Performance evaluation of a proof-of-concept 70 W internal reforming methanol fuel cell system

Performance evaluation of a proof-of-concept 70 W internal reforming methanol fuel cell system

Journal of Power Sources 307 (2016) 875e882 Contents lists available at ScienceDirect Journal of Power Sources journal homepage: www.elsevier.com/lo...

2MB Sizes 95 Downloads 73 Views

Journal of Power Sources 307 (2016) 875e882

Contents lists available at ScienceDirect

Journal of Power Sources journal homepage: www.elsevier.com/locate/jpowsour

Performance evaluation of a proof-of-concept 70 W internal reforming methanol fuel cell system G. Avgouropoulos a, b, *, S. Schlicker c, K.-P. Schelhaas c, J. Papavasiliou a, K.D. Papadimitriou a, d, E. Theodorakopoulou e, N. Gourdoupi e, A. Machocki f, T. Ioannides a, J.K. Kallitsis a, d, G. Kolb c, **, S. Neophytides a a

Foundation for Research and Technology-Hellas (FORTH), Institute of Chemical Engineering Sciences (ICE-HT), P.O. Box 1414, GR-26504 Patras, Greece Department of Materials Science, University of Patras, GR-26504 Patras, Greece Fraunhofer ICT-IMM, Carl-Zeiss-Str. 18e20, D-55129 Mainz, Germany d Department of Chemistry, University of Patras, GR-26504 Patras, Greece e Advent Technologies SA, Patras Science Park, GR-26504 Patras, Greece f Marie Curie-Sklodowska University, Faculty of Chemistry, Department of Chemical Technology, Maria Curie-Sklodowska Sq. 3, 20-031 Lublin, Poland b c

h i g h l i g h t s

g r a p h i c a l a b s t r a c t

 Proof-of-concept 70 W internal reforming methanol fuel cell system (IRMFC).  Advent TPS® membrane electrode assemblies employed for fuel cell operation at 200  C.  Anode electrocatalyst indirectly adjoined to CuMnAlOx/Cu foam (methanol reformer).  In-situ conversion of methanol/steam to the required H2 (internal reforming).  Operation of the IRMFC supported through a number of BoP components.

a r t i c l e i n f o

a b s t r a c t

Article history: Received 12 August 2015 Received in revised form 7 January 2016 Accepted 8 January 2016 Available online xxx

A proof-of-concept 70 W Internal Reforming Methanol Fuel Cell (IRMFC) stack including Balance-of-Plant (BoP) was designed, assembled and tested. Advent TPS® high-temperature, polymer electrolyte membrane electrode assemblies were employed for fuel cell operation at 200  C. In order to avoid phosphoric acid poisoning of the reformer, the anode electrocatalyst of each cell was indirectly adjoined, via a separation plate, to a highly active CuMnAlOx catalyst coated onto copper foam, which served as methanol reforming layer. The reformer was in-situ converting the methanol/steam feed to the required hydrogen (internal reforming concept) at 200  C, which was readily oxidized at the anode electrodes. The operation of the IRMFC was supported through a number of BoP components consisting of a start-up subsystem (air blower, evaporator and monolithic burner), a combined afterburner/evaporator device, methanol/water supply and data acquisition units (reactants/products analysis, temperature control, flow control, system load/output control). Depending on the composition of the liquid MeOH/H2O feed

Keywords: Hydrogen Steam reforming Internal reforming High temperature PEM fuel cell

* Corresponding author. Department of Materials Science, University of Patras, GR-26504 Patras, Greece. ** Corresponding author. E-mail addresses: [email protected] (G. Avgouropoulos), Gunther.Kolb@imm. fraunhofer.de (G. Kolb). http://dx.doi.org/10.1016/j.jpowsour.2016.01.029 0378-7753/© 2016 Elsevier B.V. All rights reserved.

876 Methanol reforming catalyst Balance-of-plant

G. Avgouropoulos et al. / Journal of Power Sources 307 (2016) 875e882

streams, current densities up to 0.18 A cm2 and power output up to 70 W could be obtained with remarkable repeatability. Specific targets for improvement of the efficiency were identified. © 2016 Elsevier B.V. All rights reserved.

