Vibration durability of Pb-free HVQFN assemblies

Combining with the practical engineering case, this paper decomposed and analyzed the process of electronic products’ reliability assessment based on the Physics of Failure (PoF) method. The first step is to build a digital prototype model and then a PoF model. The second step is to use the Monte Carlo sampling method to obtain the failure samples of different fault points based on the PoF model. The third step is forming the failure DFs by fitting the failure samples. The fourth step is using the failure DF to calculate the Time to Failure (TTF)of the components. Finally, the paper can obtain the time of the TTF and the failure DF by applying the competitive failure principle. Compared with the reliability evaluation method based on test data, the reliability evaluation method based on PoF is more economical and efficient.

In this paper, the vibration durability of both SAC305 and Sn37Pb interconnects are investigated with narrow-band harmonic vibration tests conducted at the first natural frequency of the test, printed wiring board, using constant-amplitude excitation. A time-domain approach, reported by Upadhyayula and Dasgupta (1998, “Guidelines for Physics-of-Failure Based Accelerated Stress Test,” Proceedings, Reliability and Maintainability Symposium, pp. 345–357), was adapted in this study for the fatigue analysis. The test board consists of daisy-chained components, to facilitate real-time failure monitoring. The response of the test specimens was characterized, and accelerated fatigue tests were conducted at different loading amplitudes to obtain a mix of low-cycle fatigue (LCF) and high-cycle fatigue data points. The SAC305 interconnects were found to have lower fatigue durability than comparable Sn37Pb interconnects, under the narrow-band harmonic excitation levels used in this study. This trend is consistent with most results from broadband vibration tests by Zhou et al. (2006, “Vibration Durability Comparison of Sn37Pb vs. SnAgCu Solders,” Proceedings of ASME International Mechanical Engineering Congress and Exposition, Chicago, IL, Paper No. 13555), Zhou and Dasgupta (2006, “Vibration Durability Investigation for SnPb and SnAgCu Solders With Accelerated Testing and Modeling,” IEEE-TC7 Conference on Accelerated Stress Testing & Reliability, San Francisco, CA), and Woodrow (2005, “JCAA/JG-PP No-Lead Solder Project: Vibration Test,” Boeing Electronics Materials and Processes Technical Report) and from repetitive mechanical shock tests by Zhang et al. (2005, “Isothermal Mechanical Durability of Three Selected Pb-Free Solders: Sn3.9Ag0.6Cu, Sn3.5Ag and Sn0.7Cu,” ASME J. Electron. Packag., 127, pp. 512–522), but counter to findings from quasistatic, LCF, and mechanical cycling studies by Cuddalorepatta and Dasgupta (2005, “Cyclic Mechanical Durability of Sn3.0Ag0.5Cu Pb-Free Solder Alloy,” Proceedings of the ASME International Mechanical Engineering Congress and Exposition, Orlando, FL, Paper No. 81171). Failure analysis revealed two competing failure modes, one in the solder and another in the copper trace under the component. Thus solder fatigue properties extracted with the help of finite element simulation of the test article should be treated as lower-bound estimates of the actual fatigue curves.

The thermal cycling durability of large-area Pb-free (Sn3.5Ag) solder interconnects on photovoltaic (PV) solar cells, has been studied and benchmarked against existing Sn36Pb2Ag interconnects, using a combination of accelerated testing and physics-of-failure modeling. This paper reports the accelerated testing tasks and the modeling tasks are reported elsewhere in the literature. Accelerated thermal cycling tests have been conducted on photovoltaic laminates of both solder compositions, to characterize the increase in interconnect resistance due to fatigue damage. Interconnect resistance is measured up to 1000 cycles, from dark I-V curves for PV single-cell laminates. Resistance measurements show that single-cell Pb-free laminates outperform the Sn36Pb2Ag laminates. The Sn36Pb2Ag configuration shows a steady increase of resistance while the Pb-free configuration shows a bilinear increase as a function of thermal cycles. The resistance increase rate at 1000 cycles is higher for the Sn36Pb2Ag interconnects than for the Sn3.5Ag interconnects. A simple linear extrapolation of the trends observed during the first 1000 cycles, suggests that the Sn3.5Ag interconnect is 3.5 times more durable than the Sn36Pb2Ag interconnect. However, due to nonlinearities in the damage growth rate, this ratio is expected to be non-conservative and somewhat lower in practice. Post failure analysis of the PV laminates shows cracks predominantly close to the interface between the solder and the Ag ink used as contact electrodes on the silicon wafer. These results from the accelerated test results will be combined in a future paper with acceleration factors obtained from physics of failure modeling, to predict the thermal cycling durability under field conditions

