Investigation of the liquid water distributions in a 50cm2 PEM fuel cell: Effects of reactants relative humidity, current density, and cathode stoichiometry

Depósito de Investigación de la Universidad de Sevilla https://idus.us.es/ This is an Accepted Manuscript of an article published by Elsevier in Energy, Vol. 82 on March 2015, available at: https://doi.org/10.1016/j.energy.2015.01.101 © 2015 Elsevier. En idUS Licencia Creative Commons CC BY-NC-ND 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 1 Investigation of the liquid water distributions in a 50 cm2 PEM Fuel Cell: effects of reactants relative humidity, current density, and cathode stoichiometry Alfredo Iranzo a*, Pierre Boillat b, Johannes Biesdorf b, Antonio Salva a a AICIA, School of Engineering, Sevilla, Spain. b Electrochemistry Laboratory (LEC) and Neutron Imaging and Activation Group (NIAG), Paul Scherrer Institut (PSI), Villigen, Switzerland. *Corresponding author: Tel.: (+34) 954 487471. Fax: (+34) 954 463153. Email: [email protected] Abstract A 50 cm2 commercial PEM fuel cell has been used to investigate the effects of a set of different operating conditions on the resulting liquid water distributions in the cell. A comprehensive matrix of operating conditions was analysed, varying the reactants relative humidity (anode and cathode), cathode stoichiometry, and cell current density. Neutron imaging was used to determine the liquid water distributions within the cell for each operating condition. The obtained neutron radiographs were post-processed and analysed in order to assess the effects of the different operating conditions. Cell voltage and cell resistance (High Frequency Resistance) were also monitored during the experiments and included in the analysis. Overall, the well-known water distributions corresponding to serpentine flow fields were observed, featuring a progressive water accumulation along the gas flow and towards the outlet port. Cathode channels were showing water accumulation. It was found that the cathode relative humidity had a much larger effect on the cell water content and overall performance than the anode relative humidity for this particular cell. Keywords: PEM fuel cell; water distribution; water transport; Neutron Imaging 1 Introduction Among the research efforts in Polymer Electrolyte Fuel Cells (PEFCs) [1], mass transport losses are one of the main focuses as they represent the major restriction for the operation at high current densities, which is indeed required for maximizing efficiency [2] and reducing stack weight and cost. In addition to the oxygen diffusion limitations in dry GDLs, water produced by the cathode electrode at high current densities is known to block the GDL pores and electrode reaction sites, causing the cell flooding and preventing oxygen from effectively reaching the catalyst active sites [3-6]. Also an excess of liquid water or a dry-out membrane will result in significant durability issues [7,8]. In order to ensure a successful operation of the cell an appropriate water balance must be achieved, which is the objective of many research efforts in PEFCs, mainly by means of experimental research and development of computational models [4, 9-12]. It must be considered that as stated by Li et al. [5], the optimal operating conditions depend on the given cell (flow field, MEA, GDL) and therefore, any conclusion about the effects of operating conditions depends on the component specifications of the particular fuel cell [5]. Apart from modeling studies, the effects of operating conditions on water balance and cell performance have been *Manuscript Click here to download Manuscript: Manuscript_Energy_Iranzo_fuelcell_neutronimaging_revised.docxClick here to view linked References 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 2 determined through extensive experimental work, as in Bret et al [13], Hung et al. [14], Jang et al [15], Saleh et al. [16], or Carton and Olabi [17], who studied different operating conditions for a set of three flow field designs. Most of the experimental work is focused on performance characterization at the different operating conditions by means of polarization curves and other electrochemical diagnostics, as in Iranzo et al. [18]. Other techniques such as in situ diagnostics for water visualization (neutron radiography, X-ray imaging, among others) are increasingly being used [19, 20]. In this field, Hickner et al. [21] performed in situ high-resolution neutron radiography, quantifying cross-sectional liquid water profiles as a function of cell temperature, current density, and anode and cathode gas feed flow rates. They present water distributions identifying that both water diffusion from the cathode and thermal gradients play a significant role in the cell water balance of the cell. Owejan et al. [22] analysed the effect of the flow field and the GDL properties on the water accumulation of a 50 cm2 cell using neutron radiography. The effect of the current density is also presented in the results, where it is observed that the total water content not always increases with the current density, as other effects also influences the water content and distribution, as thermal gradients and gas flows. Oberholzer and Boillat [23] recently presented results of an extensive characterization of a 0.5 cm2 differential cell operating with asymmetric relative humidities (different RH values for anode and cathode) reporting results for 25 different