1. Introduction The low energy densities of current battery systems contribute to the excessive weight and bulkiness of portable equipment and severely limit the capabilities of military or commercial operations employing portable electronic devices. While much progress has been made in battery technology, this electric storage device still has limited performance as compared to advanced H2 fuel cells, which could offer higher energy densities. In the case of methanol being the hydrogen source for the fuel cell, even higher energy densities than pressurized H2 can be obtained, making it an attractive option for advanced portable power applications. Methanol, as a hydrogen carrier, offers the advantages of low sulfur content, easy availability, production from biomass, safe handling/ storage, relatively low reforming temperatures, and low CO formation. A fuel cell supplied with methanol would be the preferred energy source for a portable power system if a rugged, reliable and lightweight fuel processor was available to efficiently convert the alcohol to H2 on-demand and drive a polymer electrolyte membrane (PEM) fuel cell, all contained in a small volume. Recently, Avgouropoulos et al. [1e5] demonstrated the functionality of an Internal Reforming Methanol Fuel Cell (IRMFC), where the methanol gets reformed by a catalyst incorporated into the anode compartment of a PEM fuel cell (internal reforming). Integration of the reformer into the fuel cell eliminates the need for additional heat exchangers and the adopted high-temperature PEM technology does not require CO-free hydrogen fuel. Thus, the design of the fuel processor-fuel cell system offers room for simplification, increase of efficiency and minimization of system weight and volume [3,6,7]. Increasing source runtime, speeding up the transient response while minimizing weight, volume and cost of the power supply system are key requirements for portable applications. In the standard IRMFC concept, the anode electrocatalyst is directly adjoined to a foam-based methanol reformer. The operation characteristics of the fuel cell, however, change significantly in the presence of small amounts of unconverted methanol in the H2 stream due to the poisoning effect of adsorbed methanol on the electrocatalyst and the proton conductivity of the membrane [8e10]. The second drawback of such a configuration refers to the dissolution of the reforming catalyst into H3PO4 released from the membrane electrode assembly (MEA) leading also to corrosion of metal substrates anchoring the reforming catalyst [5,11]. Despite the fact that Advent membrane electrode assemblies (MEAs) contain lower total amount of acid as compared to other PBI-based systems, phosphoric acid leaching from the MEA to the catalyst bed is of primary concern. An alternative approach for prevention of catalyst reaction with H3PO4 is the placement of a thin graphite plate between the reforming catalyst and the electro-catalyst, which will prevent the corrosion of the reforming catalyst while allowing current collection [4,5]. Preliminary results obtained from a single-cell laboratory prototype demonstrated the improved functionality of a unit incorporating a separation plate. Indeed, promising electrochemical performance was obtained in the first 24 h, during which the required H2 for achieving 580 mV at 0.2 A cm2 was supplied by the reformer [5].

In this paper, a proof-of-concept 70 W IRMFC system including BoP was designed, assembled and tested. The heart of the fuel cell stack is the reformer-electrode-membrane assembly. The hightemperature, membrane electrode assembly (HT-MEA) has been developed by ADVENT Technologies S.A. and the University of Patras [1e3,12e14]. Advent's original high temperature membrane is based on pyridine type structures incorporated around a stable polymer backbone. MEAs with this polymer, known as Advent TPS® MEAs, operate from 120 to 200  C e the highest for this class of MEA. The anode side of each cell has a bifunctional electro/ reforming anode, which consists of two layers. The function of the first layer is to reform methanol, while the function of the second layer is to catalyze H2 oxidation. More specifically, the second layer contains a Pt-based catalyst and is indirectly adjoined via a separation plate, to the first layer, i.e. a highly active CuMnAlOx/Cu-foam methanol reformer, which provides the required flow rate of hydrogen for fuel cell operation at 200  C. The integrated system concept and the results of sub-system tests, including IRMFC stack performance under operation with methanol/water feed streams are reported and discussed along with relevant technological aspects for portable applications. 2. System design An 100 W IRMFC system with integrated fuel cell stack and balance-of-plant was initially designed employing ASPEN software. A high-temperature PEM fuel cell model developed previously was used [15,16], while the kinetic parameters of methanol steam reforming reaction estimated in a previous report [17], were also employed in the simulation. 2.1. ASPEN model of the system Fig. 1 shows a global static ASPEN model of the integrated 100 W fuel cell stack including BoP. The off-load voltage of the fuel cell results solely from the thermodynamic properties of the reactants. When power is extracted from the fuel cell, the voltage is reduced by over-potential resulting from the activation of chemical reactions (hact), ohmic resistance over the membrane (hohm), decreasing concentration of reactants (hconc), surface coverage by hydrogen (hanode, which depends on the carbon monoxide content of the reformate):