Review of shelf life evaluation methods and a physics of failure approach for shelf life estimation for electronic components

Proton Exchange Membrane (PEM) failure is one of the key criteria in deciding the durability of the PEM fuel cell system. Physical properties such as ionic conductivity and mechanical properties depend on the microstructure. The microstructure is itself strongly dependent on the water content of the membrane (1, 2). It is therefore important to fully understand the effect of degradation stressors on the morphological changes and water-uptake characteristics in order to improve membrane durability.In this work, standard perfluorosulfonic acid (PFSA) ionomer membranes are subjected to a combined chemical and mechanical accelerated stress test (AST) used for rapid benchmarking of in-situ membrane stability. The degradation stressors in the combined chemical/mechanical AST was previously shown by our group to change the molecular structure and lead to membrane material loss (3). The chemical phase of the AST generates hydroxyl radicals that attack both the side chain and main chain of the polymer, while the mechanically generated stresses due to humidity cycling accelerate mechanical failures. The observed structural changes are anticipated to alter the water-uptake properties of the membrane, as shown previously for ex-situ testing (4). The degradation also alters the morphology, ion exchange capacity, mechanical properties, and proton conductivity of the membrane (5). Water-uptake measurements of partially AST degraded catalyst coated membrane (CCM) samples are carried out at room temperature using a dynamic vapor sorption system. In spite of severe fluoride release and membrane thinning, the water-uptake per unit mass of the AST degraded CCMs is found to be relatively constant. Since the test specimens are CCMs which contain a significant amount of catalyst layer material, an attempt is made to measure and separate the mass of the ionomer by thermogravimetric analysis (TGA). The ionomer material loss throughout the AST process is significant due to fluoride release and ionomer degradation, up to roughly 50% at end-of-life (3). The water-uptake per unit mass of ionomer in the CCM is determined by extracting the ionomer weight obtained by heating the CCM in the TGA chamber. In contrast to the previous results normalized by specimen dry weight, the water-uptake increases significantly per unit mass of the ionomer in the CCM, as shown in Fig. 1. Although the increased water-uptake may indicate enlarged solvated hydrophilic domains in the membrane, which is good for enhanced proton mobility, the proton conductivity is found to decrease (Fig. 1). This may be attributed to a decay in proton concentration due to sulfonic acid functional group loss during side chain degradation. The overall effects of the combined chemical/mechanical AST on the membrane morphology and properties are presented and compared to provide additional insight into the complex degradation mechanism.ACKNOWLEDGMENTSResearch funding provided by Automotive Partnership Canada (APC), Natural Sciences and Engineering Research Council of Canada (NSERC) and Ballard Power Systems (BPS) is gratefully acknowledged. BPS is also acknowledged for providing access to experimental facilities, material samples and technical support. REFERENCES1. J.Benziger, A. Bocarsly, M. J. Cheah, P. Majsztrik, B. Satterfield, and Q. Zhao, in Fuel Cells and Hydrogen Storage 141, Bocarly and Mingos, Editors, p. 85 Springer, Berlin Heidelberg 141 (2011).2. Y. Yang, A. Siu, T. J. Peckham, S. Holdcroft, in Advances in Polymer Sciences 215: Fuel cells 1, G. G. Scherer, Editor p. 55 Springer, Berlin Heidelberg 215 (2008).3. C. Lim, L. Ghassemzadeh, F. Van Hove, M. Lauritzen, J. Kolodziej, G.G. Wang, S. Holdcroft, E. Kjeang, J. Power Sources 257 102 (2014).4. L. Ghassemzadeh and S. Holdcroft, J.Am.Chem.Soc. 135 8181 (2013).5. S. V. Venkatesan and E. Kjeang, 11th International Fuel Cell Science, Engineering and Technology Conference, ASME, Minneapolis, MN (2013).

