operation of the burner at full load allowed heating of the stack at a temperature of 160oC in more than 90 min (!). Therefore, the stack could not be started and operate under methanol/water if there wasn’t achieved a temperature of at least 200oC, in order to obtain reasonable methanol conversions, high enough to support the MEAs operation. This negative fact took place very close to the end of the project, and the employed partners (FORTH and IMM) decided to proceed as follows: The stack was sent back to FORTH, who designed a new insulation casing aiming at much faster startup times (less than 20 min) and uniform temperature distribution across the stack at the temperature level of 200-210oC. In order to avoid further delays the stack was integrated with the new insulation casing and tested in Patras (Figure 20). The operation of the catalytic burner was “simulated” with a heat gun able to provide a quarter of the maximum flow. Indeed, this time the new design worked properly and the stack could be heated (no flows, no current) up to 200oC in less 20 min. A compact 40 X 60 mm² evaporator provided by partner IMM was connected with the anode inlet of the stack, while an HPLC pump was feeding the evaporator with the appropriate liquid flow of MeOH/H2O. MeOH and water were pre-mixed in a molar ratio of H2O/MeOH = 1,08. The stack was heated up to 200oC without any gas flows. At this temperature, anode and cathode were fed with 6,8 lt/min air (λΟ2 = 2.5) and 2,2 ccl/min MeOH+H2O (it corresponds to a gaseous mixture of 48% MeOH/52% H2O and 25,7 cc/min gaseous methanol per cell; these values give a lamda of hydrogen equal to 1,5 at 0.2 A/cm2 in the case of 80% methanol conversion at 200oC). The obtained voltage values from the whole stack and individual cells were monitored. Since, the employed MEAs cannot tolerate more than 10 on/off cycles under the IRMFC operating conditions (210oC, 0.2 A/cm2, MeOH/H2O anode feed), 3 cycles of startup/shutdown were carried out in order to demonstrate the functionality of the system and identify the target for further improvement. Taking into account the preliminary tests of heating and sealing, this stack and the employed components have been treated in more than 30 cycles of heating (up to 210oC) /cooling (down to RT) with and without gas flows.

Figure 21 shows the polarization curves of this stack obtained before the integration with the new casing (D4.5) and after this new arrangement. The last run was carried out at 210oC and 18,3 V were obtained at 0.2 A/cm2, corresponding to a power ouput of 100,7 W (3,14 W per cell; 0,114 W/cm2). This corresponds to specific size of 30 lt/kW and specific weight of 41,8 kg/kW (including heavy and voluminous compression plates and bars and insulation casing). The fact that the performance of the cell is lower in the case methanol feed as compared with pure hydrogen feed, it is attributed to the unreacted methanol effect on the MEA (a methanol conversion of ca. 92% was obtained at 210oC, which is lower than the expected value of 97%; it is quite possible that partial deactivation of the catalyst might have taken place due to sintering effect after shutdown of the stack and exposure in air).
Overall, the observed stability of the IRMFC stack and key components (MEAs, reformers, bipolar plates) is satisfactory in the timeframe of the presented experiments. Specific targets for improvement of the efficiency have been identified. These are the activity of the reforming catalyst and the thermal stability of the membrane for operation above 200ºC and especially under on/off cycles. The latter drawback of the polymer electrolyte membrane is currently under investigation by partner UPAT and promising materials have been already developed and their attractive properties are presented in the second periodic report.

Compliance to regulations, standards and codes
Synopsis of the main results:
✓ Development of regulations, standards and codes (RSC)
➢ Full deployment of required standards.
➢ Analysis of the fuel cell core components according to standards requirements.
➢ Analysis of structured reforming catalysts operational conditions with respect to the RCS framework.
➢ Analysis of standards requirements for bipolar plates.
➢ Analysis of requirements for BoP components such as heat exchangers, evaporators, pumps and blowers. Analysis of a typical fuel cell configuration.
➢ Standard requirements analysis for full scale product. Received technical data sheets of chosen subcomponents.
✓ Conformity with European safety standards
➢ Completion of technical report
Topics on the report:
➢ European Standard EN 62282-2. Check list fulfilled with 71 checks / inspections. No fail observed.
➢ European Standard EN 62282-6-100. Check list fulfilled with 190 checks / inspections. 23 fails observed. Air filter missing from air inlet, marking and technical documentation were not demonstrated, electrical circuit for voltage supply was not demonstrated and the fuel cartridge was a simple plastic tank. The above failures are minor compared to the magnitude of the project.