Ucell ¼ U0  hact  hohm  hconc  hanode   RT i þ i0 Rconc i ln  Rohm i  4aF l1 i0   RT i  sinh1 aF 2keh QH2 ðQco Þ

(1)

Ucell ¼ U0 

(2)

a ¼ a0 T þ b0 ; Rohm ¼ a1 T þ b1 ; Rconc ¼ a2 T þ b2 ; i0 ¼ a3 eb3 T

(3)

The model was validated by measurements of a high temperature PEM fuel cell operating on pure hydrogen [18]. In the model, the integrated fuel cell stack and reformer are separated into Fuel

G. Avgouropoulos et al. / Journal of Power Sources 307 (2016) 875e882

877

Fig. 1. ASPEN model of the integrated 130 W fuel cell stack including Balance-of-Plant (BoP).

Cell Anode (FCA), Fuel Cell Cathode (FCC), Steam Reformer (STR) and cooling channels, which were implemented as HX2. The fuel cell stack was designed for an electric power output of 130 W, 30 W of which are dedicated to balance-of-plant components. Direct cooling with air through fins attached to the fuel cell stack was implemented for the sake of system simplicity. 81.7 W of heat generated by the fuel cell is removed by cooling, while 23.4 W of heat will be consumed by the endothermic steam reforming reaction. The evaporation and super-heating of the methanol/water mixture needs to be performed outside the stack in an evaporator (EVA) and the 53.8 W of heat required for evaporation are supplied by combustion of the anode off-gas in the afterburner (AFB) implemented as BURN, which was integrated into the evaporator through plate heat-exchanger technology. The cathode air is preheated in heat-exchanger HX1 utilizing the cathode-off-gas heat (19.3 W). Feed pre-heating (ECO) upstream the evaporator is not required in the current system, although included into the model. 2.2. CFD modeling of the integrated reformer To determine the optimal geometry of the bipolar plate/foam couples, different geometries were investigated by CFD simulations such as tree-like structures with even channels in the bipolar plate and with channels of decreasing depth in the bipolar plate (not shown here), parallel channels in the foam (not shown here), trapezoidal channels in the foam and also a comb-like distribution of the feed in an unstructured foam. Details of the CFD modeling results together with the synthesis procedure of the methanol reformer are provided in the Supplementary Material. 2.3. Reforming catalyst 225 g of CuMnAlOx methanol reforming catalyst was coated onto 15 copper metal foams in total [19]. Ex-situ preactivation of the reformers was carried out at 260  C for 3 h under a 20% MeOH/ 30% H2O/He mixture at a flow rate of 50 cm3 min1 followed by overnight passivation at room temperature under 2 cm3 min1 flow of air. Analysis of reactants/products of the anode side of the stack

was carried out by on-line gas chromatography (micro GC, Varian CP-4900).

2.4. Bipolar plates The design of bipolar plates was driven by the requirements of the reforming reaction. The reactor volume is dictated by the required volume of the catalyst carrier material (Cu-foam) to support the respective catalyst amount for the target performance. The gas volume increase during the reforming reaction was also taken into consideration. The reactor volume was adjusted so as to achieve the appropriate contact time of 6.69e12.43 g s cm3, (depending on the feed flow rate corresponding to complete methanol conversion). The fuel cell stack was manufactured applying graphite composite bipolar and monopolar plates (raw material was provided by Nedstack, BV) because of their easier machining and higher corrosion resistance as compared to metallic plates. The bipolar plates comprise the catalytic reformer and the anodic and cathodic monopolar plates including their corresponding flow fields. As already described above, the CuMnAlOx reforming catalyst is coated onto Cu foams and encapsulated in the cathodic monopolar plate which is covered by the anodic separation/monopolar plate (Fig. S4 in Supplementary Material). The anodic monopolar plate separates the reformer from the MEA in order to avoid the poisoning effect of H3PO4. The reformer side of the bipolar plates has a typical pocket type arrangement incorporating the foambased reformer (foam dimensions: 111 mm  50 mm x 5 mm). The other side of the bipolar plate (cathode side) was machined with a zig-zag flowfield (active area: 107  46 mm2; channel height: 1 mm, width: 1 mm, separator: 1 mm). The anodic separation plate which separates the reformer from the MEA has a triple serpentine flowfield (active area: 107  46 mm2; channel height: 1 mm, width: 1 mm, separator: 1 mm). Advent TPS® MEA with an active area of 109  48 mm2 was sandwiched between the protection and cathode side of the bipolar plates. The MeOH/H2O feedstream enters the reformer plate and H2rich reformate is produced over the catalyst-coated foam.