combinations of RHs. The water distributions were visualized and quantified by high resolution in-plane imaging, and its effect on performance was analysed. The results represented a systematic analysis covering a detailed matrix of operating conditions, and it is therefore a relevant basis for the work presented here. In this work, a 50 cm2 commercial PEM fuel cell has been used to investigate the effects of a set of different operating conditions on the resulting liquid water distributions in the cell. Neutron imaging was used to determine the liquid water distributions in the cell for each operating condition. Cell voltage and cell resistance (High Frequency Resistance) were also monitored during the experiments and included in the analysis. Separate effects of anode and cathode relative humidity can therefore be assessed from the results obtained. This is presented for a medium scale fuel cell (50 cm2) operating with representative conditions. Moreover, the results presented for this parametric study include a segmentation of liquid water, where water distributions are presented for the entire cell and for the GDL/MEA only, without the contribution of the channels. 2 Experimental 2.1 Cell description The cell used in the experiment was a 50 cm2 active area cell from ElectroChem Inc., with metallic bipolar plates (five-channel serpentine flow field). The Bipolar Plates layout is cross-flow, with horizontal channels in the anode and vertical channels in the cathode. A set of Gas Diffusion Layers from SGL Group (Sigracet 24BC) were used. A Catalyst Coated Membrane (CCM) from Baltic Fuel Cells was used, with catalyst loading at anode and cathode electrodes 0.3 mg Pt/cm2 and 0.6 mg Pt/cm2 respectively. The membrane material is Nafion-117. The cell reactants are supplied through the ports in the upper right corner (Figure 1), whereas anode and cathode outlets are located in the lower left corner. The bipolar 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 3 plates are arranged in a cross-flow layout (Figure 1), where the anode gas flows in horizontal channels and the cathode gas flows in vertical channels. The cell temperature was measured in four different locations of the bipolar plates, and a cooling air flow was used to ensure that the cell was at 60 ºC (± 2 ºC) during all the experiments. 2.2 Experimental procedure The SINQ-NEUTRA beam-line at PSI [24-26] with 10 s exposure time was used for the neutron imaging experiments. Before starting each of the tests the cell was dried out by flowing dry nitrogen through anode and cathode. The cell operation was performed in galvanostatic mode setting anode and cathode flows to the desired conditions, as well as the current. The matrix of operating conditions is shown in Table 1. The cell was operated at 2.0 bar and 60 ºC. Current density, voltage, cell resistance (HFR), and neutron radiographs were logged during the entire operation. It was observed that a significant amount of liquid water was present in the channels, especially in the cathode side. In order to allow for the determination of the liquid water content distributions in GDL/MEA only, it was necessary to flush out the liquid water in the channels. This is possible by performing a sudden decompression/purging in the anode and cathode sides. This has been described in Owejan et al [27], and has been adapted as a tool for neutron imaging visualization by Boillat et al [28, 29]. The method consists of decompressing the anode and cathode channels, in order to flush out the liquid water in the channels. When this operation is performed in a short period of time, the water content in GDL/MEA is not significantly affected and therefore quantitative analysis can also be done with confidence [28]. The radiographs presented correspond to both scenarios (before and after decompressing). The image processing was done as described in a previous work [28, 29], using the Lambert-Beer law for calculating the water thickness and water content in the images. 3 Results The neutron images recorded for the different operating conditions are presented in Figure 2 to Figure 5. As general observations, liquid water clearly appears in the channels of the cathode side (vertical channels) in form of water slugs, whereas no liquid water is present in the channels of the anode side (horizontal channels). Liquid water in form of slugs within the cathode channels is actually expected for stoichiometric factors less than 5 [30] or gas velocity less than 5.0 m/s [12, 30], and the typical velocity range within the channels for the operating conditions tested was 1-2 m/s [31]. A detail analysis of the liquid water accumulation patterns was presented elsewhere [31]. Depending on the particular operating condition, a major fraction of the area of the cell is presenting channels with liquid water, where higher water contents are obtained for higher reactants relative humidity as observed in Figure 2. It is also observed that water progressively accumulates towards the outlet region of the reactants (Figure 2), as gas flow traveling along the channels progressively becomes saturated with water vapor, and therefore water removal capability decreases. This is a very wellknown behavior in serpentine flow fields [32]. Given the highly non-uniformity of the water content within the 50 cm2 area, it is expected that oxygen diffusion to the 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 4 electrode will be limited for some of the areas which are mostly blocked, negatively influencing the cell performance. 