• Energy Research
• 3. Good health close
• US close
• Journal of Power Sources close

Abstract When a lithium ion polymer battery (LiPB) is being cycled, one major cause for degradations is the irreversible side reactions between ions and solvent of electrolyte taking place at the surface of anode particles. SEM analysis of cycled battery cells has revealed that the deposits from the side reactions are dispersed not only on particles, but also between the composite anode and the separator. Thus, the solid electrolyte interface (SEI) becomes thicker and extra deposit layers are formed between composite anode and separator. Also, XPS analysis showed that the deposits are composed of Li 2 CO 3 , which is ionic conductive and electronic nonconductive. Based on the mechanisms and findings, we identified four degradation parameters, including volume fraction of accessible active anode, SEI resistance, resistance of deposit layer and diffusion coefficient of electrolyte, to describe capacity and power fade caused by the side reactions. These degradation parameters have been incorporated into an electrochemical thermal model that has been previously developed. The terminal voltage and capacity of the integrated model are compared with experimental data obtained for up to 300 cycles. Finally, the resistance of the deposit layer calculated by the model is validated against the thickness of the deposit layer measured by SEM.

Abstract Structural changes and their relationship with thermal stability of charged Li 0.33 Ni 1/3 Co 1/3 Mn 1/3 O 2 cathode samples have been studied using time-resolved X-ray diffraction (TR-XRD) in a wide temperature from 25 to 600 °C with and without the presence of electrolyte in comparison with Li 0.27 Ni 0.8 Co 0.15 Al 0.05 O 2 cathodes. Unique phase transition behavior during heating is found for the Li 0.33 Ni 1/3 Co 1/3 Mn 1/3 O 2 cathode samples: when no electrolyte is present, the initial layered structure changes first to a LiM 2 O 4 -type spinel, and then to a M 3 O 4 -type spinel and remains in this structure up to 600 °C. For the Li 0.33 Ni 1/3 Co 1/3 Mn 1/3 O 2 cathode sample with electrolyte, additional phase transition from the M 3 O 4 -type spinel to the MO-type rock salt phase takes place from about 400 to 441 °C together with the formation of metallic phase at about 460 °C. The major difference between this type of phase transitions and that for Li 0.27 Ni 0.8 Co 0.15 Al 0.05 O 2 in the presence of electrolyte is the delayed phase transition from the spinel-type to the rock salt-type phase by stretching the temperature range of spinel phases from about 20 to 140 °C. This unique behavior is considered as the key factor of the better thermal stability of the Li 1−x Ni 1/3 Co 1/3 Mn 1/3 O 2 cathode materials.

In the present work, a three-dimension (3-D) model of polymer electrolyte fuel cells (PEFCs) is employed to investigate the complex, non-isothermal, two-phase flow in the gas diffusion layer (GDL). Phase change in gas flow channels is explained, and a simplified approach accounting for phase change is incorporated into the fuel cell model. It is found that the liquid water contours in the GDL are similar along flow channels when the channels are subject to two-phase flow. Analysis is performed on a dimensionless parameter Da0 introduced in our previous paper [Y. Wang and K. S. Chen, Chemical Engineering Science 66 (2011) 3557–3567] and the parameter is further evaluated in a realistic fuel cell. We found that the GDL's liquid water (or liquid-free) region is determined by the Da0 number which lumps several parameters, including the thermal conductivity and operating temperature. By adjusting these factors, a liquid-free GDL zone can be created even though the channel stream is two-phase flow. Such a liquid-free zone is adjacent to the two-phase region, benefiting local water management, namely avoiding both severe flooding and dryness.

Abstract A dynamic three-phase transport model is developed to analyze water uptake and transport in the membrane and catalyst layers of polymer electrolyte fuel cells during startup from subfreezing temperatures and subsequent shutdown. The initial membrane water content ( λ , the number of water molecules per sulfonic acid site) is found to be an important parameter that determines whether a successful unassisted self-start is possible. For a given initial subfreezing temperature at startup, there is a critical λ ( λ h ), above which self-start is not possible because the product water completely engulfs the catalyst layers with ice before the stack can warm-up to 0 °C. There is a second value of λ ( λ l ), below which the stack can be self-started without forming ice. Between λ l and λ h , the stack can be self-started, but with intermediate formation of ice that melts as the stack warms up to 0 °C. Both λ l and λ h are functions of the initial stack temperature, cell voltage at startup, membrane thickness, catalyst loading, and stack heat capacity. If the stack is purged during the previous shutdown by flowing air in the cathode passages, then depending on the initial amount of water in the membrane and gas diffusion layers and the initial stack temperature, it may not be possible to dry the membrane to the critical λ for a subsequent successful startup. There is an optimum λ for robust and rapid startup and shutdown. Startup and shutdown time and energy may be unacceptable if the λ is much less than the optimum. Conversely, a robust startup from subfreezing temperatures cannot be assured if the λ is much higher than this optimum.

