Final Report Summary - MOREPOWER (Compact direct (m)ethanol fuel cell for portable application)
The objective of the MOREPOWER project has been the development of a low cost, low temperature, portable direct methanol fuel cell device of compact design. The device offers also limited operation on ethanol as a fuel. The potential market for portable fuel cells includes weather stations, medical devices, signal units, auxiliary power units (APUs), gas sensors and security cameras.
In order to reach the target developments in the following areas were planned and executed:
(1) new proton exchange membranes with high proton conductivity but reduced permeability of water, methanol and ethanol compared to Nafion;
(2) catalyst materials with enhanced activity for the methanol oxidation and ethanol oxidation were developed;
(3) on the cathodic side materials with enhanced oxygen reduction capacity and increased methanol tolerance were developed;
(4) the processing of MEA fabrication was optimised in order to promote efficient operation at low temperatures with practical flows and pressures;
(5) a process of system optimisation, simplification and miniaturisation were carried out.
The project was structured into work packages (WPs) as follows:
WP1: System specification and sizing
The specification in the details of the component and system characteristics (including size) for a successful prototype were established by CRF, POLITO and Nedstack. The testing protocol including membrane, MEA manufacturing, the electrochemical test and the methanol cross over analysis was defined.
WP2: Membrane development
The objective of WP2 was the development of new proton exchange materials with proton conductivity similar to Nafion117 but methanol, ethanol and water permeability tenfold lower. In order to achieve this target, organic modification by cross linking and blend development were investigated by Solvay. At GKSS inorganic modifications with sol gel technique, organically modified precursors and fillers with additionally proton conductive groups were explored.
Two membrane prototypes have been scaled up to semi-industrial reactor (35 m2). One, the 'Morgane N100-40V', has been selected as the best performing new grade:
- methanol crossover reduced down to 1/3 of Nafion, electro-osmotic water drag down to ½;
- DMFC improved performance and MEA durability of up to 2 000 h at 60 degrees Celsius demonstrated using standard electrodes;
- significantly superior mechanical properties as compared to Nafion.
At GKSS the focus was on inorganic modification of polymer exchange membranes. As baseline membranes sulfonated poly (ether ether ketone), Nafion117 and grafted membranes supplied by Solvay with different grafting levels were used. The membranes will be characterised by impedance spectroscopy, methanol and water permeability and swelling measurements, electron microscopy.
With SPEEK-mod1, a sulfonated poly (ether ether ketone) modified with polybenzimidazole modified zirconium phosphate the mid-term target of MEA performance could be reached as was shown by measurements of CNR-ITAE. The same organic treatment was applied on zirconium titanium phosphate (ZrTiPmod) leading to improved membrane properties and in preliminary FC-tests at GKSS with electrodes supplied by JM to a similar performance.
Further inorganic treatments leading to fuel cell performances similar or better than of Nafion117 were achieved for SPEEK modified with sulfonated oligomers attached to fumed silica (SPEEK-mod3).
WP3: Catalyst development and optimisation
The objective was the development of new catalyst materials with enhanced reduction capacity of oxygen at the cathode and enhanced oxidation capacity of methanol and ethanol at the anode.
Cathode: Oxygen reduction catalyst with enhanced catalytic efficiency and enhanced methanol tolerance. A colloidal preparation procedure has been developed by the CNR-ITAE for the preparation of nano-sized (1.5-3 nm) anode and cathode catalysts including alloys of Pt with transition metals, such as Fe, Cu, Co and Pt-Ru/C catalysts with crystalline structures.
Anode: Low Pt-loading and catalysts with improved methanol oxidation efficiency. For the anode Pt/Ru catalysts were optimised to reach highly effective and stable alloys. At JMFC the reduction of the Pt-loading was identified as one key technical goal for the long-term success of DMFCs.