With the rapid development of SiP (System in Package) technology, reliability evaluation cannot be achieved by a single theory and traditional methods. In this paper, the reliability analysis technology based on PoF (Physics of Failure) is proposed. The performance degradation modeling, experiment design and reliability estimation based on Monte Carlo method are integrated to address the problems existing in the reliability work of the high-reliability and long-life product of SiP modules. Combining with an applicative example of reliability analysis case for a SiP module of plastic package, systematic analysis was conducted, including historical failure information research, typical failure mode, highly accelerated stress tests’ design, fault excitation, and failure mechanism analysis, in order to make further improvement on reliability analysis technology of SiP modules in electronic industry.

The reliability of circular electrical connectors was usually assessed according to standards like MIL-HDBK-217 (Reliability Prediction of Electronic Equipment). Given to their limitations and mislead results, a new assessment method needs to be presented. Some concerns on the physics of failure (PoF) modeling and reliability assessment of circular electrical connectors are addressed in this paper to overcome this defect. An original approach combining laboratory accelerated degradation testing (ADT) and PoF modeling is proposed for effectively assessing connector reliability. Random vibration and current stress are selected as acceleration factors that affect the contact life of electrical connectors. Then, an appropriate multiple-stresses ADT scheme is derived. The PoF model for electrical connectors under multiple stresses yielded the generalized Eyring relationship. According to the system reliability model for multiposition connectors, connectors' life follows Weibull distribution, which is derived from extreme-value distribution theory. Through ADT and data gained from tests, statistical analysis result yielded the estimated value of reliability character of MIL-C-38999 I series electrical connectors under the action of vibration and current stresses.

Flip chip has become the preferred package solution for high performance IC and microprocessor device [1]. In the area of flip chip, advanced Si node is always the mainstream technology. The integration of highly fragile ELK inner layer dielectric (ILD) material plays an important role to reduce the circuit delay time. In addition, the increasing performance demand continues to drive the scaling of feature size of the silicon as well as tighter bump pitch, smaller bump diameter and large die size for future flip chip packages [2]. From the last decade, there has been a significant focus on the development of Pb-free bump interconnection in flip chip package [3]. The Pb-free bump interconnection leads to significant challenges because the elastic modulus and yield strength of Pb-free alloys are significantly higher than that of conventional eutectic Sn-Pb material, and higher melting temperature is required to meet the Pb-free soldering in chip attach process. As a result, ILD delamination has emerged as a major reliability concern with the adoption of the intrinsically higher stiffuess of Pb-free solder and lower mechanical strength of ELK dielectric materials in silicon backend structure. It is a challenge to manufacture a robust and reliable advanced flip chip product [4-6]. In this paper, we summarize some of the early development work of the high performance ASIC. The die size of the test vehicle is approximately 18x19mm fabricated using 40nm ELK technology incorporating Sn-1.8Ag plated bumps whose pitch is 150um (minimum). It is fabricated on 300mm wafer with polyimide passivation. The laminate is a 50x50mm square substrate with 800um core and 5-2-5 (12 layers) build-up structure. The Pb-free BGA has 2397 pads with 1.0mm pitch and full array format. It is well known that underfill materials and related processes are key technology for flip chip BGA (FCBGA) packaging. Underfill is an epoxy silica composite material with optimum thermo-mechanical properties (CTE, Tg, Modulus) to minimize stress transfer to die, then to prevent solder joint fatigue, ELK ILD delamination and die crack. In this study, UF selection is conducted to compare different underfill materials process capability, compatibility and reliability performance in accelerated stress tests. Common failure modes are studied, e.g., ILD delamination besides UF crack and interfacial delamination between polyimide and UF. One underfill option has passed internal qualification test including uHAST 96hrs, TCB 1000 cycles and HTST 1000hrs.

