Stack Assembly Research Articles

Investigation of the assembly for high-power proton exchange membrane fuel cell stacks through an efficient equivalent model

A high-power proton exchange membrane fuel cell (PEMFC) stack usually composes many cells, which induce high difficulty in evaluating its mechanical state of stack assembly. A methodology based on composite model and material property equivalent is developed to predict the mechanical state in PEMFC stack. In this methodology, the stack system is modeled based on a finite element model (FEM), in which bipolar plate (BPP) and membrane exchange assembly (MEA) are combined into a composite component. Experiments with stamped BPP are carried out to validate the FEM of the stack, and both the predicted clamping force and endplates deformation of FEM have a great agreement with the experimental results. Based on the methodology, it is found that more uniform pressure distribution can be generated when high-stiffness endplates are applied and cell number of the stack increases. The cells approaching to mid-stack have more uniform pressure. The pressure distribution of the stack is very sensitive to the compression ratio. High compression ratios leads to large endplate deformation, which increases the average pressure deviation between simulated value and design value, and also increases non-uniformity of pressure distribution. This methodology offers the possibility of evaluating the mechanical state of high-power fuel cell stack and greatly improves the computational efficiency.

Robustness to Large-Scale Mass Production Manufacturing Tolerances by Means of Sensitivity and Statistical Analysis for IPMSMs

This article deals with highlighting some pragmatic approaches for determining the key-parameters that one has to monitor in order not to jeopardize the fulfillment of the customer specifications, due to the tolerances, that occur in large-scale mass production. For that, two parametrization models (symmetric vs. asymmetric) and two types of parameter distributions (uniform vs. normal) are compared when considering lamination, stack assembly and material properties tolerances. Results in different scenarios for both the motor (in open-circuit and under load), and the whole drive (with and without position sensor detection errors) are provided and compared with end-of-line measurements (where available) for a 12-slots 10-poles IPM-type PMSM with fractional slot concentrated winding. After including the effect of the temperature in the full asymmetric model, rather good agreements between calculations and measurements are reported for back-EMF and torque constant. Furthermore, it is shown that the interaction between the uncertainties of the motor parameters and the rest of the drive's components (e.g. position sensor) cause large inaccuracies into the system, and they are hardly predictable by taking the individual deviations separately. Finally, towards the end, attempts are made to reduce the computational and the modeling effort using various techniques and ultimately fitted artificial neural networks (ANNs) as surrogates.

Dataset of a vanadium redox flow battery 10 membrane-electrode assembly stack.

This paper contains a vanadium redox flow battery stack with an electrode surface area 40 cm2 test data. The aim of the study was to characterize the performance of the stack of the original design. The dataset include three series of galvanostatic charge-discharge cycling in the potential region 8–16 V with current densities 75, 150 and 200 mA/cm2 for 100 cycles. Coulomb, voltaic, energy efficiencies and capacity utilization coefficient are also provided for all three series.

Photoresponsive characteristics of thin film transistors with perovskite quantum dots embedded amorphous InGaZnO channels**Project supported by the National Natural Science Foundation of China (Grant No. 61874029) and the National Key Technologies R&D Program of China (Grant No. 2015ZX02102-003).

Photodetectors based on amorphous InGaZnO (a-IGZO) thin film transistor (TFT) and halide perovskites have attracted attention in recent years. However, such a stack assembly of a halide perovskite layer/an a-IGZO channel, even with an organic semiconductor film inserted between them, easily has a very limited photoresponsivity. In this article, we investigate photoresponsive characteristics of TFTs by using CsPbX3 (X = Br or I) quantum dots (QDs) embedded into the a-IGZO channel, and attain a high photoresponsivity over 103A⋅W−1, an excellent detectivity in the order of 1016 Jones, and a light-to-dark current ratio up to 105 under visible lights. This should be mainly attributed to the improved transfer efficiency of photoelectrons from the QDs to the a-IGZO channel. Moreover, spectrally selective photodetection is demonstrated by introducing halide perovskite QDs with different bandgaps. Thus, this work provides a novel strategy of device structure optimization for significantly improving the photoresponsive characteristics of TFT photodetectors.