PtSn and PtRu catalysts were screened for EtOH oxidation activity at the anode by JMFC. The PtSn catalysts oxideses ethanol at lower over-potentials than PtRu. Sn increases the catalyst lattice and helps to break the C-C ethanol bond. However, analysis of the cathode effluent (using non optimised cathode catalysts) showed it contained ethanol, acetic acid and traces of acetaldehyde clearly indicating incomplete oxidation of ethanol and significant ethanol crossover. Therefore, the cathode was severely inhibited by EtOH and / or EtOH oxidation products. It was concluded that overcoming the issues of cathode poisoning is essential to match the performance target with EtOH.
WP4: MEA preparation and characterisation
WP4 focused on the integration of the new catalysts (from WP3) into membrane electrode assembly (MEA) and the characterisation of these under appropriate fuel cell operating conditions, as well as on the integration of the new membranes developed in WP2 and optimisation of MEAs for these membranes. The MEAs were fully tested.
WP5: Modelling, design and control
The project was followed all the time with the simulation and modelling support of CRF and POLITO, which suggested component designs, predicted the influence of different materials and components like MEAs in the final performance, simulated the methanol sensor control, etc. The modelling was validated with experimental testing results of the final prototype under operation.
System modelling: the objective of system modelling was the estimation of heat and mass fluxes and pressure drops, for the integration and optimisation of the DMFC components.
CFD modelling: The objectives of CFD modelling were the optimal definition of the flow fields to build the bipolar plates. Simulations were carried out and optimised for the anode flow field only, with the simplification of neglecting the presence of CO2 gas bubbles.
Control modelling: the first goal was to define the control strategy and the related preliminary operative philosophy (POP) to allow the correct functioning of the DMFC via the methanol sensor design, developed by CRF. The main control was related to the stack power.
Dynamic modelling: The main purpose of the realisation of dynamic simulations was the analysis of the dynamic behaviour of the DMFC's equipments during the start-up / shut-down phases and during the transients involved when a changing on the electrical power demand from the DMFC takes place.
Models validation: All the developed models were validated against experimental data. The first validation run was done on simulation of single cells, Morgane and Nafion, respectively: the simulated data perfectly fit the experimental ones. The second validation run was done on the stack performance, both the 25 Morgane cells stack and the 10 Nafion cells stack, as a comparison between experimental and simulated results. For both cases, the simulation results are slightly better than the experimental results.
WP6: System development and optimisation: Flow distribution
A set of flow distribution systems of predefined outer dimensions for further testing at the assembled fuel cell available for the partners had been developed by IMM, CRF and Nedstack.
WP7: System development and optimisation: liquid management
The start-up procedure is performed by a chain of three components: a start-up burner, an evaporator and a start-up heat-exchanger. The start-up procedure consists of electrical pre-heating followed by catalytic combustion. By these means, about 1 kW of heat are generated to heat the liquid cycle. Only the power consumption of the methanol pump and of the blower is required after the pre-heating period.
All temperature and pressure measurements are implemented into the data acquisition system as well as the control loops for temperature and pump control. These loops control the fuel cell temperature, after-burner temperature, cathode stoichiometry, methanol concentration in the liquid cycle start-up burner temperature and re-cycling of condensed water from the cathode air. CRF has conceived, manufactured and tested an electrochemical amperometric sensor for sensing the methanol concentration in aqueous solutions. The device is based on the electro-oxidation of methanol under limiting current conditions. The first experimental activity was devoted to the selection of the proper electrocatalyst, among Pt and Pt/Ru supported nano-particles. The platinum based catalyst was selected because of its high 'robustness' and output current. The influence of both physical parameters, applied potential and operating temperature, and of the anode electrode structure, binder content and noble metal load, was explored.
WP8: System integration and prototype
Nedstack has produced moulded composite cell plates using the material that has been developed in WP6. Two single cell test stack were built to evaluate the selected stack design and the moulded cell plates. A 25 cell full scale stack was assembled and tested and integrated to the final prototype. The fuel cell stack tested allowed to achieve 350 W operating with 1 M methanol solution at 60 degrees Celsius. The integrated system including stack and balance of plant was assembled in a form, which allows the easy test of each component in different conditions.
