Search

PANI-grafted carbon supports can stabilize Pt nanoparticles (NPs) during PEMFC operation. The location of the NPs depends on PANI loading, with only optimal PANI loading (33 wt%) holding the NPs close to the carbon surface – enhancing the stability.

PtM (M = 3d transition metals) alloys are known as the promising oxygen reduction reaction catalysts and have been considered as the replacement of pure Pt catalysts for the commercialization of proton exchange membrane fuel cells. Although great progress has been made in the past three decades, the performance and durability of PtM catalysts still face stringent challenges from practical applications. Functionalization of a catalyst carbon support with nitrogen-contained groups can add charges onto its surface, which can be utilized to build a more complete ionomer/catalyst interface, to reduce the catalyst particle size, and to improve particle size distribution. Nitriding of PtNi catalysts can effectively improve the catalyst activity and stability by the modification of lattice strain. Hereby, we propose a synergistic approach of combining polybenzimidazole-grafted Vulcan XC72 carbon as the catalyst carbon support and the nitriding of PtNi to develop PtNiN/XC72-polybenzimidazole catalysts. Such PtNiN/XC72-PBI catalysts exhibit the excellent performance of fuel cell membrane electrode assembly (i.e., mass activity, 440 mA mg

Proton exchange membrane fuel cells (PEMFCs) have long been studied for applications in sustainable transportation. The recent shift in focus from passenger vehicle to heavy duty applications places increased demand on the efficiency and durability of the components used in membrane electrode assemblies (MEAs).[1] Of particular interest are the oxygen reduction reaction electrocatalysts used in the fuel cell cathode. While the relaxed capital cost requirements for heavy duty applications allows for increased precious metal loadings, the four-fold increase in operating lifetime requires improved catalysts which can express their high activity not only at beginning-of-life, but also after 25k-hour equivalent accelerated stress tests (ASTs). This paradigm shift prioritizes efforts aimed at understanding and subsequently controlling catalyst degradation. In situ characterization methods can be instrumental in this effort, but it is critical that the operating conditions in these studies reliably replicate the type and degree of degradation observed in the operating fuel cell.[2,3] In this study, identical-location scanning transmission electron microscopy (IL-STEM) is used to compare catalyst degradation in an aqueous environment to degradation observed ex situ in MEAs following AST cycling.[4] For traditional catalysts, the aqueous environment appears to replicate the dissolution and coalescence mechanisms observed for Pt catalysts with small starting particle sizes ( To better replicate the type and degree of degradation in the MEA environment, various experimental parameters such as temperature, acid type/concentration, and upper and lower potential limits were explored. By extending the cycling potential window from 0.6-0.95 Vvs . to 0.4-1.0 Vvs . RHE in an sulfuric acid electrolyte containing Pt ions, the Ostwald ripening mechanism was enhanced, resulting in an end-of-test particle size distribution and composition which better match MEA tests. These refined conditions can now be used to perform in situ and identical-location experiments that will play an important role in the development of robust catalysts for heavy-duty vehicle applications.[6] References [1] D. A. Cullen, et al. Nat. Energy 2021, 6 (5), 462–474. [2] J. A. Gilbert, et al. Electrochim. Acta 2015, 173, 223–234. [3] I. Martens, et al. ACS Energy Lett. 2021, 6 (8), 2742–2749. [4] S. Rasouli, et al. Nano Lett. 2019, 19 (1), 46–53. [5] E. Padgett, et al. J. Electrochem. Soc. 2019, 166 (4), F198–F207 [6] This material is primarily based on work performed by the Million Mile Fuel Cell Truck (M2FCT) Consortium (https://millionmilefuelcelltruck.org), technology manager Greg Kleen, which is supported by the U.S. Department of Energy, Office of Energy Efficiency and Renewable Energy, Hydrogen and Fuel Cell Technologies Office. Electron microscopy was conducted at the Center for Nanophase Materials Sciences, which is a DOE Office of Science User Facility.

The biggest obstacle to the widespread implantation of polymer electrolyte membrane fuel cells (PEMFCs) is the cost, primarily due to the use of platinum catalysts. The high intrinsic catalyst activity...

Fe-N-C catalysts, the most promising platinum group metal (PGM)-free oxygen-reduction catalysts, often simultaneously contain pyrrolic N- (S1) and pyridinic N (S2) -coordinated FeN4 sites. These two types of active sites show significantly different intrinsic activity and stability. S1 sites are more active but less stable compared to S2 sites. Designing a Fe-N-C catalyst, which exclusively contains active S1 sites with enhanced intrinsic stability, is highly desirable to break the activity-stability trade-off. Herein, we report a Fe-N-C model catalyst that solely comprises S1 sites prepared by adding H2 in the pyrolysis atmosphere (i.e., 10% H2/Ar). A membrane electrode assembly (MEA) with the Fe-N-C cathode demonstrated compelling activity and generated a current density of 50.8 mA cm−2 at 0.9 ViR-free (H2-O2) and 211 mA cm−2 at 0.8 V (H2-air), which have significantly exceeded the U.S. DOE 2025 targets. The highly active Fe-N-C catalyst also demonstrated improved stability during life tests and accelerated stability tests (ASTs). The knowledge obtained from experimental and theoretical results elucidates that the FeN4 site formation process can be controlled by thermal activation atmospheres, which is essential to breaking activity-stability trade-off and design viable Fe-N-C catalysts with adequate activity and stability for proton exchange membrane fuel cells.

Fused phenazine-based organic redoxmers capable of six-electron redox activities have been designed and evaluated in aqueous flow batteries. Such a novel molecular engineering approach opens an avenue for enabling energy-dense storage strategies.