3.1 Effect of cathode relative humidity Tests 3, 6 and 9 in Table 1 were analysed in order to assess the effect of the cathode relative humidity. Figure 2 shows the related radiographs. The increase in the liquid water content in the cathode side is clearly observed, for increasing relative humidity of the cathode air inlet. The case with driest reactants (RHa/c = 35/35), presents clear differences between the membrane areas close to the inlet and to outlet. The relatively dry cathode air is drying the GDL and MEA close to the gas inlet area, whereas the areas closer to the outlet clearly presents higher water contents (and therefore membrane hydration). As the cathode air is entering the cell with progressively more humidification, this difference between inlet and outlet areas is being reduced, and water content gradients in GDL/MEA are much less pronounced in the case with RHa/c = 35/90 (Figure 2). This phenomenon is therefore clearly related to the air capacity for absorbing additional water from the cell, thus increasing its humidification. The cell resistance and voltage is also presented in Figure 1. A clear relationship between water content and cell resistance and voltage is observed, as water hydrates the membrane improving the protonic conductivity (thus reducing the cell resistance and augmenting the cell voltage). Indeed, cell resistance is a clear indicator of the MEA hydration and cell water content [33]. A decrease in the cell resistance is observed as the cathode relative humidity increases (and therefore cell water content increases). The cell voltage is significantly increased for the particular cell tested mainly due to the lower cell resistance. The impact of the cathode air relative humidity in the final cell performance is therefore relevant. The fact that cathode humidity has an important influence is a general characteristic due to the back diffusion transport of water through the membrane, which tends to equilibrate the anode and cathode flow, as reported by Saleh et al [16]. For technical cells this trend can be considered of general applicability, as it was not only reported in the work of Saleh et al [16] for Nafion-211, but it has also been shown here for the present cell with Nafion-117 (thick membrane with lower back diffusion). It is expected however that by further increasing the cathode air relative humidity a decrease in the performance will be observed, as excessive liquid water will start to hinder the oxygen diffusion in the GDL/CL towards the catalytic sites. This was however not observed in the experiments presented up to a relative humidity of 90%. It also must be considered that the results correspond to cell operation at relatively low current density, where water generation is still not high enough to induce the flooding of the GDL and cathode electrode. Operating at higher current densities would require lower cathode relative humidity for maintaining the cell performance. 3.2 Effect of anode relative humidity The neutron radiographs showing the effect of varying the anode relative humidity are presented in Figure 3, for tests 3, 4 and 5 in Table 1. It is observed that the liquid water distribution in the full cell (top) and MEA/GDL (bottom) slightly increases with increasing the anode relative humidity. This effect is however much less pronounced than observed when varying the cathode relative humidity. 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 5 The cell resistance and voltage are also shown in Figure 3. The cell resistance slightly decreases when the anode relative humidity increases but the effect is almost negligible. The cell voltage increases 2 % due to the resistance decrease obtained when operating the cell at higher anode relative humidity. The influence of the anode relative humidity in the liquid water distribution and final cell performance has therefore much less influence than the cathode relative humidity (for the particular cell tested and the particular operating conditions specified). As the results correspond to a low current density it is expected that even the lowest anode relative humidity (35% in this case) is enough for hydrating the anode side of the membrane and providing water molecules for the electro-osmotic drag mechanism. Back-diffusion mechanism transport water molecules towards the anode, even for this cell with high membrane thickness (N-117). It is expected that the effect of varying the anode gas relative humidity will be magnified when operating the cell at higher current densities, when the need for water content at the anode side for the electro-osmotic drag mechanism is higher. The differences observed for the relative importance of the anode (low influence) and cathode (high influence) humidity on the performance of the cell is consistent with other literature data, as in Saleh et al. [16]. They present results for a 25 cm2 cell with Nafion 212, where it was observed that cathode relative humidity had a more relevant impact on the cell performance, as back-diffusion provided an adequate hydration of the membrane. This is consistent with the present results obtained for the 50 cm2 cell with Nafion-117. However it is worth to mention that opposite trends were reported by Oberholzer and Boillat when performing measurements in small-scale differential cells [23]. The different behaviour is explainable by the fact that in larger cells the humidification of anode and cathode gas flows interchange water through the membrane, and the impact of the cathode gas flow is stronger due to its higher magnitude [23]. However in differential cells the flows can almost be considered infinite compared to the exchange area and do not impact the final humidity level. 