The next wave of LED lighting technology is likely to be tunable white lighting (TWL) devices which can adjust the colour of the emitted light between warm white (∼ 2700 K) and cool white (∼ 6500 K). This type of lighting system uses LED assemblies of two or more colours each controlled by separate driver channels that independently adjust the current levels to achieve the desired lighting colour. Drivers used in TWL devices are inherently more complex than those found in simple SSL devices, due to the number of electrical components in the driver required to achieve this level of control. The reliability of such lighting systems can only be studied using accelerated stress tests (AST) that accelerate the aging process to time frames that can be accommodated in laboratory testing. This paper describes AST methods and findings developed from AST data that provide insights into the lifetime of the main components of one-channel and multi-channel LED devices. The use of AST protocols to confirm product reliability is necessary to ensure that the technology can meet the performance and lifetime requirements of the intended application.

Degradation issues remain among the main factors impeding the development of commercially viable fuel cell (FC) systems for transportation and stationary power applications. The necessity to keep costs and performances at reasonable levels led to the widespread use of carbon supported platinum (Pt) for electrodes and thin PFSA membranes as electrolyte. Both are quite fragile materials but currently without any convincing alternative. To better understand, and eventually limit degradation mechanisms, several kinds of accelerated stress tests (AST) have been developed targeting each material independently: namely open circuit voltage (OCV) for PFSA membranes, potentiostatic regimes for carbon, and potential cycling for Pt [1, 2]. Other accelerated stress tests corresponding to common fuel cell operation sequences were also proposed like constant current, start/stop, hydrogen starvation/dead-ended anode protocols [3,4], temperature, relative humidity or potential/load cycling, or even freeze/thaw protocols [5]. In this work, we focus on load cycling events, which are very close to the potential cycling performed during AST to study Pt and carbon degradation mechanisms. For obvious reasons, load cycling in actual fuel cell systems and potential cycling in AST are usually performed in overfed conditions [6]: i.e. the fuel cell is supplied with the gas flow rates corresponding to the highest current before the intensity increases. And in most of the cases, the gas flow rates are even permanently kept at the highest values during AST. To avoid supplying fuel and air in excess, most of fuel cell stacks are hybridized either with batteries or supercapacitors (SC) to improve the system dynamics (which in return increases its size and cost). Thanks to hybridization, the FC voltage can drop by a few hundreds of mV per cell during the time needed by the gas supply lines to adapt the flow rates, while batteries or SC keep the system power at the required value. However, the impact of such short reactant starvation events on FC performance and durability is not very well known. Therefore, we performed repeated current Heaviside steps as AST on PEM single cells; typical values were from 0.25 to 0.95 A/cm² (with corresponding FC voltages between 0.65 and 0.8 V). The gas flow rates were set either to the maximum current intensity (with air and H2 stoichiometries of 3 and 1.5 in open-anode or 1 in dead-end mode, respectively) or we let the gas lines adjust the flow rates when the current input changed, which took about half a second (during this time, the FC could fell to about 0.1 V). As shown on the Figure, the performance and electrochemical surface area repeatedly aged significantly slower in the second case, i.e. with short gas starvation events. To better understand this difference, similar experiments were performed in a segmented cell [7] with measurements of local currents, anode and cathode potentials. We observed that local cathode potentials were higher in overfed conditions while they dropped for a short time in the other case due to (most probably) oxygen depletion in the cathode compartment. Local negative currents were also observed sometimes when the load decreased, but since they were not associated to high local potentials, they did not worsen degradation. In the same way, since the FC dynamics were limited most probably by air supply, the anode (and thus cathode) local potentials remained low. These results suggest that FC voltage drops during transients due to short lacks of reactants can improve the durability of the catalyst layer, as well as optimize the system efficiency in terms of hydrogen consumption and air compressor power demand which can be a useful observation for the design of optimized hybridization strategies.