Significantly reducing platinum group metal (PGM) loading while improving catalytic performance and durability is critical to accelerating proton-exchange membrane fuel cells (PEMFCs) for transportation. Here we report an effective strategy to boost PGM catalysts through integrating PGM-free atomically-dispersed single metal active sites in the carbon support toward the cathode oxygen reduction reaction (ORR). We achieved uniform and fine Pt nanoparticle (NP) (∼2 nm) dispersion on an already highly ORR-active FeN4 site-rich carbon (FeN4–C). Furthermore, we developed an effective approach to preparing a well-dispersed and highly ordered L12 Pt3Co intermetallic nanoparticle catalyst on the FeN4–C support. DFT calculations predicted a synergistic interaction between Pt clusters and surrounding FeN4 sites through weakening O2 adsorption by 0.15 eV on Pt sites and reducing activation energy to break O–O bonds, thereby enhancing the intrinsic activity of Pt. Experimentally, we verified the synergistic effect between Pt or Pt3Co NPs and FeN4 sites, leading to significantly enhanced ORR activity and stability. Especially in a membrane electrode assembly (MEA) with a low cathode Pt loading (0.1 mgPt cm−2), the Pt/FeN4–C catalyst achieved a mass activity of 0.451 A mgPt−1 and retained 80% of the initial values after 30 000 voltage cycles (0.60 to 0.95 V), exceeding DOE 2020 targets. Furthermore, the Pt3Co/FeN4 catalyst achieved significantly enhanced performance and durability concerning initial mass activity (0.72 A mgPt−1), power density (824 mW cm−2 at 0.67 V), and stability (23 mV loss at 1.0 A cm−2). The approach to exploring the synergy between PGM and PGM-free Fe–N–C catalysts provides a new direction to design advanced catalysts for hydrogen fuel cells and various electrocatalysis processes.

Due to their exceptional catalytic properties for the oxygen reduction reaction (ORR) and other crucial electrochemical reactions, PtCo intermetallic nanoparticle (NP) and single atomic (SA) Pt metal site catalysts have received considerable attention. However, their formation mechanisms at the atomic level during high-temperature annealing processes remain elusive. Here, the thermally driven structure evolution of Pt-Co binary catalyst systems is investigated using advanced in situ electron microscopy, including PtCo intermetallic alloys and single Pt/Co metal sites. The pre-doping of CoN4 sites in carbon supports and the initial Pt NP sizes play essential roles in forming either Pt3 Co intermetallics or single Pt/Co metal sites. Importantly, the initial Pt NP loadings against the carbon support are critical to whether alloying to L12 -ordered Pt3 Co NPs or atomizing to SA Pt sites at high temperatures. High Pt NP loadings (e.g., 20%) tend to lead to the formation of highly ordered Pt3 Co intermetallic NPs with excellent activity and enhanced stability toward the ORR. In contrast, at a relatively low Pt loading (

In this work, we developed a Fe2O3@ZIF-8 composite precursors and combined with a novel thermal activation treatment under an Ar/H2 mixture atmosphere (i.e., forming gas), which successfully prepares a highly active Fe-N-C catalyst. 57Fe Mossbauer spectroscopy analysis experimentally verified that the catalyst only contains S1 sites. The highly active Fe-N-C catalyst achieved exceptional ORR initial activity in acids, exceeding that of a Pt/C baseline catalyst (60 µgPt cm−2) in rotating disk electrode (RDE) tests. The MEA tests further verified that the Fe-N-C catalyst's activity is competitive with Pt/C cathode (0.1 mgPt cm−2) in the kinetic range. Notably, the catalyst demonstrated remarkable activity of 50.8 mA cm−2 (@0.9 ViR-free) under H2-O2 conditions, exceed the U.S. DOE target. As expected, the catalyst degraded significantly during the stability ASTs. To increase S1 sites and address the stability issues, we further developed in-situ CVD methods to treat Fe2O3@ZIF-8 precursors under forming gas flow but with additional N/C precursors that was put at the upstream sides of the tube furnace. Therefore, additional nitrogen and carbon sources are added to the Fe2O3@ZIF-8 precursors during the thermal activation of catalysts. As a results, we can populate the intrinsically stable S2 sites to design highly durable Fe-N-C catalyst. The resulting catalyst presented outstanding stability in both RDE and fuel cell tests. The E 1/2 gained 21 mV after 100,000 cycles of accelerated degradation test (ADT), achieving 0.869 VRHE at the end of the test, which is close to Pt/C (60 µgPt cm−2) in acid electrolyte. The current density of MEA at 0.8 V retained unchanged, and the peak power density increased from 454.0 to 512.0 mW cm−2 in H2-fuel cell after 30,000 cycles of AST.