3.3 Effect of current density The effect of the current density over the liquid water distributions within the cell has been analyzed and is presented in this section. The liquid water distribution in the cell and GDL/MEA with air used as cathode feed at 10 and 25A is shown in Figure 4 (tests 1 and 2 in Table 1). It is observed that the cell is drier at 25A than at 10A. This result could be in principle not expected as water generation at 25A is higher according to Faraday´s law. However, it is important to note that the air flow in the cathode is also higher at 25 A as the stoichiometry is maintained constant in both cases. In addition, the local heat generation at the cathode is higher at 25 A, resulting in a higher MEA temperature (and therefore a larger amount of the water generated in the form of vapour, and finally less liquid water within the cell for this particular case). This result was also discussed by Owejan et al. [22] where they introduced that other effects also influences the water content and distribution, as thermal gradients and gas flows. 3.4 Effect of cathode stoichiometry The effect of the cathode stoichiometry over the liquid water distribution is presented in Figure 5 (for c = 3.5 and 5, tests 1 and 8 in Table 1). Cell resistance and voltage is presented as well. The degree of humidification in the membrane is similar in both cases 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 6 according to the HFR values. However the accumulated water in the cathode channels is decreasing (slightly) when increasing the cathode stoichiometry, due to the fact that the higher cathode flow has a better water removal capability. The cell voltage is clearly higher at higher cathode stoichiometry. The voltage increases given the higher oxygen availability at the electrode for higher stoichiometry, thus reducing activation and concentration losses. It is expected that the effect of cathode stoichiometry will be higher as the current increases as the water generation in the cathode increases. For this reason, operating with high cathode stoichiometry is recommended for enhancing cell performance at high current densities. 4 Conclusions Neutron Imaging was used for determining water liquid content/distribution in a 50 cm2 PEM fuel cell with serpentine flow fields. A comprehensive matrix of operating conditions was tested in order to analyse the influence of the operating conditions on the liquid water distributions. Cell resistance and voltage were also included in the analysis. In particular, the effects of reactants relative humidity, cell current density and cathode stoichiometry were investigated. The results presented for this parametric study include a segmentation of liquid water within the cell, where water distributions are presented for the entire cell and for the GDL/MEA only, without the contribution of the channels. The liquid water patterns were identified for the different cases, where the general wellknown serpentine-type patterns were clearly observed (dry inlet and progressive water accumulation along the gas flow path towards the outlet). Severe flooding in the cathode channels were observed in the radiographs. Given the flow field design with five parallel channels per serpentine pass, with a common gas manifold, the accumulated water in some of the channels remains blocked without the possibility of being flushed out. Regarding the influence of the different operating conditions, the trends observed are summarized in Table 3. It must be noted that separate effects of anode and cathode relative humidity has been assessed from the results (with cathode RH featuring a much more significant influence than anode RH for the cell and conditions tested). In the experimental results it is clearly observed that the cathode relative humidity has a high and positive impact in the fuel cell performance (until a 90% relative humidity, the maximum tested). The water content in the MEA increases and therefore, the ohmic resistance decreases improving the final performance of the cell. This occurs at the relatively lower current density tested, but it is expected that higher current densities would require lower cathode relative humidity for maintaining the cell performance, given the higher water generation. The conclusions derived from the study are representative of a medium scale fuel cell (50 cm2) operating with a set of realistic conditions. Acknowledgements This work was carried out with the support of the European Community. We appreciate the support of the European Research Infrastructure project “H2FC European Infrastructure” (funded under the FP7 specific programme Capacities, Grant agreement Number FP7-284522) and its partner Paul Scherrer Institut. 