A PEM fuel cell’s durability and lifetime are affected by various degradation phenomena, among which chemical degradation of PEM material exerts the most damage. Chemical degradation of PEM occurs when reactive radicals such as hydroxyl (·OH) and hydroperoxyl (·OOH) are generated from the decomposition of hydrogen peroxide. The radicals chemically attack the main and side chains of the perfluorosulfonic acid-based PEM and change its internal chemical structure. This leads to excessive membrane thinning and pinhole formation that causes higher gas crossover and electrode shorting inside the PEM fuel cell1. The generation of harmful radicals accelerates in the presence of Fenton’s metal, such as iron (Fe)2. Evaluation of PEM fuel cell chemical degradation durability is performed using long-term pure chemical accelerated stress test (AST) under high temperature (T), low relative humidity (RH), and open circuit voltage (OCV) conditions3. Physical characterization of a membrane sample used in a fuel cell can help to visualize and assess damage incurred during an AST. These results, however, fail to encompass the formation rate of radicals and intermediate degradation products generated during the AST3. To overcome these issues, computational fluid dynamics models are developed that simulate the evolution of degraded products and predict the lifetime of PEM fuel cells4,5.The present work aims to develop an in-situ one-dimensional (1-D) transient chemical degradation model that describes reaction-transport phenomena within the membrane electrode assembly (MEA) in the presence of Fe and Ce cations. This work considers mechanically reinforced GORE-SELECT® 15 µm baseline (MemA) and chemically and mechanically reinforced 12 µm (MemB-Ce) membranes. The developed in-situ model simulates pure chemical AST operating conditions with variable Fe ion loading and follows the computational algorithms originally presented by Wong and Kjeang4. The model successfully predicted that side chain scission is followed by main chain unzipping, as reported in the literature6. The evaluation of the change in ionomer structure by chemical degradation was done by computing the temporal concentration profile of intermediate degraded ionomer species, hydrogen peroxide, and membrane thickness reduction. The 1-D model also encapsulates the concentration profile of Fe2+ and Fe3+ ions across the PEM and predicts their temporal distribution throughout the simulated AST duration under OCV conditions. It was observed that the MemA thickness decreased rapidly in the presence of Fe ions, as shown in Figure 1. The computational parameters and results were calibrated and validated using in-house AST results for MEAs with embedded Fe metallic particles, which were shown to completely dissolve during conditioning and hence treated as uniformly distributed Fe cations with an initial concentration of 900 ppm at the beginning-of-life7. The model identified a limiting Fe ion concentration limit of 100 ppm that accentuates the chemical attack. Experimentally, a low Fe concentration of 10 ppm was determined in the baseline MemA, which also reduced the membrane thickness within 1125 h of operation, as shown in Figure 1. It is known that cerium (Ce) ions can scavenge harmful radicals and increase the lifetime of PEM fuel cells8. The model, therefore, simulates the effectiveness of Ce ions in the MemB-Ce membrane and the temporal distribution of Ce3+/Ce4+ ions and shows a lower rate of membrane thickness reduction, as shown in Figure 1. The modelling framework was, therefore, able to predict PEM fuel cell lifetime in conjunction with the transport behaviour of both harmful and beneficial cations as well as their interactions during the chemical degradation process at OCV conditions. Keywords: PEM fuel cell, fuel cell durability, chemical degradation model, in-situ modelling, iron, cerium Acknowledgements This research was supported by the Natural Sciences and Engineering Research Council of Canada, Canada Foundation for Innovation, British Columbia Knowledge Development Fund, Western Economic Diversification Canada, Ballard Power Systems, and W.L. Gore & Associates. It was also undertaken in part thanks to funding from the Canada Research Chairs program. References J. Healy et al., Fuel Cells, 5, 302–308 (2005).S. Kundu, L. C. Simon, and M. W. Fowler, Polym. Degrad. Stab., 93, 214–224 (2008).Y. Singh, F. P. Orfino, M. Dutta, and E. Kjeang, J. Electrochem. Soc., 164, F1331–F1341 (2017).K. H. Wong and E. Kjeang, J. Electrochem. Soc., 161, F823–F832 (2014).R. B. Ferreira, D. S. Falcão, and A. M. F. R. Pinto, Int. J. Hydrogen Energy, 46, 1106–1120 (2021).L. Ghassemzadeh, K. D. Kreuer, J. Maier, and K. Müller, J. Phys. Chem. C, 114, 14635–14645 (2010).N. Kumar et al., ECS Meet. Abstr., MA2023-02, 1919–1919 (2023).S. M. Stewart, D. Spernjak, R. Borup, A. Datye, and F. Garzon, ECS Electrochem. Lett., 3, F19–F22 (2014). Figure 1