Platinum group metal (PGM) catalysts are the major electrocatalysts for oxygen reduction reaction (ORR) in the polymer electrolyte membrane fuel cells (PEMFCs). However, the high cost of PGM catalysts is the major huddler for the widespread applications of fuel cell electric vehicles. To remove this cost obstacle of fuel cell commercialization, PGM-free catalysts have been considered as the replacement of PGM catalysts for ORR because of the low cost and relatively comparable performance with PGM catalyst. Fe-C-N complex is the one of the most active centers in PGM-Free catalyst groups. This type of catalyst shows excellent activity characterized using the rotation disk electrode (RDE), i.e., the half wave potential (E1/2 ) could reach 0.91 V versus standard hydrogen electrode (SHE). However, in a membrane electrode assembly (MEA), the performance of PGM-Free catalysts cannot achieve the comparable performance to PGM catalyst. Since there are so many differences between PGM-free, and PGM catalysts e.g., activity, stability, surface conditions, particle size etc. The fabrication of PGM-Free catalyst MEA cannot simply borrow the methods from that of making PGM MEA. In addition, the thicknesses of catalyst layers of PFM-free are significantly thicker than that of PGM, i.e., 10 times. Hereby, we proposed a novel method of fabricating PGM-Free catalyst MEA, so that the intrinsic catalyst activity from RDE can be translated into MEA performance. The method is based on the catalyst coated membrane (CCM) method using optimized ionomer to carbon (I/C) ratio and solvent mixture of catalyst ink. Such method pushes PGM-free MEA first ever achieved the current density of 50.8 mA cm-2 at 0.9 V iR-free in H2/O2 and over 150 mA cm-2 at 0.8 V in H2/air, which surpassed the 2025 performance targets of US Department of Energy (DOE) for PGM-Free catalyst MEA. Further, the property (solvent composition, dispersion of catalyst and ionomer in an ink), structure (pore structure) and the MEA performance have been characterized using mercury intrusion porosimetry (MIP), MEA testing. A property-structure-performance relationship has been established.

Atomically dispersed and nitrogen-coordinated single iron sites (FeN4) embedded in carbon (Fe-N-C) catalysts are the most promising platinum group metal (PGM)-free catalysts. However, they have yet to match their Pt counterparts for oxygen reduction reaction (ORR) activity and stability in proton exchange membrane fuel cells (PEMFCs). Here, we developed viable Fe-N-C catalysts, which, for the first time, demonstrated competitive activity to that of Pt/C catalysts and dramatically enhanced stability and durability under practical PEMFC operating conditions. The most active Fe-N-C catalyst achieved a record half-wave potential (E1/2 = 0.915 V vs. RHE at 0.6 mgcatcm-2) and an ORR mass activity of 10.5 mA mgcat at 0.9 V in (RDE) tests, exceeding a Pt/C baseline catalyst (60 µgPt cm-2) by 40 mV in acidic electrolytes. This compelling activity of the Fe-N-C catalyst in aqueous acids on rotating disk electrode (RDE) was successfully transferred to a fuel cell membrane electrode assemblies (MEAs), generating an initial current density of 44.2 mA cm-2 exceeding the U.S. DOE 2025 target (i.e., 44 mA cm-2) at 0.9 VIR-free under O2. Under practical hydrogen-air conditions, record 151 mA cm-2 at 0.8 V and peak power density of 601 mW cm-2 were achieved. Importantly, we discovered that depositing nitrogen-carbon species on the catalyst surface via chemical vapor deposition (CVD) dramatically enhanced catalyst stability, evidenced by performance durability after accelerated stress tests (30 000 square-wave voltage cycles under H2/air) and long-term steady-state life tests (> 300 hours at 0.67 V). Innovative identical location-scanning transmission electron microscopy (IL-STEM) experiments confirmed that the CVD process leads to deposition of nitrogen-doped carbon onto the catalyst surfaces. Along with theoretical modeling, a reconstruction of the carbon structure adjacent to FeN4 sites leads to increased robustness against demetallation and carbon oxidation. This work opens new avenues for developing earth-abundant iron-based catalysts with extraordinary activity and stability, thus competing with Pt and addressing the cost barrier of current PEMFCs.

Rational design of catalyst layers in a membrane electrode assembly (MEA) is crucial for achieving high-performance polymer electrolyte membrane fuel cells. Establishing a clear understanding of the property (catalyst ink)-structure (catalyst layer)-performance (MEA) relationship lays the foundation for this rational design. In this work, a synergistic approach was taken to correlate the ink formulation, the microstructure of catalyst layers, and the resulting MEA performance to establish such a property-structure-performance relationship. The solvent composition (n-PA/H2O mixtures) demonstrated a strong influence on the performance of the MEA fabricated with an 830-EW (Aquivion) ionomer, especially polarization losses of cell activation and mass transport. The performance differences were studied in terms of how the solvent composition affects the catalyst/ionomer interface, ionomer network, and pore structure of the resulting catalyst layers. The ionomer aggregates mainly covered the surface of catalyst aggregates acting as oxygen reduction reaction active sites, and the aggregate sizes of the ionomer and catalyst (revealed by ultrasmall angle X-ray scattering and cryo-transmission electron microscopy) were dictated by tuning the solvent composition, which in turn determined the catalyst/ionomer interface (available active sites). In n-PA/H2O mixtures with 50∼90 wt % H2O, the catalyst agglomerates could be effectively broken up into small aggregates, leading to enhanced kinetic activities. The boiling point of the mixed solvents determined the pore structure of ultimate catalyst layers, as evidenced by mercury porosimetry and scanning electron microscopy. For mixed solvents with a higher boiling point, the catalyst-ionomer aggregates in the ink tend to agglomerate during the solvent evaporation process and finally form larger catalyst-ionomer aggregates in the ultimate catalyst layer, resulting in more secondary pores and thus lower mass transport resistance. Both the enlarged catalyst/ionomer interface and appropriate pore structure were achieved with the catalyst layer fabricated from an n-PA/H2O mixture with 90 wt % H2O, leading to the best MEA performance.

Our previous work demonstrated that the decorated Pt nano particles (PNP) distributed uniformly over the surface of the PBI-functionalized graphene. Considering the excellent chemical stability and conductivity, graphene are the promising catalyst supports for PEMFCs. However, the mass transfer issue is a huge challenge for the application of graphene as catalyst supports. In this paper, a method to overcome such an issue was proposed, in which the highly stable spacers were introduced during catalyst ink preparation to create channels/pores which facilitates the mass transfer, consequently, increase the high current density performance while maintain the excellent stability.