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 7 Alfredo Iranzo also gratefully acknowledges the support of the Spanish Ministry of Economy and Competitiveness under the program INNCORPORA, Torres-Quevedo PTQ-11-04703. References [1] Neef HJ. International overview of hydrogen and fuel cell research. Energy 2009;34:327-333 [2] Li D, Yu Y, Jin Q, Gao Z. Maximum power efficiency operation and generalized predictive control of PEM (proton exchange membrane) fuel cell. Energy 2014;68:210- 217 [3] Dai W, Wang H, Yuan XZ, Martin JJ, Yang D, Qiao J, et al. A review on water balance in the membrane electrode assembly of proton exchange membrane fuel cells. Int J Hydrogen Energ 2009;34:9461-9478 [4] Jiao K, Li X. Water transport in polymer electrolyte membrane fuel cells. Prog Energ Combust 2011;37:221-291 [5] Li H, Tang Y, Wang Z, Shi Z, Wu S, Song D, et al. A review of water flooding issues in the proton exchange membrane fuel cell. J Power Sourc 2008;178:103-117 [6] Berg P, Promislow K, St Pierre J, Stumper J, Wetton B. Water management in PEM fuel cells. J Electrochem Soc 2004;151:A341-353 [7] Schmittinger W, Vahidi A. A review of the main parameters influencing long-term performance and durability of PEM fuel cells. J Power Sourc 2008;180:1-14 [8] Wu J, Yuan XZ, Martin JJ, Wang H, Zhang J, Shen J, et al. A review of PEM fuel cell durability: degradation mechanisms and mitigation strategies. J Power Sourc 2008;184:104-119 [9] Wang CY. Fundamental models for fuel cell engineering. Chem Rev 2004;104:4727-4765 [10] Siegel C. Review of computational heat and mass transfer modelling in polymerelectrolyte-membrane (PEM) fuel cells. Energ 2008;33:1331-1352 [11] Iranzo A, Boillat P, Oberholzer P, Guerra J. A novel approach coupling neutron imaging and numerical modelling for the analysis of the impact of water on fuel cell performance. Energ 2014;68: 971-981 [12] Carton JG, Lawlor V, Olabi AG, Hochenauer C, Zauner G. Water droplet accumulation and motion in PEM (Proton Exchange Membrane) fuel cell minichannels. Energ 2012;39:63-73 [13] Brett DJL, Atkins S, Brandon NP, Vasileiadis N, Vesovic V, Kucernak AR. 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 8 Membrane resistance and current distribution measurements under various operating conditions in a polymer electrolyte fuel cell. J Power Sourc 2007;172:2–13 [14] Hung AJ, Sung LY, Chen YH, Yu CC. Operation-relevant modeling of an experimental proton exchange membrane fuel cell. J Power Sourc 2007;171:728–737 [15] Jang JH, Chiu HC, Yan WM, Sun WL. Effects of operating conditions on the performances of individual cell and stack of PEM fuel cell. J Power Sourc 2008;180: 476–483 [16] Saleh MM, Okajima T, Hayase M, Kitamura F, Ohsaka T. Exploring the effects of symmetrical and asymmetrical relative humidity on the performance of H2/air PEM fuel cell at different temperatures. J Power Sourc 2007;164:503–509 [17] Carton JG, Olabi AG. Design of experiment study of the parameters that affect performance of three flow plate configurations of a proton exchange membrane fuel cell. Energy 2010;35:2796-2806 [18] Iranzo A, Muñoz M, López E, Pino J, Rosa F. Experimental fuel cell performance analysis under different operating conditions and bipolar plate designs. Int J Hydrogen Energ 2010; 35:11437-11447 [19] Tsushima S, Hirai S. In situ diagnostics for water transport in proton exchange membrane fuel cells. Prog Energ Comb Sci 2011;37:204-220 [20] Bazylak A. Liquid water visualization in PEM fuel cells: A review. Int J Hydrogen Energ 2009 34:3845-3857 [21] Hickner MA, Siegel NP, Chen KS, Hussey DS, Jacobson DL, Arif M. In situ high resolution neutron radiography of cross-sectional liquid water profiles in proton exchange membrane fuel cells. J Electrochem Soc 2008;155:B427-B434 [22] Owejan JP, Trabold TA, Jacobson DL, Arif M, Kandlikar SG. Effects of flowfield and diffusion layer properties on water accumulation in a PEM fuel cell. Int J Hydrogen Energ 2007;32:4489–4502 [23] Oberholzer P, Boillat P. Local Characterization of PEFCs by Differential Cells: Systematic Variations of Current and Asymmetric Relative Humidity. J Electrochem Soc 2014;161:F139-F152 [24] Oberholzer P, Boillat P, Kaestner A, Lehmann EH, Scherer GG, Schmidt TJ, et al. Characterizing local O2 diffusive losses in GDLs of PEFCs using simplified flow field patterns (“2D”,“1D”,“0D”). J Electrochem Soc 2013;160:F659-669 [25] Oberholzer P, Boillat P, Siegrist R, Ka¨stner A, Lehmann EH, Scherer GG, et al. Simultaneous neutron imaging of six operating PEFCs: experimental set-up and study of the MPL effect. Electrochem Comm 2012;20:67-70 4 electrode will be limited for some of the areas which are mostly blocked, negatively influencing the cell performance. 3.1 Effect of cathode relative humidity Tests 3, 6 and 9 in Table 1 were analysed in order to assess the effect of the cathode relative humidity. Figure 2 shows the related radiographs. The increase in the liquid water content in the cathode side is clearly observed, for increasing relative humidity of the cathode air inlet. The case with driest reactants (RHa/c = 35/35), presents clear differences between the membrane areas close to the inlet and to outlet. The relatively dry cathode air is drying the GDL and MEA close to the gas inlet area, whereas the areas closer to the outlet clearly presents higher water contents (and therefore membrane hydration). As the cathode air is entering the cell with progressively more humidification, this difference between inlet and outlet areas is being reduced, and water content gradients in GDL/MEA are much less pronounced in the case with RHa/c = 35/90 (Figure 2). This phenomenon is therefore clearly related to the air capacity for absorbing additional water from the cell, thus increasing its humidification. The cell resistance and voltage is also presented in Figure 1. A clear relationship between water content and cell