Potential Impact:
Expected impact
The specific consortium leading to the establishment of new products will result in numerous benefits for the participating organizations and the industrial partners as well. The technological impact of the participants will be highly improved, since some of them are already active and leaders in hydrogen and fuel cell technologies and the proposed technology will further increase their business activities in a rapidly growing market of “green” applications taking also advantage of the energy and environmental parameters. Additional opportunities will be created by intellectual property rights ownership and possible user rights among participants. The successful implementation of novel, fuel cell based devices will open new perspectives in future energy technology.
Innovation impact
The principal output of the project and the corresponding impact will be the following:
1. New generation of high temperature polymer electrolyte membranes operating at 200-210°C. Impact: Wide application in HT PEM stacks, promoting their long term operation
2. Highly active and stable methanol reforming catalysts operating at 200-210oC. Impact: It will boost an efficient methanol reforming technology at lower operating temperatures thus simplifying the portable HT PEM systems
3. Design and construction of anti-corrosive metallic bipolar plates. Impact: Increase of the life time of HT PEM stacks and will enhance the fast development of the technology and its market penetration
4. Design of light weight and volume stack. Impact: Enabling the application of HT-PEM fuel cells in mobile applications to substitute off hybrid generators or to reduce high weight batteries to decrease weight and costs of electric vehicles
5.Fuel cell-fuel reformer in a single compact unit. Impact: lower the system costs by avoiding external reformer and expensive oil cooling. Easier penetration of the fuel cell system in the energy market.
Economic impact
Issues of increased power demand, efficiency, and emission reduction, security of supply and diversification of energy sources have led to renewed interest in fuel cells as systems for electricity generation in multiple applications. In particular the reformed methanol fuel cell as a portable energy supply equipment will be cheap and easy to handle compared to the existing battery equipment. In this respect it will very fast penetrate the market, thus allowing not only for the economic growth of the participating companies but also it will promote, disseminate and increase the public awareness for the social, economic and environmental benefits of the new energy systems.
The main direct economic benefits from IRMFC project will be the following:
1. Development of low cost robust membranes for fuel cell applications
2. High cost reduction of the methanol fuel processor-fuel cell system
3. High efficiency approaching the efficiency of a stand-alone H2 fuel cell system
4. Significant increase in volume power density with respect to separate fuel processor/fuel cell systems
5. Economic development of rural and isolated regions. Creation of new jobs related to the renewable energy carriers
The aforementioned direct benefits will result to several corresponding indirect economic benefits:
1. Use of membranes in other applications comprising H2 economy e.g. electrolysers, H2 pumps, CO2 electroreduction
2. Easier market penetration of the fuel cells and renewables with high positive consequences in reduction of air pollution, on health and environment
3. Increase in fuel savings with both economic and environmental benefits
4. Significant increase in volume power density will allow the application of the methanol fuel cells in mobile applications
5. Development of efficient energy sources for renewable back up energy
Quality of Life
The every-day life benefits of Fuel Cells primarily relate to the great reduction in air pollution emissions that FCs can achieve in comparison with the combustion of fossil fuels. In addition the proposed technology will improve the wealthiness of the everyday life as it will give the chance to be independent from the grid for electricity supply with larger autarky. Fuel cell engines will create (i) economic benefits (energy savings, creation of highly specialised jobs, massive production of renewable fuels, reduction of dependence from oil markets), (ii) energy benefits (high quality energy produced from fuel cells, energy security, energy production distribution and decentralization, (iii) environmental and health benefits (long-term solution for drastically reducing greenhouse gases, air quality). Moreover, according to the US DoE, H2 and fuel cell technologies development will result in 750,000 new jobs by 2030. The potential of the H2 and fuel cell economy is very high. The impact in European job market will also be proportional. The IRMFC will contribute through high efficient and low-cost fuel cell technologies to their rapid penetration into market, while the established consortium is able to stimulate and ensure high level scientific career opportunities.
Financial and commercial exploitation of results
Markets
The IRMFC project complies with the EU legislation for sustainable development as its objectives aim at opening new scientific and engineering prospects, which may allow easier penetration of the fuel cell system in the energy market. Existing market for the reformed methanol fuel cell encompasses portable/transportable fuel cells to be used as:
• An APU or a Plug in Hybrid for electric vehicles like cars or delivery vans with heat and power combination for winter use (heating) and summer use (cooling) without lowering the range
• Telco and industry applications like measurements (environment, traffic) or construction (wind, railways)
• Security and military applications
• Off-grid home energy supply e.g. leisure homes, micro grids
• Application with low power consumption like: Internet of things, battery chargers, laptops.