Atomically dispersed and nitrogen-coordinated single iron sites (FeN4) embedded in carbon (Fe-N-C) catalysts are the most promising platinum group metal (PGM)-free catalysts. However, they have yet to match their Pt counterparts for oxygen reduction reaction (ORR) activity and stability in proton exchange membrane fuel cells (PEMFCs). Here, we developed viable Fe-N-C catalysts, which, for the first time, demonstrated competitive activity to that of Pt/C catalysts and dramatically enhanced stability and durability under practical PEMFC operating conditions. The most active Fe-N-C catalyst achieved a record half-wave potential (E1/2 = 0.915 V vs. RHE at 0.6 mgcatcm-2) and an ORR mass activity of 10.5 mA mgcat at 0.9 V in (RDE) tests, exceeding a Pt/C baseline catalyst (60 µgPt cm-2) by 40 mV in acidic electrolytes. This compelling activity of the Fe-N-C catalyst in aqueous acids on rotating disk electrode (RDE) was successfully transferred to a fuel cell membrane electrode assemblies (MEAs), generating an initial current density of 44.2 mA cm-2 exceeding the U.S. DOE 2025 target (i.e., 44 mA cm-2) at 0.9 VIR-free under O2. Under practical hydrogen-air conditions, record 151 mA cm-2 at 0.8 V and peak power density of 601 mW cm-2 were achieved. Importantly, we discovered that depositing nitrogen-carbon species on the catalyst surface via chemical vapor deposition (CVD) dramatically enhanced catalyst stability, evidenced by performance durability after accelerated stress tests (30 000 square-wave voltage cycles under H2/air) and long-term steady-state life tests (> 300 hours at 0.67 V). Innovative identical location-scanning transmission electron microscopy (IL-STEM) experiments confirmed that the CVD process leads to deposition of nitrogen-doped carbon onto the catalyst surfaces. Along with theoretical modeling, a reconstruction of the carbon structure adjacent to FeN4 sites leads to increased robustness against demetallation and carbon oxidation. This work opens new avenues for developing earth-abundant iron-based catalysts with extraordinary activity and stability, thus competing with Pt and addressing the cost barrier of current PEMFCs.

Significantly reducing platinum group metal (PGM) loading while improving catalytic performance and durability is critical to accelerating proton-exchange membrane fuel cells (PEMFCs) for transportation. Here we report an effective strategy to boost PGM catalysts through integrating PGM-free atomically-dispersed single metal active sites in the carbon support toward the cathode oxygen reduction reaction (ORR). We achieved uniform and fine Pt nanoparticle (NP) (∼2 nm) dispersion on an already highly ORR-active FeN4 site-rich carbon (FeN4–C). Furthermore, we developed an effective approach to preparing a well-dispersed and highly ordered L12 Pt3Co intermetallic nanoparticle catalyst on the FeN4–C support. DFT calculations predicted a synergistic interaction between Pt clusters and surrounding FeN4 sites through weakening O2 adsorption by 0.15 eV on Pt sites and reducing activation energy to break O–O bonds, thereby enhancing the intrinsic activity of Pt. Experimentally, we verified the synergistic effect between Pt or Pt3Co NPs and FeN4 sites, leading to significantly enhanced ORR activity and stability. Especially in a membrane electrode assembly (MEA) with a low cathode Pt loading (0.1 mgPt cm−2), the Pt/FeN4–C catalyst achieved a mass activity of 0.451 A mgPt −1 and retained 80% of the initial values after 30 000 voltage cycles (0.60 to 0.95 V), exceeding DOE 2020 targets. Furthermore, the Pt3Co/FeN4 catalyst achieved significantly enhanced performance and durability concerning initial mass activity (0.72 A mgPt −1), power density (824 mW cm−2 at 0.67 V), and stability (23 mV loss at 1.0 A cm−2). The approach to exploring the synergy between PGM and PGM-free Fe–N–C catalysts provides a new direction to design advanced catalysts for hydrogen fuel cells and various electrocatalysis processes.

Fuel cells for heavy-duty vehicles (HDVs) have attracted considerable attention because of their unique scalability, better fuel economy, the less demand for hydrogen refilling infrastructure. However, the potential application requires more stringent fuel cell durability up to 25,000 h. Membrane electrode assemblies (MEAs) made from platinum group metal (PGM) catalyst with relatively high loading 0.3 mgPt cm−2 play a crucial role in ensuring high-power and long-term durability. Integrating fine PGM nanoparticles and robust carbon support with strengthened interactions is critical for improving MEA performance and durability. Herein, a unique atomically dispersed and nitrogen coordinated single Mn site-rich carbon (M–N–C) support was developed for high content (40 wt%) platinum catalysts for the oxygen reduction reaction (ORR) cathode with reduced thickness. Compared with two controls studied in this work (e.g., a porous graphitic carbon-supported Pt and a commercial TKK Pt/C catalysts), the Pt (40 wt%)/Mn–N–C catalyst exhibited much enhanced catalytic activity and stability for the ORR in both aqueous acidic electrolyte and polymer electrolyte-based MEA. We carefully elucidated the—role of the Mn–N–C support in promoting Pt catalyst concerning its high surface area, partially graphitic structure, and nitrogen dopants, providing better Pt nanoparticle dispersion, and strengthened interactions between Pt and carbon. Consequently, the MEA from the Pt (40 wt%)/Mn–N–C catalyst generated a 1.61 A cm−2 at 0.7 V based on HDV conditions (0.2 mgPt cm−2 and 250 kPa air). More importantly, the MEA is highly durable and can retain 1.31 A cm−2 at 0.7 V after 30,000 voltage cycles (∼19% loss), surpassing the commercial Pt/C catalyst (loss of ∼56%). Therefore, the Mn–N–C carbon-supported Pt catalyst holds a great promise to meet the challenging DOE target (1.07 A cm−2 at 0.7 V after 150,000 cycles) for HDVs.