resistance and voltage is observed, as water hydrates the membrane improving the protonic conductivity (thus reducing the cell resistance and augmenting the cell voltage). Indeed, cell resistance is a clear indicator of the MEA hydration and cell water content [33]. A decrease in the cell resistance is observed as the cathode relative humidity increases (and therefore cell water content increases). The cell voltage is significantly increased for the particular cell tested mainly due to the lower cell resistance. The impact of the cathode air relative humidity in the final cell performance is therefore relevant. The fact that cathode humidity has an important influence is a general characteristic due to the back diffusion transport of water through the membrane, which tends to equilibrate the anode and cathode flow, as reported by Saleh et al [16]. For technical cells this trend can be considered of general applicability, as it was not only reported in the work of Saleh et al [16] for Nafion-211, but it has also been shown here for the present cell with Nafion-117 (thick membrane with lower back diffusion). It is expected however that by further increasing the cathode air relative humidity a decrease in the performance will be observed, as excessive liquid water will start to hinder the oxygen diffusion in the GDL/CL towards the catalytic sites. This was however not observed in the experiments presented up to a relative humidity of 90%. It also must be considered that the results correspond to cell operation at relatively low current density, where water generation is still not high enough to induce the flooding of the GDL and cathode electrode. Operating at higher current densities would require lower cathode relative humidity for maintaining the cell performance. 3.2 Effect of anode relative humidity The neutron radiographs showing the effect of varying the anode relative humidity are presented in Figure 3, for tests 3, 4 and 5 in Table 1. It is observed that the liquid water distribution in the full cell (top) and MEA/GDL (bottom) slightly increases with increasing the anode relative humidity. This effect is however much less pronounced than observed when varying the cathode relative humidity. 5 The cell resistance and voltage are also shown in Figure 3. The cell resistance slightly decreases when the anode relative humidity increases but the effect is almost negligible. The cell voltage increases 2 % due to the resistance decrease obtained when operating the cell at higher anode relative humidity. The influence of the anode relative humidity in the liquid water distribution and final cell performance has therefore much less influence than the cathode relative humidity (for the particular cell tested and the particular operating conditions specified). As the results correspond to a low current density it is expected that even the lowest anode relative humidity (35% in this case) is enough for hydrating the anode side of the membrane and providing water molecules for the electro-osmotic drag mechanism. Back-diffusion mechanism transport water molecules towards the anode, even for this cell with high membrane thickness (N-117). It is expected that the effect of varying the anode gas relative humidity will be magnified when operating the cell at higher current densities, when the need for water content at the anode side for the electro-osmotic drag mechanism is higher. The differences observed for the relative importance of the anode (low influence) and cathode (high influence) humidity on the performance of the cell is consistent with other literature data, as in Saleh et al. [16]. They present results for a 25 cm2 cell with Nafion 212, where it was observed that cathode relative humidity had a more relevant impact on the cell performance, as back-diffusion provided an adequate hydration of the membrane. This is consistent with the present results obtained for the 50 cm2 cell with Nafion-117. However it is worth to mention that opposite trends were reported by Oberholzer and Boillat when performing measurements in small-scale differential cells [23]. The different behaviour is explainable by the fact that in larger cells the humidification of anode and cathode gas flows interchange water through the membrane, and the impact of the cathode gas flow is stronger due to its higher magnitude [23]. However in differential cells the flows can almost be considered infinite compared to the exchange area and do not impact the final humidity level. 3.3 Effect of current density The effect of the current density over the liquid water distributions within the cell has been analyzed and is presented in this section. The liquid water distribution in the cell and GDL/MEA with air used as cathode feed at 10 and 25A is shown in Figure 4 (tests 1 and 2 in Table 1). It is observed that the cell is drier at 25A than at 10A. This result could be in principle not expected as water generation at 25A is higher according to Faraday´s law. However, it is important to note that the air flow in the cathode is also higher at 25 A as the stoichiometry is maintained constant in both cases. In addition, the local heat generation at the cathode is higher at 25 A, resulting in a higher MEA temperature (and therefore a larger amount of the water generated in the form of vapour, and finally less liquid water within the cell for this particular case). This result was also discussed by Owejan et al. [22] where they introduced that other effects also influences the water content and distribution, as thermal gradients and gas flows. 