The adoption of PEM fuel cells for heavy-duty applications, such as maritime transport, often necessitates the use of liquid hydrogen carriers like methanol due to storage and handling considerations.[1] Methanol reforming is an efficient method to generate hydrogen and CO₂ on-site; however, trace amounts of CO invariably remain in the reformate gas, despite subsequent purification steps such as water-gas shift and selective oxidation.[2,3] These residual CO levels, typically in the low ppm range, significantly exceed the tolerance threshold of platinum-based catalysts on the anode (<0.2 ppm),[4] resulting in catalyst poisoning and diminished cell performance. To address this, RuPt alloy nanoparticles supported on carbon are frequently employed for their enhanced tolerance to CO under reformate conditions.[3] Nevertheless, operational challenges arise as Ruthenium (Ru) leaches from the anode catalyst during prolonged use. This Ru migrates through the membrane and deposits on the cathode, poisoning the Pt/C catalyst and severely inhibiting oxygen reduction reaction (ORR) activity.[3,5]Efforts to mitigate Ru leaching have focused on modifying catalyst surface composition and structure. However, because Ru leaching is a long-term phenomenon, accelerated stress tests (ASTs) are essential for investigating this behavior in a practical timeframe. Traditional MEA-based ASTs, while effective, are hindered by high material requirements, long preparation times, and elevated costs, making them less suitable for early-stage catalyst screening. Additionally, quantifying the total amount of Ru leached has remained a significant challenge.In this study, we developed a robust, material-efficient AST protocol using the rotating disk electrode (RDE) to quantitatively evaluate Ru leaching. Three commercial RuPt catalysts with Ru:Pt atomic ratios of 1:1, 1.5:1, and 2:1 were examined, along with a heat-treated 2:1 ratio catalyst with larger particle size and reduced electrochemical surface area (ECSA). The RDE-based AST involved cyclic voltammetry to simulate catalyst aging, with key parameters such as cycle number (100–1000), scan rate (10–100 mV s⁻¹), and upper potential limit (0.8–0.9 V_RHE) systematically varied. Ru dissolution was quantified using inductively coupled plasma optical emission spectroscopy (ICP-OES) from collected electrolyte samples (0.1M H₂SO₄), and changes in catalyst surface composition were assessed through shifts in CO-oxidation peaks.MEA-based AST measurements were also performed to evaluate the stability of these commercial catalysts and compare the results with those obtained from the RDE protocol. This study examines the correlations between RDE- and MEA-derived stability rankings and discusses the limitations of RDE-based ASTs when applied to practical fuel cell conditions.The validated RDE protocol was further applied to self-synthesized RuPt/C and Ru_corePt_shell materials, providing a reliable and cost-effective approach for assessing Ru leaching during early-stage catalyst development. These findings highlight the potential of the RDE-based AST for advancing PEM fuel cell anode catalysts tailored for reformate hydrogen applications.