Inference is one key objective in a Bayesian network (BN), and it aims to estimate the posterior distributions of state variables based on evidence (observations). While efficient analytical inference algorithms (either approximate or exact) for BN with discrete variables have been well-established in the literature, the inference in BN with continuous variables is still challenging if the BN is non-linear and/or non-Gaussian. In this case we can either discretize the continuous variable and utilize the inference approaches for discrete BN; or we have to use sampling-based methods such as MCMC for static BN and particle filter for dynamic BN. This paper proposes a network collapsing technique based on the concept of probability integral transform to convert a multi-layer BN to an equivalent simple two-layer BN, so that the unscented Kalman filter can be applied to the collapsed BN and the posterior distributions of state variables can be obtained analytically. For dynamic BN, the proposed method is also able to propagate the state variables to the next time step analytically using the unscented transform, based on the assumption that the posterior distributions of state variables are Gaussian. Thus the proposed method achieves a very fast approximate solution, making it particularly suitable for dynamic BN where inference and uncertainty propagation are required over many time steps.

Current airframe health monitoring generally relies on deterministic physics models and ground inspections. This paper uses the concept of a dynamic Bayesian network to build a versatile probabilis...

The reduction of platinum use and improvement of its catalytic performance has become the most critical steps to accelerate the development of proton-exchange membrane fuel cells (PEMFCs). Here we report a novel and facile method to boost the performance of the Pt-cased catalyst by integrating with an atomic platinum-group metal (PGM)–free active sites. Using iron-doped zeolitic imidazolate framework-8 (ZIF-8) derived carbon as support, we achieved a uniform deposition of Pt and intermetallic Pt-Co nanoparticles (NPs) through one-step synthesis, with average particle size 2.4 nm for Pt, 3.4 nm for L10 PtCo, and 4.2 nm for L12 Pt3Co. Besides, easy phase transfer during synthesis between ordered intermetallic structures L10 PtCo and L12 Pt3Co was achieved, and a comprehensive comparison between their catalytic performance was established. Synergistic catalysis between Pt or Pt-Co NPs over a PGM–free catalytic substrate derived from iron-doped ZIF-8 led to excellent oxygen reduction reaction (ORR) performance under both rotating disk electrode (RDE) and membrane electrode assembly (MEA) testing. Resulting pure Pt catalysts achieved ORR mass activities of 0.451 A/mgPt and retained 80% initial values after 30,000 voltage cycles in a fuel cell, superior to the DOE 2020 targets without using Pt alloy. Both PtCo and Pt3Co catalysts achieved better performance and exhibited all-sided excellence among mass activity, stability, and power density. Among them, Pt3Co reached the power density at 0.67 V of 923 mW/cm2 and retained 86% initial mass activity after 30,000 voltage cycles in a fuel cell. Their performance dependences on support material particle size and ionomer content were discussed and optimized in MEAs.

The proton conduction within a catalyst layer is one of the critical factors affecting the performance of membrane electrode assembly in polymer electrolyte fuel cells. In this work, a simple and effective approach for providing surface proton conduction over carbon supports was developed utilizing chemically grafting SO3H and COOH groups covalently bonded onto carbon surface. A method for accurately measuring the proton conductivity within a catalyst layer was also developed which physically excluded the conductivity contribution from the membrane. This method has several advantages: 1) providing easy sample preparation with high reproducibility, 2) allowing the control of measurement conditions such as gas flow rate, relative humidity and temperature to mimic the exact fuel cell operating conditions. The proton and electron conductivities of the catalyst layers made with and without functional groups at different relative humidifies were characterized with a specially designed 4-probe cell using the AC impedance spectroscopy technique. The results clearly demonstrated that the introduction of functional groups (i.e., SO3H and COOH) did result in a significant improved surface proton conduction over carbon surface and the resulted conductivity depends on the relative humidity, temperature and porosimetry of the carbon blacks.

Sobol׳ index is a prominent methodology in global sensitivity analysis. This paper aims to directly estimate the Sobol׳ index based only on available input–output samples, even if the underlying model is unavailable. For this purpose, a new method to calculate the first-order Sobol׳ index is proposed. The innovation is that the conditional variance and mean in the formula of the first-order index are calculated at an unknown but existing location of model inputs, instead of an explicit user-defined location. The proposed method is modularized in two aspects: 1) index calculations for different model inputs are separate and use the same set of samples; and 2) model input sampling, model evaluation, and index calculation are separate. Due to this modularization, the proposed method is capable to compute the first-order index if only input–output samples are available but the underlying model is unavailable, and its computational cost is not proportional to the dimension of the model inputs. In addition, the proposed method can also estimate the first-order index with correlated model inputs. Considering that the first-order index is a desired metric to rank model inputs but current methods can only handle independent model inputs, the proposed method contributes to fill this gap.