3.4 Effect of cathode stoichiometry The effect of the cathode stoichiometry over the liquid water distribution is presented in Figure 5 (for c = 3.5 and 5, tests 1 and 8 in Table 1). Cell resistance and voltage is presented as well. The degree of humidification in the membrane is similar in both cases 6 according to the HFR values. However the accumulated water in the cathode channels is decreasing (slightly) when increasing the cathode stoichiometry, due to the fact that the higher cathode flow has a better water removal capability. The cell voltage is clearly higher at higher cathode stoichiometry. The voltage increases given the higher oxygen availability at the electrode for higher stoichiometry, thus reducing activation and concentration losses. It is expected that the effect of cathode stoichiometry will be higher as the current increases as the water generation in the cathode increases. For this reason, operating with high cathode stoichiometry is recommended for enhancing cell performance at high current densities. 4 Conclusions Neutron Imaging was used for determining water liquid content/distribution in a 50 cm2 PEM fuel cell with serpentine flow fields. A comprehensive matrix of operating conditions was tested in order to analyse the influence of the operating conditions on the liquid water distributions. Cell resistance and voltage were also included in the analysis. In particular, the effects of reactants relative humidity, cell current density and cathode stoichiometry were investigated. The results presented for this parametric study include a segmentation of liquid water within the cell, where water distributions are presented for the entire cell and for the GDL/MEA only, without the contribution of the channels. The liquid water patterns were identified for the different cases, where the general wellknown serpentine-type patterns were clearly observed (dry inlet and progressive water accumulation along the gas flow path towards the outlet). Severe flooding in the cathode channels were observed in the radiographs. Given the flow field design with five parallel channels per serpentine pass, with a common gas manifold, the accumulated water in some of the channels remains blocked without the possibility of being flushed out. Regarding the influence of the different operating conditions, the trends observed are summarized in Table 3. It must be noted that separate effects of anode and cathode relative humidity has been assessed from the results (with cathode RH featuring a much more significant influence than anode RH for the cell and conditions tested). In the experimental results it is clearly observed that the cathode relative humidity has a high and positive impact in the fuel cell performance (until a 90% relative humidity, the maximum tested). The water content in the MEA increases and therefore, the ohmic resistance decreases improving the final performance of the cell. This occurs at the relatively lower current density tested, but it is expected that higher current densities would require lower cathode relative humidity for maintaining the cell performance, given the higher water generation. The conclusions derived from the study are representative of a medium scale fuel cell (50 cm2) operating with a set of realistic conditions. Acknowledgements This work was carried out with the support of the European Community. We appreciate the support of the European Research Infrastructure project “H2FC European Infrastructure” (funded under the FP7 specific programme Capacities, Grant agreement Number FP7-284522) and its partner Paul Scherrer Institut. 7 Alfredo Iranzo also gratefully acknowledges the support of the Spanish Ministry of Economy and Competitiveness under the program INNCORPORA, Torres-Quevedo PTQ-11-04703. References [1] Neef HJ. International overview of hydrogen and fuel cell research. Energy 2009;34:327-333 [2] Li D, Yu Y, Jin Q, Gao Z. Maximum power efficiency operation and generalized predictive control of PEM (proton exchange membrane) fuel cell. Energy 2014;68:210- 217 [3] Dai W, Wang H, Yuan XZ, Martin JJ, Yang D, Qiao J, et al. A review on water balance in the membrane electrode assembly of proton exchange membrane fuel cells. Int J Hydrogen Energ 2009;34:9461-9478 [4] Jiao K, Li X. Water transport in polymer electrolyte membrane fuel cells. Prog Energ Combust 2011;37:221-291 [5] Li H, Tang Y, Wang Z, Shi Z, Wu S, Song D, et al. A review of water flooding issues in the proton exchange membrane fuel cell. J Power Sourc 2008;178:103-117 [6] Berg P, Promislow K, St Pierre J, Stumper J, Wetton B. Water management in PEM fuel cells. J Electrochem Soc 2004;151:A341-353 [7] Schmittinger W, Vahidi A. A review of the main parameters influencing long-term performance and durability of PEM fuel cells. J Power Sourc 2008;180:1-14 [8] Wu J, Yuan XZ, Martin JJ, Wang H, Zhang J, Shen J, et al. A review of PEM fuel cell durability: degradation mechanisms and mitigation strategies. J Power Sourc 2008;184:104-119 [9] Wang CY. Fundamental models for fuel cell engineering. Chem Rev 2004;104:4727-4765 [10] Siegel C. Review of computational heat and mass transfer modelling in polymerelectrolyte-membrane (PEM) fuel cells. Energ 2008;33:1331-1352 [11] Iranzo A, Boillat P, Oberholzer P, Guerra J. A novel approach coupling neutron imaging and numerical modelling for the analysis of the impact of