Although electrochemical water splitting currently represents a small amount of the US hydrogen supply, its role is expected to grow with the emergence of a hydrogen-based infrastructure.[1] Electrolysis is reasonably cost-competitive with a variety of technologies, including batteries (solid state and flow), pumped hydro, and compressed air.[2] When coupled with renewable energy sources, such as wind and solar, electrochemical water splitting can produce hydrogen using a nearly carbon free pathway.[3] The commercial progress of electrolyzers is limited in part by the cost of the catalyst layer. The oxygen evolution reaction is the focus of electrocatalyst development due to slow kinetics relative to hydrogen evolution; iridium nanoparticles are commonly used to catalyse oxygen evolution, with platinum added to ensure conductivity throughout the catalyst layer. Extended surface nanostructures have previously been developed as electrocatalysts in hydrogen fuel cells, where they have produced an order of magnitude higher specific activity than nanoparticles in oxygen reduction and demonstrated a potential benefit in accelerated stress tests.[4] Extended nanostructures have recently been developed for the oxygen evolution reaction, where a similar specific activity improvement was observed in rotating disk electrode (RDE) half-cells. Catalysts formed by spontaneous galvanic displacement are ideally situated, being able to take advantage of the specific activities generally associated with the catalyst type, while significantly improving upon the traditionally low surface areas of extended surfaces. This approach has been used in forming iridium-nickel and iridium-cobalt nanowires.[5] While nanoparticles have modest site-specific activity in oxygen reduction and evolution, nanoparticle evolution catalysts are at a particular disadvantage with regards to surface area, since carbon supports cannot be used at high potentials. Iridium-cobalt and iridium-nickel nanowires have been synthesized with surface areas of 70 m2 gIr ‒ 1, more than double the nanoparticles. In RDE half-cells, the extended nanostructures produce an oxygen evolution mass activity 7 to 8 times greater throughout the kinetic region. Accelerated stress tests were completed by potential holds and cycling over a variety of conditions to study catalyst degradation and establish standardized testing protocols. Acid treatment has been used to remove excess nickel and cobalt template, improving catalyst durability. Following accelerated stress tests in RDE half-cells, iridium-cobalt and iridium-nickel nanowires exceed the mass activity of nanoparticles by more than an order of magnitude. Membrane electrode assembly (MEA) fabrication of these materials has produced electrolyzers with higher performance than the benchmark device. Recent improvements in MEA performance suggest that the promise of extended surface catalysts observed in half-cells can be realized in the application. [1] R. Forgie, G. Bugosh, K.C. Neyerlin, Z. Liu, P. Strasser, Electrochemical and Solid-State Letters, 13 (2010) B36-B39. [2] K. Harrison, M. Peters, in: U.S. Department of Energy (Ed.), http://www.hydrogen.energy.gov/pdfs/review13/pd031_harrison_2013_o.pdf, 2013. [3] K. Harrison, M. Peters, C. Ainscough, in: U.S. Department of Energy (Ed.), http://www.hydrogen.energy.gov/pdfs/progress13/ii_a_2_harrison_2013.pdf, 2013. [4] S.M. Alia, Y.S. Yan, B.S. Pivovar, Catalysis Science & Technology, 4 (2014) 3589-3600. [5] H. Xu, High-Performance, Long-Lifetime Catalysts for Proton Exchange Membrane Electrolysis, in: U.S. Department of Energy (Ed.), http://www.hydrogen.energy.gov/pdfs/review14/pd103_xu_2014_o.pdf, 2014. Figure 1. Surface areas (x-axis) and site-specific oxygen evolution activities (y-axis) of iridium-cobalt nanowires, iridium-nickel nanoparticles, and iridium nanoparticles prior to (red, blue, black) and following (dark red, dark blue, grey) durability testing in RDE half cells. Arbitrary mass activities (1, 4, 7 A mgIr ‒ 1) are included as solid black lines. Figure 1