878

G. Avgouropoulos et al. / Journal of Power Sources 307 (2016) 875e882

Fig. 2. Simplified cross-section view of the fuel cell stack arrangement (left); integrated fuel cell stack with 15 MEAs and 15 reformers (right).

Thereafter the reformate travels through the serpentine flowfield facing the anode electrocatalyst. Air-cooling is obtained through fins machined at the long sides of each plate (Fig. S4 in Supplementary Material), while stack pre-heating is performed with the hot combustion gases flowing over the same fins. According to this configuration, 14 bipolar plates, 2 end plates (one for the reformer and one for the cathode side) and 15 protection plates were constructed and integrated into the fuel cell module. 2.5. Membrane electrode assemblies (MEAs) The present 70 W fuel cell stack includes 15 ADVENT TPS® MEAs. A detailed description of the MEA preparation procedure has been reported previously [4,6,7]. The MEA has the following characteristics: (i) PPy(60)coT(40)S copolymer film of 180e200 mm thickness with 180e200 wt.%. H3PO4 doping level (on copolymer film weight basis), (ii) 50 cm2 electrode active area with a platinum loading of 1.5 mg Pt cm2 and a ratio of 2 g H3PO4 per g of Pt. MEAs were fabricated by hot pressing the Pt/C electrodes onto the acid-doped, polymer electrolyte membrane at 150  C for 15 min in a die setup using PTFE and FEP gaskets to achieve the appropriate compression and sealing in the single cell. 2.6. Fuel cell stack The fuel cell stack consists of unit cells of reformer, anode,

membrane and cathode connected in series through bipolar (reformer and cathode flowfield arrangements) and monopolar (separation plate with anode flowfield arrangement) plates (Fig. 2 and Fig. S4 in Supplementary Material). The plates also serve as gas distributors for methanol/water, hydrogen reformate and air that are fed to the anode (reformer and anode electrocatalyst) and cathode compartments, respectively. The integrated stack (Fig. 2) was rather bulky with dimensions of 25 cm H x 15 cm W x 20 cm L with a total weight of 7 kg (excluding the stainless steel compression bars). This resulted in a high energy demand during start-up. However, the demonstration of the functionality of a proof-ofconcept system was the primary concern of this work. 2.7. Balance-of-plant (BoP) Taking into account all the aforementioned parameters the design focused on a proof-of-concept bread-board system. A CAD model of this system was set up as shown in Fig. 3. It contains the start-up subsystem, the fuel cell stack, the methanol/water dosing and evaporation and the integrated afterburner. Applying the results from ASPEN simulation, the BoP components were designed and built. Fig. S5 in Supplementary Material shows the heat-exchanger (HX1) and the evaporator/afterburner. HX-1 was used to pre-heat the cathode air as explained in Section 2.1. It is composed of 316 TI stainless steel plates with dimensions 105 mm L x 22 mm W, 20 channels of

Fig. 3. CAD model of the integrated system; start-up subsystem is on the left; fuel cell stack in the centre; methanol/water dosing, evaporation and integrated afterburner on the right.

G. Avgouropoulos et al. / Journal of Power Sources 307 (2016) 875e882 Table 1 Electric power consumption of BoP components. Start-up

Power consumption (W)

Methanol pump Air blower Evaporator Total Normal operation Methanol/water pump Cathode compressor Air blower Total

3 15 24 42 3 10.5 15 28.5

879

stream supplied by the blower described above. The combustion gases then flow over the fins, which are also used for stack cooling during normal operation and through the EVA/AFB. The total power consumption of the balance-of-plant components for this system design is summarised in Table 1. The start-up power demand amounts to about 42 W, while the heat provided by the catalytic combustion amounts to more than 800 W. 2.8. System integration