water on fuel cell performance. Energ 2014;68: 971-981 [12] Carton JG, Lawlor V, Olabi AG, Hochenauer C, Zauner G. Water droplet accumulation and motion in PEM (Proton Exchange Membrane) fuel cell minichannels. Energ 2012;39:63-73 [13] Brett DJL, Atkins S, Brandon NP, Vasileiadis N, Vesovic V, Kucernak AR. 8 Membrane resistance and current distribution measurements under various operating conditions in a polymer electrolyte fuel cell. J Power Sourc 2007;172:2–13 [14] Hung AJ, Sung LY, Chen YH, Yu CC. Operation-relevant modeling of an experimental proton exchange membrane fuel cell. J Power Sourc 2007;171:728–737 [15] Jang JH, Chiu HC, Yan WM, Sun WL. Effects of operating conditions on the performances of individual cell and stack of PEM fuel cell. J Power Sourc 2008;180: 476–483 [16] Saleh MM, Okajima T, Hayase M, Kitamura F, Ohsaka T. Exploring the effects of symmetrical and asymmetrical relative humidity on the performance of H2/air PEM fuel cell at different temperatures. J Power Sourc 2007;164:503–509 [17] Carton JG, Olabi AG. Design of experiment study of the parameters that affect performance of three flow plate configurations of a proton exchange membrane fuel cell. Energy 2010;35:2796-2806 [18] Iranzo A, Muñoz M, López E, Pino J, Rosa F. Experimental fuel cell performance analysis under different operating conditions and bipolar plate designs. Int J Hydrogen Energ 2010; 35:11437-11447 [19] Tsushima S, Hirai S. In situ diagnostics for water transport in proton exchange membrane fuel cells. Prog Energ Comb Sci 2011;37:204-220 [20] Bazylak A. Liquid water visualization in PEM fuel cells: A review. Int J Hydrogen Energ 2009 34:3845-3857 [21] Hickner MA, Siegel NP, Chen KS, Hussey DS, Jacobson DL, Arif M. In situ high resolution neutron radiography of cross-sectional liquid water profiles in proton exchange membrane fuel cells. J Electrochem Soc 2008;155:B427-B434 [22] Owejan JP, Trabold TA, Jacobson DL, Arif M, Kandlikar SG. Effects of flowfield and diffusion layer properties on water accumulation in a PEM fuel cell. Int J Hydrogen Energ 2007;32:4489–4502 [23] Oberholzer P, Boillat P. Local Characterization of PEFCs by Differential Cells: Systematic Variations of Current and Asymmetric Relative Humidity. J Electrochem Soc 2014;161:F139-F152 [24] Oberholzer P, Boillat P, Kaestner A, Lehmann EH, Scherer GG, Schmidt TJ, et al. Characterizing local O2 diffusive losses in GDLs of PEFCs using simplified flow field patterns (“2D”,“1D”,“0D”). J Electrochem Soc 2013;160:F659-669 [25] Oberholzer P, Boillat P, Siegrist R, Ka¨stner A, Lehmann EH, Scherer GG, et al. Simultaneous neutron imaging of six operating PEFCs: experimental set-up and study of the MPL effect. Electrochem Comm 2012;20:67-70 9 [26] Boillat P, Frei G, Lehmann EH, Scherer GG, Wokaun A. Neutron imaging resolution improvements optimized for fuel cell applications. Electrochem Solid St Lett 2010;13:B25-27 [27] Owejan JP, Gagliardo JJ, Falta SR, Trabold TA. Accumulation and removal of liquid water in proton exchange membrane fuel cells. J Electrochem Soc 2009;156:B1475-B1483 [28] Boillat P, Iranzo, A. Biesdorf J. Layer by layer segmentation of water distribution from neutron imaging of large scale cells 2015 (under review) [29] Iranzo, A. Boillat P, Rosa F. Validation of a three dimensional PEM fuel cell CFD model using local liquid water distributions measured with neutron imaging. Int J Hydrogen Energ 2014;39:7089-7099 [30] Anderson R, Zhang L, Ding Y, Blanco M, Bi X, Wilkinson DP. A critical review of two-phase flow in gas flow channels of proton exchange membrane fuel cells. J Power Sourc 2010;195:4531–4553 [31] Iranzo A, Boillat P, Biesdorf J, Tapia E, Salva A, Guerra J. Liquid water preferential accumulation in channels of PEM fuel cells with multiple serpentine flow fields. Int J Hydrogen Energ 2014;39:15687-15695 [32] Li X, Sabir I. Review of bipolar plates in PEM fuel cells: flowfield designs. Int J Hydrogen Energ 2005;30:359-371 [33] Tajiri K, Wang CY, Tabuchi Y. Water removal from a PEFC during gas purge. Electrochim Acta 2008;53:6337–6343 10 Figure captions Figure 1: Bipolar Plate design used for the experimental tests. Figure 2: Liquid water content distributions in the cell (top) and in MEA/GDL (bottom) at 10 A, λc = 3.5 and λa = 1.5, 60 ºC, 2 bar with H2/Air. Left: RHa 35%, RHc 35%. Middle: RHa 35%, RHc 60%. Right: RHa 35%, RHc 90%. Cell resistance and voltage displayed in the bottom part. Figure 3: Liquid water content distributions in the cell (top) and in MEA/GDL (bottom) at 10 A, λc = 3.5 and λa = 1.5, 60 ºC, 2 bar with H2/Air. Left: RHa 35%, RHc 35%. Middle: RHa 60%, RHc 35%. Right: RHa 90%, RHc 35%. Cell resistance and voltage displayed in the bottom part. Figure 4: Liquid water content distributions in the cell (top) and in MEA/GDL (bottom) at RHa = RHc = 90% (60ºC, 2 bar) with air feed. Left: I = 10 A. Right: I = 25 A. Cell resistance and voltage displayed in the bottom part. Figure 5: Liquid water content in the cell (top) and in MEA/GDL (bottom) at 10 A with RHa = RHc = 90% (60ºC, 2 bar) with air feed. Left: λc = 3.5. Right: λc = 5.0. Cell resistance and voltage displayed in the bottom part. 11 Table captions: Table 1: Operating conditions used in the experiment. Table 2: Liquid water content within the cell for the conditions analysed. Table 3: Summary of the influence observed at different operating conditions. Figure 01 Click here to download high resolution image Figure 02 Click here to download high resolution image