In order to reach the target developments in the following areas were planned and executed:
(1) new proton exchange membranes with high proton conductivity but reduced permeability of water, methanol and ethanol compared to Nafion;
(2) catalyst materials with enhanced activity for the methanol oxidation and ethanol oxidation were developed;
(3) on the cathodic side materials with enhanced oxygen reduction capacity and increased methanol tolerance were developed;
(4) the processing of MEA fabrication was optimised in order to promote efficient operation at low temperatures with practical flows and pressures;
(5) a process of system optimisation, simplification and miniaturisation were carried out.
The project was structured into work packages (WPs) as follows:
WP1: System specification and sizing
The specification in the details of the component and system characteristics (including size) for a successful prototype were established by CRF, POLITO and Nedstack. The testing protocol including membrane, MEA manufacturing, the electrochemical test and the methanol cross over analysis was defined.
WP2: Membrane development
The objective of WP2 was the development of new proton exchange materials with proton conductivity similar to Nafion117 but methanol, ethanol and water permeability tenfold lower. In order to achieve this target, organic modification by cross linking and blend development were investigated by Solvay. At GKSS inorganic modifications with sol gel technique, organically modified precursors and fillers with additionally proton conductive groups were explored.
Two membrane prototypes have been scaled up to semi-industrial reactor (35 m2). One, the 'Morgane N100-40V', has been selected as the best performing new grade:
- methanol crossover reduced down to 1/3 of Nafion, electro-osmotic water drag down to ½;
- DMFC improved performance and MEA durability of up to 2 000 h at 60 degrees Celsius demonstrated using standard electrodes;
- significantly superior mechanical properties as compared to Nafion.
At GKSS the focus was on inorganic modification of polymer exchange membranes. As baseline membranes sulfonated poly (ether ether ketone), Nafion117 and grafted membranes supplied by Solvay with different grafting levels were used. The membranes will be characterised by impedance spectroscopy, methanol and water permeability and swelling measurements, electron microscopy.
With SPEEK-mod1, a sulfonated poly (ether ether ketone) modified with polybenzimidazole modified zirconium phosphate the mid-term target of MEA performance could be reached as was shown by measurements of CNR-ITAE. The same organic treatment was applied on zirconium titanium phosphate (ZrTiPmod) leading to improved membrane properties and in preliminary FC-tests at GKSS with electrodes supplied by JM to a similar performance.
Further inorganic treatments leading to fuel cell performances similar or better than of Nafion117 were achieved for SPEEK modified with sulfonated oligomers attached to fumed silica (SPEEK-mod3).
WP3: Catalyst development and optimisation
The objective was the development of new catalyst materials with enhanced reduction capacity of oxygen at the cathode and enhanced oxidation capacity of methanol and ethanol at the anode.
Cathode: Oxygen reduction catalyst with enhanced catalytic efficiency and enhanced methanol tolerance. A colloidal preparation procedure has been developed by the CNR-ITAE for the preparation of nano-sized (1.5-3 nm) anode and cathode catalysts including alloys of Pt with transition metals, such as Fe, Cu, Co and Pt-Ru/C catalysts with crystalline structures.
Anode: Low Pt-loading and catalysts with improved methanol oxidation efficiency. For the anode Pt/Ru catalysts were optimised to reach highly effective and stable alloys. At JMFC the reduction of the Pt-loading was identified as one key technical goal for the long-term success of DMFCs.