Characterizing Stack Degradation in Proton-Conducting Ceramic Fuel Cells and Electrolyzers

We present our efforts to scale-up proton-conducting ceramic fuel cells (PCFC) for integration into small multi-cell stacks. Proton-conducting ceramics are an exciting new class of materials that are now emerging from research laboratories to address societal challenges in electricity generation, energy storage, and fuels synthesis. While there are many technologies under investigation, intermediate temperature proton-conducting fuel cells (PCFC) have been drawing attention due to their high performance at low temperatures in fuel cell mode (up to ~2 Wcm2 at 600 ºC) and their high efficiency (90 – 98 % Faradic efficiency) when operated in electrolysis mode. However despite the impressive achievements at lab-scale PCFC technology limited investments have been made on the scaling up process due to the challenges that the PCFC fabrication procedure represents. In fact, there are no stack demonstrations reported to date in the archival literature. In this presentation we will review our efforts at Colorado School of Mines to increase the size of PCFCs beyond the button-cell level, and to integrate these cells into multi-cell stack assemblies. The stability of our PCFC stack reached the degradation rates of 3.5% and 2.1% per 1000 hours, respectively, under fuel cell and electrolysis modes.An illustration of our stack design is shown in Figure 1, and includes a photograph of a fully assembled three-cell stack after performance testing. The design is centered on three repeating components: the protonic-ceramic electrolysis cells, the composite-ceramic frames in which the cells are bonded, and the thin metallic interconnect / bipolar plates that conduct electricity between adjacent cells. The cells feature a dense, 20 μm-thick BaCe0.4Zr0.4Y0.1Yb0.1O3- d (BCZYYb4411) electrolyte on a porous Ni-BCZYYb4411 anode fabricated through reactive-sintering methods. The cathode is a novel BaCo0.4Fe0.4Zr0.1Y0.1O3-δ (BCFZY) atop of the BCZYYb4411 electrolyte with the maximum active area of ~ 7cm2 per single unit cell. Reactive gases are fed through fuel and oxidizer ports machined into the frame. The metallic interconnect is a ferritic steel developed for intermediate-temperature solid-oxide fuel cells and electrolyzers, and is compression sealed to the frame. Metallic meshes (not shown) electrically connect the electrodes to the interconnects.High protonic-ceramic cell performance operating at 550 0C has been previously demonstrated at Colorado School of Mines. While the lower operating temperature should present benefits for the long-term stability of the stack it also brings questions regarding the behavior of stack materials such as the interconnects which represent a source of degradation in performance depending on the composition of the scale that is grown on them. In an effort to better understand the root causes behind performance degradation processes in our protonic-ceramic stacks, we have integrated numerous voltage-taps throughout the stack shown in Figure 1. In this presentation, we will review our results for long-term protonic-ceramic stack operation under both fuel cell and electrolysis mode, discuss after-test interconnects analysis showing high-iron-content scale and present strategies for reducing degradation rates and extending stack life. Figure 1

Versatile Pt(II) Pyrazolate Complexes: Emission Tuning via Interplay of Chelate Designs and Stacking Assemblies.

Three homoleptic Pt(II) metal complexes [Pt(imPz)2] (1), [Pt(imiz)2] (2), and [Pt(imMz)2] (3) were synthesized from the treatment of Pt(DMSO)2Cl2 and functional imidazolyl pyrazole in refluxing tetrahydrofuran (THF). Alternatively, the heteroleptic Pt(II) complexes [Pt(imPz)(fppz)] (4), [Pt(imiz)(fppz)] (5), and [Pt(imMz)(fppz)] (6) were obtained from the treatment of a common intermediate [Pt(fppzH)Cl2] with a corresponding imidazolyl chelate. Pt(II) complexes 1, 2, and 5 were studied by single-crystal X-ray diffraction to reveal the corresponding packing arrangement in their crystal lattices, among which both homoleptic complexes 1 and 2 formed monomeric species, while heteroleptic 5 aligned as a dimer with a nonbonding Pt···Pt contact of 3.574 Å. Subsequent photophysical examinations showed that the homoleptic 1-3 and heteroleptic 4-6 exhibited the structured sky-blue ππ* emission and structureless light-green-emitting metal-metal-to-ligand charge transfer (MMLCT) emission in the solid state, respectively. A shortened Pt···Pt interaction of approximately 0.34-0.35 nm was confirmed in thin films of all heteroleptic Pt(II) complexes 4-6 by grazing-incidence X-ray diffraction (GIXD) analyses. Finally, Pt(II) complex 6 was employed as a dopant in the fabrication of organic light-emitting diode (OLED) devices with varied doping ratios, among which OLEDs with only 1 wt % 6 in the SimCP host exhibited a maximum external quantum efficiency (EQE) of 5.8% and CIEx,y values of 0.20, 0.31. In contrast, OLEDs using a nondoped architecture (i.e., 100% of 6 in the emitting layer (EML)) achieved a maximum EQE of 26.8%, current efficiency (CE) of 91.7 cd A-1, and power efficiency (PE) of 80.1 lm·W-1 and CIEx,y values of 0.41, 0.55, manifesting their versatility in various degrees of stacking assemblies and hence facile color-tuning capability on OLEDs.