Examples for Emerging markets
APUs (auxiliary power units) for mobile applications
A relevant mass market for APU applications are electric vehicles. To rise the number of such environment friendly cars, national incentive programs have been set up in many countries. Germany just announced a 1 billion € supportive program for 400.000 EVs. But the subsidies will phase out after 2020. Therefore a competitive product has to be offered to customers. Most cars, powered by a battery only, have the problem of a non-predictable range and especially a very low range in winter times. A small HT-PEM system e.g. the Ecoport 800 can overcome the disadvantages, when it is used as an APU for electric consumers like lamps for example. It provides also heat, to keep the range stable in winter times. Furthermore, exhaust heat and heat from a methanol burner could be used. The unit provides also the energy, which is necessary for air conditioning in summer. Those applications allow predictable range and carefree operation.
APUs fuel cells systems running on methanol hold a great deal of market potential, as they offer a ready and widely available source of power for mobile applications. The market for leisure applications like RV’s and boats is currently served by diesel or gasoline generators, RVs seem an ideal market for fuel cell systems because the customer need a quiet, lower emission, and they offer vibration free power source. The marine segment for >30 ft boats has a global production of approximately 500,000 per year mainly in the US and Europe.
Hybrid systems for battery powered applications and micro grids. To apply a methanol fuel cell to a Plug-In-Hybrid raises the quality and range of electric vehicles. Methanol can be easily and quickly refueled. The driver is more independent from the next charging station. Even a small range extender provides the car driver from a still stand. A fuel cell with a power of 1 -2,5 kW avoids or lowers battery discharge in traffic jams. Typical first mobile applications are: light commercial vehicles (freight), utility vehicles (taxi, etc.), small refrigerated trucks. Development and higher production capabilities helping to reduce weight and price, are necessary to enter the market. High interest is shown from logistic company because the deliveries within city will be electric (German post needs more than 10.000 delivery electric vans).
Rural infrastructure. The largest energy market is the rural electrification market in light of 1.6bn people without access to the grid, or 3 bn with poor grid stability. On the other hand energy systems based on regenerative sources are strongly emerging (micro grids). Because of the unsteady electricity of PV and Wind power plants a high demand for secure and ecological backup system for regenerative energy is given. Even low price lead batteries could not store economically the necessary amount of energy for critical infrastructure applications. For these applications a diesel generator is used. However, viable solutions and strong intentions to make emerging countries oil independent by hydrogen or methanol, are being discussed and studied.
First market is the less price sensitive industry market. The customers are willing to pay for high quality, if the product can assure reliability and security for their application. Especially in segments like measurement (Oil and gas, Mining) or surveillance the benefit of continuous and reliable measurement data is very important, as the losses of data may cause high costs e.g. within a wind survey. Another example is the observation of suspects by the police. The telecom market namely off-grid power supply for mobile base stations has become an attractive segment for fuel cell providers. This is due to the low maintenance of fuel cells compared to diesel engines but also due to regulations with the objective of reducing exhaust gases in these applications. The market potential is over 1 bn. € p.a.
The industrial partners involved in the project will define the most promising market segments based on the type of project exploitable results (i.e. flow field plates, membranes, catalysts, etc.). Specific knowledge management and protection (IPRs) will be detailed in the Consortium Agreement (CA) signed by all partners. Market trends and competition will be continuously monitored during the project. As described in previous subsection for the dissemination plan, all items of exploitable knowledge will be identified to create the project business portfolio. The exploitation plan document includes the strategy for patent applications and analysis of prospects of each patent to be exploited either through internal activities by the industrial project partners and the establishment of new start-up or through licensing/joint development with external stakeholders, previously identified during the Market Survey. The exploitation activities already started by the partners will be extended even beyond the end of the project. Alternative exploitation is possible is possible via private investments for start-ups or joint venture establishment.
Exploitable results - markets
The exploitable results regarding (i) high temperature polymer electrolytes, (ii) high temperature PEM fuel cell membrane electrode assemblies, (iii) methanol reformers, (iv) high temperature bipolar plates, (v) flowfields design, (vi) fuel cell stacks, will be addressing the needs of universities, polymer and membrane producers, electrocatalysts and electrode producers, MEA producers, stack and energy production systems manufacturers, while the expected results of the project will lead to the low cost production of materials (electrolytes, catalysts, bipolar plates and MEAs), and substitution of the voluminous fuel-processor-assisted fuel cell systems with the high volume power density of a reformed methanol fuel cell. The individual results can be used in other applications of the hydrogen economy as well, e.g. the membranes can be used in electrolysers, hydrogen pumps, and CO2 electroreduction.