Calibration of model parameters is an essential step in predicting the response of a complicated system, but the lack of data at the system level makes it impossible to conduct this quantification directly. In such a situation, system model parameters are estimated using tests at lower levels of complexity which share the same model parameters with the system. For such a multi-level problem, this paper proposes a methodology to quantify the uncertainty in the system level prediction by integrating calibration, validation and sensitivity analysis at different levels. The proposed approach considers the validity of the models used for parameter estimation at lower levels, as well as the relevance at the lower level to the prediction at the system level. The model validity is evaluated using a model reliability metric, and models with multivariate output are considered. The relevance is quantified by comparing Sobol indices at the lower level and system level, thus measuring the extent to which a lower level test represents the characteristics of the system so that the calibration results can be reliably used in the system level. Finally the results of calibration, validation and relevance analysis are integrated in a roll-up method to predict the system output.

This paper develops a novel computational framework to compute the Sobol indices that quantify the relative contributions of various uncertainty sources towards the system response prediction uncertainty. In the presence of both aleatory and epistemic uncertainty, two challenges are addressed in this paper for the model-based computation of the Sobol indices: due to data uncertainty, input distributions are not precisely known; and due to model uncertainty, the model output is uncertain even for a fixed realization of the input. An auxiliary variable method based on the probability integral transform is introduced to distinguish and represent each uncertainty source explicitly, whether aleatory or epistemic. The auxiliary variables facilitate building a deterministic relationship between the uncertainty sources and the output, which is needed in the Sobol indices computation. The proposed framework is developed for two types of model inputs: random variable input and time series input. A Bayesian autoregressive moving average (ARMA) approach is chosen to model the time series input due to its capability to represent both natural variability and epistemic uncertainty due to limited data. A novel controlled-seed computational technique based on pseudo-random number generation is proposed to efficiently represent the natural variability in the time series input. This controlled-seed method significantly accelerates the Sobol indices computation under time series input, and makes it computationally affordable.

This research is concerned with how to use available experimental data from tests of lower complexity to inform the prediction regarding a complicated system where no test data is available. Typically, simpler test configurations are used to infer the unknown parameters of an engineering system. Then the calibration results are propagated through the system model to predict the uncertainty in the system response. However, it is important to note that parameter estimation results are affected by the quality of the model used to represent the test configuration. Therefore, it is necessary that the model of the test configuration be also subjected to rigorous validation testing. Then the calibration and validation results for the test configurations need to be integrated to produce the distributions of the parameters to be used in the system-level prediction. Such a systematic roll-up methodology that integrates calibration and validation results at multiple levels of test configurations has been previously established (Sankararaman and Mahadevan, Reliab Eng Syst Saf 138:194–209, 2015).

In a Bayesian network (BN), how a node of interest is affected by the observation at another node is a main concern, especially in backward inference. This challenge necessitates the proposed global sensitivity analysis (GSA) for BN, which calculates the Sobol’ sensitivity index to quantify the contribution of an observation node toward the uncertainty of the node of interest. In backward inference, a low sensitivity index indicates that the observation cannot reduce the uncertainty of the node of interest, so that a more appropriate observation node providing higher sensitivity index should be measured. This GSA for BN confronts two challenges. First, the computation of the Sobol’ index requires a deterministic function while the BN is a stochastic model. This paper uses an auxiliary variable method to convert the path between two nodes in the BN to a deterministic function, thus making the Sobol’ index computation feasible. Second, the computation of the Sobol’ index can be expensive, especially if the model inputs are correlated, which is common in a BN. This paper uses an efficient algorithm proposed by the authors to directly estimate the Sobol’ index from input–output samples of the prior distribution of the BN, thus making the proposed GSA for BN computationally affordable. This paper also extends this algorithm so that the uncertainty reduction of the node of interest at given observation value can be estimated. This estimate purely uses the prior distribution samples, thus providing quantitative guidance for effective observation and updating.

As an important tool for knowledge representation and decision-making under uncertainty, Dempster-Shafer evidence theory (D-S theory) has been used in many fields. The application of D-S theory is critically dependent on the availability of the basic probability assignment (BPA). The determination of BPA is still an open issue. A non-parametric method to obtain BPA is proposed in this paper. This method can handle multi-attribute datasets in classification problems. Each attribute value of the dataset sample is treated as a stochastic quantity. Its non-parametric probability density function (PDF) is calculated using the training data, which can be regarded as the probability model for the corresponding attribute. The BPA function is then constructed based on the relationship between the test sample and the probability models. The missing attribute values in datasets are treated as ignorance in the framework of the evidence theory. This method does not have the assumption of any particular distribution. As a result, it can be flexibly used in many engineering applications. The obtained BPA can avoid high conflict between evidence, which is desired in data fusion. Several benchmark classification problems are used to demonstrate the proposed method and to compare against existing methods. The constructed classifier based on the proposed method compares well to the state-of-the-art algorithms.

