Ballard Power Systems (Canada)

Enhancing the catalytic stability of Fe-N-C catalysts for cathodic oxygen reduction in proton-exchange membrane fuel cells (PEMFCs) necessitates an in-depth understanding of their degradation mechanisms. This study identifies key stressors affecting the stability of Fe-N-C catalysts, specifically acidic environment, oxygen (O2), and reactive oxygen species (ROS). Through ex situ/operando experiments, we show that the oxidation of local carbon by acidic environment + O2 + ROS, along with the demetalation of catalytic FeNxCy sites by O2 or O2 + ROS, is the primary factor responsible for the initial fast degradation of Fe-N-C catalysts. The demetalation of FeNxCy sites, influenced by O2, in particular by O2 + ROS, leads to the subsequent gradual degradation of Fe-N-C. Notably, FeN4C12-type active sites are more susceptible to demetalation than FeN4C10-type sites in O2 or O2 + ROS. Our findings indicate that, besides constructing more stable FeNxCy sites, preventing local carbon oxidation and scavenging of ROS are all critical for maintaining the stability of Fe-N-C catalysts.

The variability of electricity prices, driven by demand and supply fluctuations, creates challenges in deregulated markets. This paper proposes a two-stage probabilistic framework to address these issues. The first stage employs Structured Gated Recurrent Units (SGRU), an enhancement over conventional GRUs, to improve spatial-temporal feature extraction through structured layers and nonlinear connections. The second stage introduces a nonparametric maximum likelihood estimator (NMLE) for probability density function (PDF) estimation, offering smooth and accurate density predictions without assuming a predefined distribution. The numerical results of the actual data of California independent system operators (CAISO) within a $30-\mathrm{min}$ time resolution demonstrate the effectiveness and superiority of the proposed framework by comparing it with several state-of-the-art models and one PDF estimator (kernel density estimator (KDE)).

The European Union has established ambitious targets for lowering carbon dioxide emissions in the residential sector, aiming for all new buildings to be “zero-emission” by 2030. Integrating solar generators with hydrogen storage systems is emerging as a viable solution for achieving these goals in homes. This paper introduces a linear programming optimization algorithm aimed at improving the installation capacity of residential solar–hydrogen systems, which also utilize waste heat recovery from electrolyzers and fuel cells to increase the overall efficiency of the system. Analyzing six European cities with diverse climate conditions, our techno-economic assessments show that optimized configurations of these systems can lead to significant net present cost savings for electricity and heat over a 20-year period, with potential savings up to EUR 63,000, which amounts to a 26% cost reduction, especially in Southern Europe due to its abundant solar resources. Furthermore, these systems enhance sustainability and viability in the residential sector by significantly reducing carbon emissions. Our study does not account for the potential economic benefits from EU subsidies. Instead, we propose a novel incentive policy that allows owners of solar–hydrogen systems to inject up to 20% of their total solar power output directly into the grid, bypassing hydrogen storage. This strategy provides two key advantages: first, it enables owners to profit by selling the excess photovoltaic power during peak midday hours, rather than curtailing production; second, it facilitates a reduction in the size—and therefore cost—of the electrolyzer.

Visual anomaly detection plays a crucial role in manufacturing to ensure product quality by identifying image patterns that deviate from the expected ones. Existing methods that rely on distribution estimation struggle with the complexity of real-world images, resulting in complex and inefficient procedures. This study leverages normalizing flow techniques to address the cold start anomaly detection problem, where no prior examples of anomalies are available during the training phase. In such scenarios, models must learn exclusively from defect-free images and still accurately identify anomalies. We propose a novel unsupervised multi-scale and multi-semantic normalizing flow model, enhanced with an ensemble of neural networks, to detect anomalies based on their feature distributions. Our model estimates the likelihood of non-defective features, identifying anomalies as out-of-distribution values. Extensive experiments on three state-of-the-art anomaly detection datasets demonstrate that our proposal achieves superior AUROC performance and improves computational efficiency compared to existing approaches. Furthermore, we validate the robustness and adaptability of our proposal through low-shot training experiments using only 20% of available training data, highlighting its potential as an efficient solution for cold start anomaly detection.

Studies on the hydration properties, proton conductivity, and water content of perfluorinated ionomer thin films at temperatures relevant to fuel cell operation temperatures (around 80 °C) and the effect of ionomer chemistry are scarce. In this work, we report the water content and proton conductivity properties of thin-film ionomers (30 nm) at 80 °C over a wide range of relative humidity (0-90%) for seven different ionomers differing in the side-chain structure, including the number of protogenic groups, with the equivalent weight ranging from 620 to 1100 g/mol of sulfonic acid. The results show that the acid content or equivalent weight of the ionomer is the strongest determinant of both the swelling and the proton conductivity of ionomer films at a given relative humidity. The molar water content (λ) of ionomer films normalized to the molar protogenic group is observed to be equivalent-weight-dependent, implying that the affinity for water is acid-content-dependent. At high relative humidity conditions (>70%) pertinent to fuel cell operations, the proton conductivity of low-equivalent-weight ionomers was higher than that of higher-equivalent-weight ionomers. However, upon correlating the proton conductivity with molar water content (λ), the differences reduce dramatically, highlighting that water content is the controlling factor for proton conduction. Significantly higher values of both water content and proton conductivity are observed at 80 °C compared to those at 30 °C, implying that room temperature data are not reliable for estimating ionomer properties in the fuel cell catalyst layer.