More specifically, the successful demonstration of the proposed reformed methanol fuel cell can be viewed as an enabling technology, opening the road to a wide variety of new applications and also a wide variety of strategic fuels which could increasingly come from renewable sources. The main innovation stems from the fact that liquid fuel, in this case methanol is used. This makes the system more simple (no complex hydrogen storage bottles needed) and more compact, thus saving valuable space. Hence, the first group of applications for this system consists of mobile applications, including military applications and hence both dissemination and exploitation of this technology should target this sector first. Mobile applications include a large variety of applications including forklifts, 2/3 wheeled, locomotive (train), barge, automobiles, sailing yacht, long-haul truck, mid-large marine, commercial aircraft as well as portable devices such as mobile generators, soldier power and electronics. Moreover, the proposed power unit is the perfect addition to off-grid home systems. It compensates photovoltaic installations and batteries where they are weak – during low solar radiation and cold temperatures, thus offering reliable heat and electricity all the year round.
The Enterprise Europe Network will also play an important role in the exploitation of the technology, which provides a large pool of potential buyers and future development partners, including individual companies, industrial associations and clusters of companies. The consortium already took part in relevant technology transfer brokerage events (Hannover Fair). Technology missions were and willl be contacted through the Enterprise Europe Network in order to present the project results to groups of interested companies, eg. Regional or National Clusters or associations will be organized. Various European Technology Platforms related to the industries targeted were and will also be target groups as they concentrate the most relevant and key players in the sector. Indicative platforms which were/will be targeted for dissemination and exploitation purposes are the: Advisory Council for Aeronautics Research in Europe, the European Rail Research Advisory Council, the European Space Technology Platform and the Waterborne ETP.
As already mentioned the technology utilizes methanol as fuel. Hence this will open the road to the use of more types of liquid fuels, and especially those coming from renewable energy sources such as bio-methanol (from biomass) or methanol products from renewable sources other than the use of natural gas. In this case, the European technology platforms dealing with this, such as the Renewable Heating & Cooling ETP, the European Biofuels Technology Platform and the European Technology Platform for the Electricity Networks of the Future will also be kept informed.
Detailed Dissemination Plan
The dissemination plan includes the following:
Who: role of each Consortium member for the execution of dissemination actions
How: adoption of dissemination approaches based on target audiences and allocation of funds and resources
What: items, layout and information to be included
To whom: list of target audiences.
In particular:
•The Coordinator set up the project web site, which was the main point of reference to obtain detailed information on its objectives, partners, results, and project deliverables
•The SME/industrial partners disseminated the relevant information to stakeholders in their field by participation in specialized fairs and contact with specialized press and media.
•Links with the Research and Industry Groupings of the Fuel Cells & Hydrogen Joint Undertaking were/will be established to demonstrate the possibilities of the new-developed technology to this large target group.
•Networking with other similar R&D projects (common events with ongoing projects in the framework of FCH-JU2) and networks (e.g. Enterprise Europe Network), clustering activities with regional and national SMEs networks on eco-innovative and material technologies (e.g. INSME, European Forum on Eco-innovation)
•Academic partners did/will widely disseminate the project within the scientific community through national & international scientific meetings, scientific articles for publication to peer review journals of high impact.
The knowledge management and protection (IPRs) is detailed in the Consortium Agreement (CA) signed by all partners. The CA defines among other issues the rules for the following points: Confidentiality, secrecy and patenting; Publication and dissemination; IP ownership: Background IP and Foreground IP; IP Use and Access Rights and Access Rights to foreground and background IP.
The exploitation plan includes two main activities:
Market Survey: the industrial partners defined the most promising market segments based on the type of project exploitable results (i.e. individual components, fuel cell, membranes, catalysts, etc.). Market trends and competition were/will be continuously monitored during and after the project.
Exploitation of results: All items of exploitable knowledge were identified to create the project business portfolio. The exploitation plan document also includes the strategy for patent applications and analysis of prospects of each patent to be exploited either through internal activities by the industrial project partners and the establishment of new start-up and joint ventures or through licensing/joint development with external stakeholders, previously identified during the Market Survey. The exploitation activities started in the second year of the project after the proof of concept and will be extended even beyond the end of the project especially in the case that or start-up joint venture will be established attracting private investments.
Communication activities
The information of the project was/will be posted in the project website, as means of mass communication with the wider public. The project has created a common database for all the project documentation (deliverables, publications, management procedures, strategies, research experimental data, regular reports and minutes, etc.) as a common tool for communication.
Dissemination activities
Dissemination strategy
Dissemination of the information to the scientific community was/will be done through the publication of selected information from the experiments and data analysis of the project in international “peer-reviewed” journals and relevant conferences. The information is also posted in the project website, as means of mass communication with the wider public.