0.5 mm W x 0.25 mm H x 80 mm L each. Due to the high pressure drop of the fuel cell cathode, the cathode air is not supplied by the radial blower described above, but rather by a small diaphragm gas pump (Thomas Gardner 1420DP BLDC) with a power consumption of 10.5 W. The evaporator/afterburner is employed for the combustion of unconverted hydrogen in the anode exhaust and supplies the required heat to evaporate the reformer feed during normal operation as explained in Section 2.1. It consists of two evaporator plates with 260 evaporation channels with dimensions: 0.03 mm W x 0.03 mm H x 40 mm L. The combustion air for the afterburner is provided by a blower, which is also used during system start-up as described below. For the system start-up, combustion of methanol in a monolithic start-up burner was applied. To introduce the methanol into this burner along with the required air flow (supplied by a blower, CPAP ECI 30.20 from EBM Papst) as gaseous feed, an evaporator was used, which was heated by silicone rubber heating foil (Minco HR5583R5.2L12A). It consists of 2 plates with 30 mm W x 60 mm L and carried 480 channels of 0.03 mm W, 0.03 mm H and 40 mm L for the evaporation process. The monolithic burner has a diameter of 50 mm and a length of 50.8 mm with a cell density of 500 cpsi. It was also pre-heated to 60  C by an integrated electric heater. The methanol/water mixture is dosed into the system by a micro-gear pump (HNP, mzr®-2921). The system start-up (S/U) is performed through the monolithic methanol burner, which is supplied with methanol vapour generated in a second, small evaporator heated by electricity. The methanol vapour is then mixed into a cold air

System integration combines the start-up subsystem, the feed evaporator and the fuel cell stack. The performance of the fuel cell stack could be monitored with a multimeter via electrical connections at each individual cell plate. During start-up, the methanol/water-mixture could be heated by the hot combustion gases produced by the start-up burner, while the hot combustion gases of the anode off-gas burner provide the heat required for evaporation of the feed during normal operation. The fins on the bipolar plates allowed the heating (during start-up) and cooling (during operation) of the fuel cell stack. Leak tests and Open Circuit Voltage (OCV) measurements utilizing hydrogen as fuel were carried out before final system integration and testing. An OCV of 14.3 V (mean value of 0.920e0.970 V for each of the 15 cells) was measured at room temperature. On-line sampling/analysis of the fuel cell anode off-gas was carried out by micro-GC. A nitrogen flow was used to dilute higher concentrations of methanol and water to avoid condensation in the analysis instrument. 3. Results and discussion The sub-system of methanol evaporator, air blower and monolithic burner was tested separately. As shown in Fig. 4, 200 L min1 of air could be pre-heated by a methanol flow of 2 ml min1 to a final temperature of more than 200  C within less than 5 min. The electrical pre-heating power required for the burner amounted to only 71 W for 45 s duration. The performance of the cathode heatexchanger was also tested separately. The heat flow of 19.3 W from the hot to the cold side of the heat-exchanger, which had been calculated for a hot gas temperature of 200  C was almost achieved, which is satisfactory considering the heat losses of such a heatexchanger of the smallest scale.

Fig. 4. Start-up (S/U) burner testing.

880

G. Avgouropoulos et al. / Journal of Power Sources 307 (2016) 875e882

Fig. 5. IRMFC e S/U and thermal management.

densities up to 0.15 A cm2 with stoichiometric ratio l ¼ 1.2. Higher current densities were obtained with even higher feed flow rates, though accompanied with lower methanol conversion (<80%). This had a detrimental effect on MEA performance, due to the high partial pressure of unconverted methanol (>3e4 kPa) [8,9]. Thus, in order to avoid rapid MEA deactivation by unconverted methanol, the system was operated at lower feed flow rates to limit the partial pressure of unconverted methanol below 3e4 kPa. The H2 IeV curves recorded before and after the measurements under methanol/water mixtures show no degradation of the stack and the MEAs under the short term test. In general, the OCV, as well as the performance of the stack under methanol/water mixture, appear to be lower than those under H2 flow. Such behavior can be attributed to the lower partial pressure of H2 and the higher partial pressure of