Abstract The catalyst layer (CL) of the polymer electrolyte membrane (PEM) fuel cell must be modeled accurately in order to resolve the effects of complex interactions between charge and mass transport on the fuel cell's electrochemical reactions. In previous work, we developed an agglomerate model [1] which correctly accounts for variations in the agglomerate surface area as the CL constituents are varied to provide a better estimate of diffusion losses. Here, this improved agglomerate model is employed to investigate a PEM fuel cell catalyst layer with a functionally-graded composition. We present results for varying catalyst and ionomer loadings in both the through-thickness and in-plane directions. In agreement with experimental observations, we find that a higher catalyst and/or ionomer loading at the membrane/CL interface improves performance especially in the ohmic loss regime. Similarly, improved performance is observed for higher catalyst and/or ionomer loadings under the channel in the mass transport loss regime. In addition, we investigated bidirectionally graded CLs for the first time. It is observed that higher performance can be obtained with bidirectionally graded CLs in both ohmic and mass transport loss regimes.

Abstract In this study, the lattice Boltzmann method (LBM) is used to investigate liquid water transport in the microporous layer (MPL) and gas diffusion layer (GDL) of polymer electrolyte membrane fuel cells (PEMFCs). Two-phase LB simulations are performed with modeled porous geometries that imitate multi-layer porous transport layers (PTLs) consisting of an MPL and a GDL. The simulation conditions are closely matched to the actual liquid water transport conditions in the PEMFCs. The results indicate that invasion-percolation processes due to strong capillary effects govern liquid water transport in PEMFCs. In addition, LB simulations are conducted by varying the intrusion thickness of the MPL and the surface wettability of the PTL. The results clearly show that the liquid water content can be reduced in the PTL by employing a thicker MPL and/or more hydrophobic surfaces. The steady-state water distribution is observed to occur more rapidly as the MPL becomes thicker or as the solid surfaces become more hydrophobic. Furthermore, several dynamic liquid water transport behaviors are identified from the results and explained in detail.

Abstract The performance of a ten-cell 50 cm 2 100 W polymer electrolyte membrane fuel cell (PEMFC) stack was evaluated under dynamic load cycling conditions utilizing the 2005 United States Department of Energy durability test protocol for PEFCs. An enhancement of performance was observed during the first 240 h, while an irreversible degradation of stack performance was observed after 480 h (∼4700 cycles). In particular, the stack voltage at 100 mA cm − 2 was decreased by 2.8% after 480 h and individual cell voltage was decreased up to 8%. An analysis of cell overpotentials for activation, Ohmic, and mass transport losses revealed that the predominant source of performance degradation was due to kinetic losses. The loss of catalyst utilization was estimated to be 39% based on the electrochemically active surface area (ECSA) measurements. Electron microscopic images of some of the cells showed growth in cathode Pt particle size from 5.3 to 6.2 nm. However, these microscopic images did not show any membrane damage or electrode thinning. Severe degradation of both the anode and cathode silicone gasket seals was observed during the durability test.

The electrochemical stability of electrolytes at lithium, or lithium-intercalating anodes, is achieved via ionically conducting surface films termed as solid electrolyte interphase (SEI). Since the lithium deposition or intercalation process occurs on the electrode covered with the SEI, the characteristics of the SEI determine the kinetics of lithiation/delithiation, stability of the interface, and thus, the overall cell performance, especially at low temperatures. In this paper, we have reiterated the significance of the SEI characteristics over the solution properties, using a few illustrative examples from our research on low temperature Li ion battery electrolytes at JPL. The examples specifically include the beneficial aspects of a ternary carbonate mixture of ethylene carbonate (EC), dimethyl carbonate (DMC) and diethyl carbonate (DEC) compared to the binary mixtures (of EC and either DMC or DEC) and quaternary solutions with appropriate co-solvents, such as alkyl esters.

Abstract Porous metal foams have gained much attention as the reactant distributor in polymer electrolyte membrane (PEM) fuel cell due to their high porosity, good distribution capacity, and excellent electrical and thermal conductivity. Water management is a major concern hindering their wide adoption in PEM fuel cells. In this study, two different pore size metal foam flow fields are digitally reconstructed based on their morphological analysis. Then, the two-phase flow field microstructures are simulated using the volume of fluid method (VOF) to investigate liquid droplet removal and dispersion. Various droplet diameters and ligament surface wettability are investigated. It is found that a comparable sized liquid droplet to the foam pore requires a high critical air velocity for droplet removal. Liquid dispersion in the metal foam flow field improves liquid removal, especially for large droplets. Moreover, increasing the ligament surface hydrophobicity or reducing the foam pore size helps liquid removal at a high-velocity air flow. The proposed method can be used to optimize the contact angle and pore size in the metal foam design for water removal in porous media flow fields.