Dissemination of the information to the relevant industrial world follows more or less the same mechanism of publications, scientific conferences and public website. On top of this and because not all of the industrial world follows scientific conferences, the dissemination of the work progress and most importantly the project results will be made through the Enterprise Europe Network, the largest European network offering services primarily to SMEs with more than 550 organizations members in 44 countries supported by the Competitiveness and Innovation programme and providing amongst others Innovation, technology and knowledge transfer services. A project partner, Advent Technologies is already working closely with the Enterprise Europe Network – Hellas partner, HELP-FORWARD. The Enteprise Europe Network is the mechanism recommended by the IPR help desk concerning the dissemination and exploitation of research results coming from EU funded projects.
Whole period dissemination activities
The results of the project were presented in 36 international/and national scientific conferences, such as 2013 & 2014 AIChE Annual Meeting, EHEC2014, 4th & 5th European PEFC & H2 Forum, EPF2013, 10th European Congress of Chemical Engineering, 230th Electrochemical Society Fall Meeting.
4 publications in peer-reviewed international journals (Journal of Power Sources (http://dx.doi.org/10.1016/j.jpowsour.2015.07.055 and http://dx.doi.org/10.1016/j.jpowsour.2016.01.029) , International Journal of Hydrogen Energy (http://dx.doi.org/10.1016/j.ijhydene.2014.03.101). New designs, reformers arrangements, performance evaluation was reported in these publications.
Moreover, exploitation of the technology resulting from this project was done through the participation of project partners in brokerage events, mainly Hannover Fair 2015 & 2016. Three partners of the consortium, namely ZBT, IMM and Advent participated in this exhibition, and the core technology, main objectives, and current progress were presented via posters and interviews.
The project was also presented during the FCH JU Programme Review Days 2013-2014-2015-2016 held in Brussels every year as poster presentation for boarder dissemination of the obtained results.
The information was also posted in the project website, as means of mass communication with the wider public. The project has created a common database for all the project documentation (deliverables, publications, management procedures, strategies, research experimental data, regular reports and minutes, etc.) as a common tool for communication. The content of the database is regularly updated and restricted by secure access control.
http://irmfc.iceht.forth.gr/index.php
Intellectual property rights management
Background information
Partners have already indentified in the signed consortium agreement all Background information which they are ready to grant Access Rights to, as well as the ones the wish to be excluded.
Foreground information
According to the consortium agreement signed before the contract signature, knowledge shall be the property of the contractor generating it. Where several contractors have jointly carried out work generating the knowledge and where their respective share of the work cannot be ascertained, they shall have joint ownership of such knowledge and shall be entitled to use and license such knowledge without owing any financial compensation to each other. The contractors shall agree amongst themselves the allocation and terms of exercising ownership of that knowledge with a clear procedure stated in the consortium agreement (CA).
The primary means of protection will be through the use of European and global patents. The goal will be to patent and copyright the technology as early as possible in the project as commercialising under patent-pending conditions will not offer the optimal level of IPR protection.
Each partner will retain any developed IPR and the consortium will enter into joint agreements to ensure all IPR can be exploited fairly. These will be laid out in the CA. The CA will also ensure that any IP developed during this product is retained for the benefit of the European companies involved.
It has been agreed by the partners that the consortium will retain ownership of any new IPR directly arising from work related to and stemming from the Project. Full recognition will be given to any existing prior art, especially where a Consortium member owns it.
The Dissemination & Exploitation Manager, will ensure that any such relevant pre-existing IPR is given maximum visibility to all members from the beginning of the Project. He will also ensure the regular searching for new patents in the subject in order to be alert in case another team i working on very similar field. Regular patent search coupled with technology watch through the Enterprise Europe Network will ensure staying ahead of any possible competition.
Patents:
1. “Crosslinked Wholly Aromatic Polymeric Electrolyte Proton Conductors Based On Quinoline Units” K.K. Kallitsis, S.C. Kakogianni, A.K. Andreopoulou, M.K. Daletou, S.G. Neophytides, J.K. Kallitsis Greek Patent Application Appl. No. 20160100491/26.09.2016.
2. “Sposób otrzymywania katalizatora do reformingu parowego metanolu oraz katalizator otrzymywany tym sposobem (A method of production of a catalyst for the steam reforming of methanol and the catalyst produced by this method)” A. Machocki, W. Gac, M. Greluk, G. Słowik, W. Zawadzki, Polish Patent application No. P.419439 (2016).




List of Websites:
http://irmfc.iceht.forth.gr