15

2,5

14 12

2

2,0

11 9

1,5

8 6

1,0 Anode feed streams: Feed1: 1500 ml/min H

5

Feed2: 1.083 ml /min 35% MeOH/65% H O

3

Feed3: 1.75 ml /min 35% MeOH/65% H O

2 0 0,00

0,5

Power Density, W/cm

Fuel Cell Stack Voltage, V

As shown in Fig. 5, the start-up time for the fuel cell to reach the operating temperature was relatively long (exceeding 90 min) due to the high inertia mass of the fuel cell stack. The power output of the catalytic burner ranged between 800 and 1000 W at a burner outlet temperature between 260  C and 240  C. The temperature difference along the stack as measured at the bipolar plates varied in the range 150e180  C. At this temperature, hydrogen was introduced and the stack was set in operation by allowing a current flow of 5 A until the average stack temperature had risen to 200  C. The stack remained stable at this temperature for time periods of more than 10 h. The recorded IeV curves (averaged over time at 200  C) are shown in Fig. 6. The average stack performance is within the single MEA performance as specified by ADVENT indicating good electrical characteristics along the reformers and in general an adequate functionality of stack design. Thereafter, the stack was switched off and the following day was switched on again following the same procedure. This was done for three cycles within three days. The internal methanol-steam reforming was performed with a liquid methanol/water mixture at a volumetric flow rate of 1.083 ml min1 at a steam to methanol ratio of 1.86. As shown in Fig. 7, the methanol conversion reached approximately 98%, with a carbon dioxide selectivity of 99%, while the hydrogen molar fraction in the reformate gas as amounted to 61%. At a cathode air flow rate of 7500 ml min1 and an applied current of 4 A, the hydrogen utilization at the MEA was between 60% and 75% (Fig. 7). During this test, the power output of the IRMFC system amounted to 40 W. The composition of the anode off-gas was 38.3% hydrogen, 22.8% steam and 37.7% carbon dioxide under these conditions. Other carbon containing species in the off-gas were methane (0.25%), carbon monoxide (0.2%) and unconverted methanol (0.75%). Methanol conversion higher than 95% could be obtained at 200  C employing higher liquid feed flow rates up to 1.75 ml min1. The produced hydrogen allowed operation of the fuel cell at current

switch to Feed1 after 3 days Cathode feed stream: 7500 ml/min air 0,05

0,10

0,15

0,20

0,25

0,0 0,30

2

Current Density, A/cm

Fig. 6. Performance of IRMFC power system (15 MEAs; active area 50 cm2/cell; 15 foam based reformers with 15 g CuMnAlOx loaded on each foam) operated at 180e200  C in a 3-day period.

G. Avgouropoulos et al. / Journal of Power Sources 307 (2016) 875e882

881

Fig. 7. IRMFC-System under operating conditions in a thermally stable state, 4 A applied current and 40 W power output.

steam in the reformate gas mixture. Nevertheless, prolonged operation under methanol/water mixtures results in a severe decrease of the performance of some cells as shown in Fig. 8. This can be attributed either to incomplete conversion of methanol due to the uneven flow distribution in some of the reformer foams or to the very high content of steam (>38 kPa) in the reformate gas at the inlet and along the anode [10], causing significant decrease in MEA performance at higher current densities. The performance loss is reversible (Fig. 6, switch to Feed1 after 3 days) and is caused by the restructuring and shrinkage of the phosphoric acid electrochemical interface upon the absorption of steam in phosphoric acid [10]. This negative behavior of the MEAs 1000 800 600

Cell Voltage, mV

400 200 0 0,00 1000 800 600 400 200 0 0,00

200 C (3rd day of operation) Anode: 1.266 ml /min 35%MeOH/65%H O (97% MeOH conversion) Cathode: 7500 ml/min air

0,02

C1 C2 C3 C4 C5 C6 C7 C8 C9 C10 C11 C12 C13 C14 C15

0,04

0,06

0,08

0,10

0,12

0,14

can be encountered by controlling the amount of phosphoric acid distributed within the catalytic layer of the anode electrode, as described in Ref. [10]. 4. Conclusions To the authors' knowledge, this is the first time that a High Temperature PEM fuel cell stack is operated in combination with a reformer integrated into the stack itself. The stability of the system and especially of the MEAs that operated at elevated temperatures, was remarkable. The results showed promising methanol conversion and hydrogen utilization in the reformer and fuel cell, respectively. However, the stack design requires further optimization to reduce its volume, weight and consequently start-up time so that it can be used for portable applications. The durability of the combined operation of reformer catalyst and fuel cell MEA needs to be demonstrated (experiments in progress). Indeed, voluminous and heavy foam-based reformers may be replaced with ultrathin (<1 mm) designs with significantly reduced reforming catalyst loading and new flow distribution design [4]. This will be combined with ultrathin and lightweight metallic bipolar plates aiming to drastic reduction of the volume and weight of the stack over one order of magnitude. Depending on the liquid feed flow rate, current densities up to 0.18 A cm2 and a power output up to 70 W could be obtained with remarkable repeatability in the internal reforming methanol fuel cell system. The integrated power system concept (fuel cell stack and Balance-of-Plant) and the results of sub-systems tests demonstrated the functionality of the whole system. Acknowledgment Financial support from The Fuel Cells and Hydrogen Joint Undertaking (FCH JU; Area SP1-JTI-FCH-2009.4.2; Grant agreements no. 245202 and no. 325358) is gratefully acknowledged. Appendix A. Supplementary data

switch to pure H (800 ml/min) after 4 h (simulating the produced H from methanol reformer)