For the use of polymer electrolyte membrane fuel cells (PEMFCs) in heavy duty applications, on-site generation of hydrogen rich gas by reforming hydrocarbons or methanol are a suitable alternative to pure H2 due to higher availability and higher energy density of hydrocarbon fuels [1]. The so-called reformate contains in addition of H2 also CO, which is a catalyst poison. PtRu alloys are the most promising catalysts for PEM fuel cells operated with reformate due to their high CO-tolerance [2]. It is known that the less noble Ru is unstable in acidic PEMFC conditions and is lowering the durability of the membrane electrode assembly (MEA) by dissolution and crossing over through the membrane to the cathode [3]. Ru dissolution and crossover result in a decreased CO-tolerance on the anode and a lower activity of the cathode towards oxygen reduction due to Ru blocking the Pt surface of the cathode catalyst [4]. To prolong the lifetime and CO-tolerance of reformate PEMFCs, the stability of the anode catalysts must be improved. One approach is nano-structuring the catalysts by encapsulating the less noble Ru-core with a Pt-shell. In doing so, the corrosion resistant Pt-shell supposes to protect the Ru-core from dissolution and thus mitigates the Ru crossover phenomenon.In this work, the influence of the Pt-shell thickness on the stability and performance of the Ru@Pt catalysts is investigated. Two catalysts with varying Pt-shell thicknesses were synthesized on Vulcan XC72R carbon via a two-step polyol. The catalyst with the thinner Pt-shell is called Ru@1Pt and with the thicker shell Ru@2Pt. The as-synthesized catalysts were physically characterized by XRD, ICP-OES, TEM and EDS. A representative EDS map of Ru@1Pt is shown in fig. 1a. The physical characterization of the obtained catalysts unveils a Ru@Pt core-shell structure with crystalline hcp Ru and fcc Pt. To further investigate the catalysts electrochemically, MEAs were manufactured using Ru@Pt catalysts as anodes and commercial membranes and cathode catalysts. An accelerated stress test (AST) was developed to target Ru dissolution by potential cycling and was applied to investigate and compare the stability of the catalysts. The MEAs were electrochemically characterized by cyclic voltammetry, CO-stripping and U-I curves before, during and after the AST. Post-mortem cross-section STEM analysis of the stressed cells were performed to evaluate the elemental composition and morphology of the anode and cathode after the stress test.During the AST, CO-stripping on the Ru@1Pt anode showed a peak shift to higher potentials, indicating a decrease in CO-tolerance due to Ru loss. By contrast, on the Ru@2Pt anode, the peak shifted to lower potentials implying a stronger promoting influence of Ru. On the cathode side, CV measurements before and after the AST revealed a double layer increase and a decrease in the HUPD. This change implies the poisoning of Pt with Ru and were more pronounced for Ru@1Pt compared to Ru@2Pt. As seen by the CO-stripping and CV results, the Ru dissolution and crossover was more severe for Ru@1Pt than for Ru@2Pt. Hence, a higher fuel cell performance loss was observed for the cell with Ru@1Pt in H2, as can be seen in fig. 1b. In addition, further U-I curves were measured under varying reformate conditions to identify the influence of CO on the cell performance. Cross-section EDS maps of the stressed MEAs were used to evaluate the Ru dissolution and its distribution in the different layers of the MEA. A higher Ru:Pt ratio was found on the cathode for the cell with Ru@1Pt compared to a cell with Ru@2Pt, which further verifies a higher degree of Ru crossover for the catalyst with the thinner Pt-shell. The above findings could evidence that the shell thickness plays a significant role in the corrosion resistance of core-shell reformate anode fuel cell catalysts.

Platinum-cobalt alloy nanoparticles are of great interest as cathode catalysts for PEMFCs as they have been shown to have enhanced activity versus platinum. However, their relative stability against loss of electrochemically-active surface area in relation to Pt catalysts is still debated. In this study, the evolutions of Pt3Co particle size distributions (PSDs) in fuel cell and aqueous environments were followed during accelerated stress tests (ASTs) using in-operando ASAXS. The measured evolutions showed a degradation mechanism dominated by loss of particles smaller than the critical particle diameter (<5.2 to 6.1 nm, CPD), which depended on environment and the AST. These evolutions were compared to that of a Pt catalyst with a similar initial PSD, which was found to have a lower degradation rate than Pt3Co. The ASAXS data, as well as data from aqueous dissolution, X-ray absorption spectroscopy, individual particle energy-dispersive spectroscopy, X-ray fluorescence spectrometry, and kinetic Monte Carlo calculations support a loss mechanism of increased Pt dissolution from Pt3Co versus Pt due to destabilization caused by extensive dealloying of particles <∼5 nm during ASTs.