Lithium-ion batteries (LIBs) are the major power sources for electronic and transportation applications due to their high specific energy, good cyclability and environmental friendliness1, 2 Intercalation compounds were commercially employed as the positive (cathode) electrode material. However, they have a limited capacity due to the mono-valence change of host materials and the accompanied 1-e- transfer process, which accommodates 1-Li ion intercalation. These materials can’t meet the growing demand on higher specific energy and energy/power density. To achieve high specific energy cathode materials, efforts are being made to find alternative cathode materials for LIBs: (1) materials with transition metal ions capable of multi valence changes, and (2) materials with high potentials (vs. Li/Li+). Compared with intercalation cathode materials, conversion cathode materials can store multiple electrons instead of 1 electron of the intercalation materials. However, the conversion materials have not been used as practical LIB cathode materials due to the low electric conductivity, and structural stability. Iron oxyfluoride (FeOF) was proposed3, 4 as a promising candidate because its high theoretical specific capacity of 885 mAh/g, leading to an exceptionally high theoretical specific energy of 2938 Wh/kg for 3- electron reactions. Graphene was employed to solve the low conductivity and structural stability. As the results from the improved FeOF structure, the nanostructured FeOF with the incorporated graphene sheets shows the superior performance to its blank. The FeOF/G initial specific capacity reached 621 mAh/g (Fig. 1) with an initial Coulombic efficiency of 94.7%, while the blank FeOF initial specific capacity can reach 586 mAh/g but has a drastically low Coulombic efficiency of 38.7%. SEM and TEM (Fig. 2) were employed to observe the morphology. The structure evolution during charge/discharge process were also characterized by in situ XAS and high resolution XRD. The results showed that the graphene nano-sheets serve as substrates to stabilize the structure of FeOF and form a framework to stabilize the discharged species through bonding them to their original sites without migration, in which the graphene sheets serve as a “cage” to trap the discharged species. Thus, the FeOF/Graphene composite can keep the (de)lithiation reaction reversible during discharge and charge process and could be a promising cathode material for Li-ion batteries. Fig. 1 Charge/discharge curves of FeOF (BLK) and FeOF/graphene (GRP). Fig. 2 SEM image (a) and TEM image (b) of FeOF/Graphene hybrid material. J. B. Goodenough and K.-S. Park, Journal of the American Chemical Society, 2013, 135, 1167-1176. J.-M. Tarascon and M. Armand, Nature, 2001, 414, 359-367. A. Kitajou, H. Komatsu, R. Nagano and S. Okada, Journal of Power Sources, 2013, 243, 494-498. K. M. Wiaderek, O. J. Borkiewicz, E. Castillo-Martínez, R. Robert, N. Pereira, G. G. Amatucci, C. P. Grey, P. J. Chupas and K. W. Chapman, Journal of the American Chemical Society, 2013, 135, 4070-4078. Figure 1

Graphene is a single atomic-tick graphite layer. The carbon atoms in the graphene plane are saturated due to the bonding structure: each has 3 covalent bonds with three neighboring carbon atoms in the graphene plane, the lone px e- form a long range conjugated system, a big π bond. This leads to some unique properties: excellent electric conductivity and chemical stability, which are critically needed for catalyst supports in polymer electrolyte membrane fuel cells (PEMFCs).[1] The catalyst support stability plays a significant role in determining the overall durability of PEMFC systems when PEMFCS are subjected to very high potential cycling , for example, from 1 V to 1.5 V. However, there are three major obstacles for graphene supported Pt based catalysts in PEMFCs: 1) there are limited bonding sites for catalyst landing on the graphene basal plane, which lead to migration/aggregations of Pt nanoparticles when subjected to harsh accelerated durability tests (ADTs); 2) it is very difficult for Nafion ionomer particles and graphene based catalyst mixing together to form a uniform catalyst ink, which results in the poor catalyst/ionomer interface. 3) For MEA preparation using ink spraying method, when ink is dried, the π-π interaction makes 2D graphene sheets easily restacking back to graphite structure, which could severely block the mass transport of gas and water, leading to poor MEA performance at high current. To overcome these barriers, we propose a novel approach to transfer the 2D graphene sheets into 3D composite with sufficient channels and pores for facile mass transport. This is realized by developing the highly stable hierarchical polybenzimidazole (PBI) -grafted nano-graphene supported Pt catalysts for PEMFCs and applying spacers during ink formulation. Nano- graphene, comparing with normal graphene, has smaller dimensions as the catalyst supports, which makes the pores/channels between graphene sheets much shorter, facilitating the mass transport. It is expected to see some extra voltage gains at the high current density. In order to construct appropriate pore structures in the catalyst layers, spacers are introduced. However, although the initial performance somehow increases with such modification, yet the support durability remains same and to some extents, decrease little bit. In this case, the hypothesis raised that during durability test, the spacers themselves are corroded and graphene sheets are restacked. To overcome this barrier, we apply more stable spacer that match the durability of graphene. Nano-silicon powder is very stable during voltage scan between 1.0V to 1.5V but is not conductible. To remedy this drawback, we coat highly graphitized carbon over its surface, which gives very promising durability and conductivity for overall MEA electrode. [1]. Geim, A. K.; Novoselov, K. S. Nat. Mater. 2007, 6, 183

Proton exchange membrane fuel cells (PEMFCs) are promising energy conversation devices due to their high power/energy density, high efficiency and the absence of carbon dioxide emission. Yet, one of the largest obstacles for the commercial application of PEMFC is the degradation of platinum group metal (PGM)-based electrocatalysts during long term operation. Four degradation mechanisms have been widely reported, namely, 1) Ostwald ripening; 2) particle migration and coalescence; 3) carbon corrosion and 4) Pt dissolution and reprecipitation in the membrane. However, the role played by each degradation mechanism is not fully understood. In this work, scanning transmission electron microscopy (STEM) tomography coupled with identical location TEM (IL-TEM) are utilized to delineate the roles of each degradation mechanism for Pt catalysts on Vulcan XC72 carbon support. Compared with conventional TEM, 3D STEM tomography is capable of identifying the following: 1) interparticle distances; 2) 3D structure of carbon black, 3) specific positions of Pt catalysts on carbon support; and 4) shapes and morphologies of Pt catalysts[1]. In addition, IL-TEM allows a comparison of Pt catalysts and carbon support of the same region of interest (R.O.I), at the beginning of life and after potential cycling, which provides direct insight into the effects of potential cycling on the degradation of Pt catalysts. Figure 1(a-b) shows the 3D illustration of Pt catalysts on the Vulcan XC72 carbon support before and after potential cycling, respectively. We observe that volume shrinkage is insignificant in the bulk of the carbon black particles and that there are no Pt catalysts falling off after potential cycling (0.6V-1.0V vs RHE, 6000 cycles). In addition, we note the appearance of newly-formed Pt nanoparticles or ionic clusters (0.5nm). Instead, the growth of Pt nanoparticles result from Ostwald ripening and subsequent sintering. In summary, 3D STEM tomography coupled with IL-TEM (~1nm spatial resolution) provides unique insight into the roles of each degradation mechanisms for Pt electrocatalysts on Vulcan XC72 during potential cycling. Reference: [1]Horn, S. Y.; Sheng, W. C.; Chen, S.; Ferreira, P. J.; Holby, E. F.;Morgan, D. Instability of Supported Platinum Nanoparticles in Low-Temperature Fuel Cells. Top. Catal. 2007, 46, 285-305. Figure 1