The membrane electrolyte assembly (MEA) is the key component of a proton exchange membrane fuel cell (PEMFC). The MEA usually consists of gas diffusion layers as outer layers to the inner catalyst coated membrane (CCM). There are various methods to prepare CCMs and decal transfer is currently a common method which was first introduced by Wilson and Gottesfeld [1] and further developed by other researchers [2]–[4]. The initial step of this method is that the catalyst ink is coated onto an intermediate substrate material creating a catalyst coated film (CCF). This step is followed by a transfer of the active layer onto the membrane by hot pressing the membrane and CCF between heated press platens maintained at a specific high temperature for an optimized time and pressure to ensure bonding.Extensive studies on the procedure of decal transfer, substrate material and its preparation, catalyst solvents, etc. have increased the transfer yield of decal transfer to more than 95% [5]. Additionally, in the last decade, industrial development has reduced its cost; however, the total cost for fuel cell manufacturing is still relatively high. One reason for this high production cost is the contamination sensitivity of the CCMs by external particles that can potentially be detrimental to fuel cell operation [6]. This requirement necessitates the use of cleanrooms and quality control equipment to prevent contaminants entering the manufactured components, thus increasing the overall costs. The purpose of the present research is to improve the understanding of the interactions between foreign contaminants and the manufacturing process in the context of MEA production quality. The specific objectives are to i) understand the impact of external particles on the catalyst layer decal transfer process and ii) improve the robustness of the decal transfer method in the presence of external particles.This research investigates the impact of incidental external particles that may be found on the membrane surface or the CCF prior to hot pressing. 60µm Silica microspheres (Si-M) were selected as a representative of solid particles and CCMs with purposely introduced Si-Ms were fabricated and several decal transfer methods and support materials were tested and imaged using the X-ray computed tomography (XCT) technique to analyze the impact on the fuel cell integrity and operation at the presence of the Si-M. While regular decal transfer protocols result in excessive membrane thinning in the presence of these particles, it was observed that by changing the rate of applied pressure and using alternative support materials when transferring the cathode catalyst onto a half-coated membrane, it was possible to reduce membrane thinning under the particles by more than 20% while maintaining transfer quality and cell performance (Fig. 1). In addition, it was observed that although anisotropic mechanical properties of the membrane can adversely affect the CCM topography after the decal transfer, a tuned protocol is able to prevent unwanted deformations and potential stress concentrations. Overall, it is envisioned that the outcomes of this work may enable relaxed quality control measures and manufacturing site cleanroom standards by reducing the potential effects of external particles on MEA production quality. 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. This research was undertaken, in part, thanks to funding from the Canada Research Chairs program. Keywords: fuel cell; membrane durability; X-ray computed tomography; decal transfer; manufacturing References Wilson, M. S. & Gottesfeld, S. Thin-film catalyst layers for polymer electrolyte fuel cell electrodes. J. Appl. Electrochem. 22, 1–7 (1992).Shahgaldi, S., Alaefour, I. & Li, X. Impact of manufacturing processes on proton exchange membrane fuel cell performance. Appl. Energy 225, 1022–1032 (2018).Cho, H. J. et al. Development of a novel decal transfer process for fabrication of high-performance and reliable membrane electrode assemblies for PEMFCs. Int. J. Hydrogen Energy 36, 12465–12473 (2011).Thanasilp, S. & Hunsom, M. Effect of MEA fabrication techniques on the cell performance of Pt-Pd/C electrocatalyst for oxygen reduction in PEM fuel cell. Fuel 89, 3847–3852 (2010).Liang, X., Pan, G., Xu, L. & Wang, J. A modified decal method for preparing the membrane electrode assembly of proton exchange membrane fuel cells. Fuel 139, 393–400 (2015).James, B. D., Moton, J. M. & Colella, W. G. Mass Production Cost Estimation of Direct H2 PEM Fuel Cell Systems for Transportation Applications: 2018 Update. ASME 2014 12th Int. Conf. Fuel Cell Sci. Eng. Technol. collocated with ASME 2014 8th Int. Conf. Energy Sustain. V001T07A002–V001T07A002 (2018). Figure 1

This paper presents results from a set of experiments performed on a Ballard air-cooled PEM stack equipped with an Electrochemical Impedance Spectroscopy (EIS) system that permits measuring impedance over different cell-counts. The EIS scans were collected over a single cell as well as a collection of 2/5/10/20/35 & 50 cells spanning the original single cell. The intent of this research testing was to first understand how the EIS of single-and multiple-air-starved cells 'add' in a stack of mostly 'good' cells, and then to learn the required conditions for detecting significant internal transfers or hydrogen crossover within an air-cooled stack. Nitrogen was injected into the air stream entering the target cell at various rates to dilute the oxygen on the cathode side to emulate the phenomena of crossover recombination. It was found that EIS signatures of healthy cells within a stack are fully additive. Furthermore, adding a single cell with a relatively low oxygen concentration resulting from higher nitrogen injection rates generated large low-frequency arcs that showed even in scans of one air-starved cell in a large pool of healthy cells.

EDITORIAL article Front. Chem., 09 July 2020 | https://doi.org/10.3389/fchem.2020.00576