0,02

0,04

0,06

0,08

0,10

0,12

0,14

Supplementary data related to this article can be found at http:// dx.doi.org/10.1016/j.jpowsour.2016.01.029.

2

Current Density, A/cm

Fig. 8. Polarization curves of 15 cells of IRMFC power system (15 MEAs; active area 50 cm2/cell; 15 foam based reformers with 15 g CuMnAlOx loaded on each foam) operated at 200  C.

Nomenclature Ucell U0

cell voltage open circuit voltage

882

G. Avgouropoulos et al. / Journal of Power Sources 307 (2016) 875e882

hact hohm hconc hanode

activation related voltage loss ohmic voltage loss voltage loss by concentration anode voltage loss i current density T temperature R universal gas constant Rohm, Rconc electrical resistance F Faraday constant QH2 hydrogen coverage of the anode l cathode stoichiometric ratio References [1] G. Avgouropoulos, J. Papavasiliou, M.K. Daletou, J.K. Kallitsis, T. Ioannides, S. Neophytides, Appl. Catal. B Environ. 90 (2009) 628e632. [2] G. Avgouropoulos, J. Papavasiliou, M. Daletou, M. Geormezi, N. Triantafyllopoulos, T. Ioannides, J. Kallitsis, S. Neophytides, US Patent 200861/ 095779, September 10 (2008). [3] G. Avgouropoulos, T. Ioannides, J.K. Kallitsis, S. Neophytides, Chem. Eng. J. 176e177 (2011) 95e101. [4] G. Avgouropoulos, A. Paxinou, S. Neophytides, Int. J. Hydrogen Energy 39

(2014) 18103e18108. [5] G. Avgouropoulos, J. Papavasiliou, T. Ioannides, S. Neophytides, J. Power Sources 296 (2015) 335e343. [6] Samms, Savinell, J. Power Sources 112 (2002) 13e29. [7] T. Derieth, G. Bandlamudi, P. Beckhaus, C. Kreuz, F. Mahlendorf, A. Heinzel, J. New Mater. Electrochem. Syst 11 (2008) 21e29. [8] S.S. Araya, I.F. Grigoras, F. Zhou, S.J. Andreasen, S.K. Kær, Int. J. Hydrogen Energy 39 (2014) 18343e18350. [9] G. Avgouropoulos, S. Neophytides, J. Appl. Electrochem. 42 (2012) 719e726. [10] M. Geormezi, F. Paloukis, A. Orfanidi, N. Shroti, M.K. Daletou, S.G. Neophytides, J. Power Sources 285 (2015) 499e509. [11] J. Papavasiliou, G. Avgouropoulos, T. Ioannides, Int. J. Hydrogen Energy 37 (2012) 16739e16747. [12] M.K. Daletou, N. Gourdoupi, J.K. Kallitsis, J. Membr. Sci. 252 (2005) 115e122. [13] K.D. Papadimitriou, M. Geormezi, S.G. Neophytides, J.K. Kallitsis, J. Membr. Sci. 433 (2013) 1e9. [14] N. Gourdoupi, K. Andreopoulou, V. Deimede, J.K. Kallitsis, Chem. Mater. 15 (2003) 5044e5050. [15] G. Kolb, S. Keller, D. Tiemann, K.P. Schelhaas, J. Schuerer, O. Wiborg, Chem. Eng. J. 207e208 (2012) 388e402. [16] A.R. Koorsgard, M.P. Nielsen, K.K. Soeren, Int. J. Hydrogen Energy 33 (2008) 1909e1920. [17] J. Papavasiliou, G. Avgouropoulos, T. Ioannides, J. Catal. 251 (2007) 7e20. [18] E.U. Ubong, Z. Shi, X. Wang, J. Electrochem. Soc. 156 (2009) B1276eB1282. [19] J. Papavasiliou, G. Avgouropoulos, T. Ioannides, Appl. Catal. B Environ. 66 (2006) 168e174.