PtSn and PtRu catalysts were screened for EtOH oxidation activity at the anode by JMFC. The PtSn catalysts oxideses ethanol at lower over-potentials than PtRu. Sn increases the catalyst lattice and helps to break the C-C ethanol bond. However, analysis of the cathode effluent (using non optimised cathode catalysts) showed it contained ethanol, acetic acid and traces of acetaldehyde clearly indicating incomplete oxidation of ethanol and significant ethanol crossover. Therefore, the cathode was severely inhibited by EtOH and / or EtOH oxidation products. It was concluded that overcoming the issues of cathode poisoning is essential to match the performance target with EtOH.
WP4: MEA preparation and characterisation
WP4 focused on the integration of the new catalysts (from WP3) into membrane electrode assembly (MEA) and the characterisation of these under appropriate fuel cell operating conditions, as well as on the integration of the new membranes developed in WP2 and optimisation of MEAs for these membranes. The MEAs were fully tested.
WP5: Modelling, design and control
The project was followed all the time with the simulation and modelling support of CRF and POLITO, which suggested component designs, predicted the influence of different materials and components like MEAs in the final performance, simulated the methanol sensor control, etc. The modelling was validated with experimental testing results of the final prototype under operation.
System modelling: the objective of system modelling was the estimation of heat and mass fluxes and pressure drops, for the integration and optimisation of the DMFC components.
CFD modelling: The objectives of CFD modelling were the optimal definition of the flow fields to build the bipolar plates. Simulations were carried out and optimised for the anode flow field only, with the simplification of neglecting the presence of CO2 gas bubbles.
Control modelling: the first goal was to define the control strategy and the related preliminary operative philosophy (POP) to allow the correct functioning of the DMFC via the methanol sensor design, developed by CRF. The main control was related to the stack power.
Dynamic modelling: The main purpose of the realisation of dynamic simulations was the analysis of the dynamic behaviour of the DMFC's equipments during the start-up / shut-down phases and during the transients involved when a changing on the electrical power demand from the DMFC takes place.
Models validation: All the developed models were validated against experimental data. The first validation run was done on simulation of single cells, Morgane and Nafion, respectively: the simulated data perfectly fit the experimental ones. The second validation run was done on the stack performance, both the 25 Morgane cells stack and the 10 Nafion cells stack, as a comparison between experimental and simulated results. For both cases, the simulation results are slightly better than the experimental results.
WP6: System development and optimisation: Flow distribution
A set of flow distribution systems of predefined outer dimensions for further testing at the assembled fuel cell available for the partners had been developed by IMM, CRF and Nedstack.
WP7: System development and optimisation: liquid management
The start-up procedure is performed by a chain of three components: a start-up burner, an evaporator and a start-up heat-exchanger. The start-up procedure consists of electrical pre-heating followed by catalytic combustion. By these means, about 1 kW of heat are generated to heat the liquid cycle. Only the power consumption of the methanol pump and of the blower is required after the pre-heating period.
All temperature and pressure measurements are implemented into the data acquisition system as well as the control loops for temperature and pump control. These loops control the fuel cell temperature, after-burner temperature, cathode stoichiometry, methanol concentration in the liquid cycle start-up burner temperature and re-cycling of condensed water from the cathode air. CRF has conceived, manufactured and tested an electrochemical amperometric sensor for sensing the methanol concentration in aqueous solutions. The device is based on the electro-oxidation of methanol under limiting current conditions. The first experimental activity was devoted to the selection of the proper electrocatalyst, among Pt and Pt/Ru supported nano-particles. The platinum based catalyst was selected because of its high 'robustness' and output current. The influence of both physical parameters, applied potential and operating temperature, and of the anode electrode structure, binder content and noble metal load, was explored.
WP8: System integration and prototype
Nedstack has produced moulded composite cell plates using the material that has been developed in WP6. Two single cell test stack were built to evaluate the selected stack design and the moulded cell plates. A 25 cell full scale stack was assembled and tested and integrated to the final prototype. The fuel cell stack tested allowed to achieve 350 W operating with 1 M methanol solution at 60 degrees Celsius. The integrated system including stack and balance of plant was assembled in a form, which allows the easy test of each component in different conditions.