In a Bayesian network, how a node of interest is affected by the observation of another node is of interest in both forward propagation and backward inference. The proposed global sensitivity analysis (GSA) for Bayesian network aims to calculate the Sobol’ sensitivity index of a node with respect to the node of interest. The desired GSA for Bayesian network confronts two challenges. First, the computation of the Sobol’ index requires a deterministic function while the Bayesian network is a stochastic model. Second, the computation of the Sobol’ index can be expensive, especially if the model inputs are correlated, which is common in a Bayesian network. To solve the first challenge, this paper uses the auxiliary variable method to convert the path between two nodes in the Bayesian network to a deterministic function, thus making the Sobol’ index computation feasible in a Bayesian network. To solve the second challenge, this paper proposes an efficient algorithm to directly estimate the first-order Sobol’ index from Monte Carlo samples of the prior distribution of the Bayesian network, so that the proposed GSA for Bayesian network is computationally affordable. Before the updating, the proposed algorithm can predict the uncertainty reduction of the node of interest purely using the prior distribution samples, thus providing quantitative guidance for effective observation and updating.

To predict the response of a system with unknown parameters, a common route is to quantify the parameters using test data and propagate the results through a computational model of the system. Activities in this process may include model calibration and/or model validation. Test data value uncertainty has a significant effect on model calibration and model validation, and therefore affects the response prediction. Limited testing budget creates the challenge of test resource allocation, i.e., how to optimize the number of calibration and validation tests to be conducted. In this paper, a novel computational technique based on pseudo-random numbers is proposed to efficiently quantify the uncertainty in the data value of each type of test. This technique helps quantifying the contribution of data value uncertainty to the uncertainty in the prediction through Sobol indices. Consistent predictions using different sets of data are expected if this contribution is small. Then the numbers of each type of test are optimized to minimize this contribution. A simulated annealing algorithm is applied to solve this discrete optimization problem.

To predict the response of a system with unknown parameters, a common route is to quantify the parameters using test data and propagate the results through a computational model of the system. Activities in this process may include model calibration and/or model validation. Data uncertainty has a significant effect on model calibration and model validation, and therefore affects the response prediction. Data uncertainty includes the uncertainty regarding the amount of data and numerical values of data. Although its effect can be qualitatively observed by trying different data sets and visually comparing the response predictions, a quantitative methodology assessing the contributions of these two types of data uncertainty to the uncertainty in the response prediction is necessary in order to solve test resource allocation problems. In this paper, a novel computational technique based on pseudo-random numbers is proposed to efficiently quantify the uncertainty in the data value of each type of test. Then the method of auxiliary variable based on the probability integral transform theorem is applied to build a deterministic function so that variance-based global sensitivity analysis can be conducted. The resultant global sensitivity indices quantify the contribution of data value uncertainty of each type of test to the uncertainty in the response prediction. Thus a methodology for robust test resource allocation is proposed, i.e., quantifying the number of each type of tests so that the response predictions using different data set are consistent.

In the preparation process for scanning electron microscopy (SEM), flexed silkworm embryos typically assume several curled shapes with irregular postures that obscure morphological details during SEM observation. We describe a preparation technique based on glycerol substitution for better SEM visualization of straight and flat silkworm embryos. Glycerol has high viscosity, low vapor pressure, and sufficient electrical conductivity. Silkworm embryos were infiltrated with glycerol and arranged in a straight posture or flattened using a cover slip. Samples were directly observed by SEM without additional dehydration, drying, or coating procedures. The complete ventral side could be easily viewed in one image. Recoating alleviated the charging phenomenon. This represents a simple method for preparation of straight and flat samples from curled biological specimens for SEM observation.

The calibration of model parameters is essential to predict the output of a complicated system, but the lack of data at the system level makes it impossible to conduct this quantification directly. This situation drives analysts to obtain information on model parameters using experimental data at lower levels of complexity which share the same model parameters with the system of interest. To solve this multi-level problem, this paper first conducts model calibration using lower level data and Bayesian inference to obtain the posterior distribution of each model parameter. However, lower level models are not perfect; thus model validation is also needed to evaluate the model that was used in model calibration. In the model validation, this paper extends the model reliability metric by using a stochastic representation of model reliability, and model with multivariate output is also considered. Another contribution of this paper is the consideration of physical relevance through sensitivity analysis, in order to measure the extent to which a lower level test represents the physical characteristics of the actual system of interest so that the calibration results can be extrapolated to the system level. Finally all the information from calibration, validation and relevance analysis is integrated to quantify the uncertainty in the system level prediction.