Investigation of the non-uniform distribution of current density in commercial-size proton exchange membrane fuel cells

It is necessary to obtain the current density distribution in order to understand local reactions inside the proton exchange membrane (PEM) fuel cells. Enlarging the active area to commercial-size makes it difficult to maintain uniform assembly pressure, gas supply and heat dissipation inside the fuel cell stack. Existing studies for laboratory-level fuel cells cannot effectively guide the development of large-area fuel cells. In this study, current density distribution inside a 250 cm2 and 3 kW level fuel cell stack is measured and analyzed using the printed circuit board (PCB) technology. A four-layer PCB sensor plate covering 144 current collecting segments and shunt resistors for current measurement is well-designed. The in-house developed measurement system is proved with good function, and the accuracy of current density measurement is within 2%. Results show that the flow field configuration and stack assembly condition greatly affects the uniformity of local current densities. Higher current passes through the area with larger contact pressure, and the current density is low in the poor contact area. The local current density has larger deviations for higher output current, but the normalized current is more uniform. Effects of operating conditions including stoichiometric ratio and flow direction are investigated.

Self-Assembly and Molecular Recognition in Water: Tubular Stacking and Guest-Templated Discrete Assembly of Water-Soluble, Shape-Persistent Macrocycles.

Supramolecular chemistry in aqueous media is an area with great fundamental and practical significance. To examine the role of multiple noncovalent interactions in controlled assembling and binding behavior in water, the self-association of five water-soluble hexakis(m-phenylene ethynylene) (m-PE) macrocycles, along with the molecular recognition behavior of the resultant assemblies, is investigated with UV-vis, fluorescence, CD, and NMR spectroscopy, mass spectrometry, and computational studies. In contrast to their different extents of self-aggregation in organic solvents, all five macrocycles remain aggregated in water at concentrations down to the micromolar (μM) range. CD spectroscopy reveals that 1-F6 and 1-H6, two macrocycles carrying chiral side chains and capable of H-bonded self-association, assemble into tubular stacks. The tubular stacks serve as supramolecular hosts in water, as exemplified by the interaction of macrocycles 1-H6 and 2-H6 and guests G1 through G4, each having a rod-like oligo(p-phenylene ethynylene) (p-PE) segment flanked by two hydrophilic chains. Fluorescence and 1H NMR spectroscopy revealed the formation of kinetically stable, discrete assemblies upon mixing 2-H6 and a guest. The binding stoichiometry, determined with fluorescence, 1H NMR, and ESI-MS, reveals that the discrete assemblies are novel pseudorotaxanes, each containing a pair of identical guest molecules encased by a tubular stack. The two guest molecules define the number of macrocyclic molecules that comprise the host, which curbs the "infinite" stack growth, resulting in a tubular stack with a cylindrical pore tailoring the length of the p-PE segment of the bound guests. Each complex is stabilized by the action of multiple noncovalent forces including aromatic stacking, side-chain H-bonding, and van der Waals interactions. Thus, the interplay of multiple noncovalent forces aligns the molecules of macrocycles 1 and 2 into tubular stacks with cylindrical inner pores that, upon binding rod-like guests, lead to tight, discrete, and well-ordered tubular assemblies that are unprecedented in water.

An analytical model to predict interlayer burr size following drilling of CFRP-metallic stack assemblies

A novel analytical model to describe entrance burr formation when drilling ductile metals as a function of tool geometry (point/helix angles, diameter), operating parameters (cutting speed, feed rate) and workpiece material properties, was initially formulated. The model was further developed to account for interlayer burr dimensions when drilling CFRP-metallic stack arrangements. Data from validation trials performed on Ti–6Al–4V, AA7010 and AA2024 workpieces together with their associated CFRP stack assemblies over 4 different combinations of cutting speed and feed rate showed that the predicted sizes of entrance and interlayer burrs were all accurate to within 20% of the experimentally measured results.

Effect of Side-Chain Variation on Single-Crystalline Structures for Revealing the Structure–Property Relationships of Organic Solar Cells

The molecular stacking assembly in the active layer plays a significant role in the photovoltaic performance of organic solar cells (OSCs). Here, we report two new small molecular donors with different side chains, FBT-O and FBT-H, and their corresponding fullerene-based OSCs. A slight change in the side chains led to a big difference in the power conversion efficiencies (PCEs). Although the molecular structures of the two donors are similar to each other, PCEs of the devices based on FBT-O were almost three times higher than those of the devices based on FBT-H, with manifold short-circuit current density, fill factor, as well as three orders of magnitude enhancement in the hole mobility. The difference in their single crystal structures was thoroughly investigated, whereby the FBT-O exhibited better planarity leading to appropriate phase separation and domain size. Furthermore, two-dimensional grazing-incidence wide-angle X-ray scattering results of the blend films revealed that the two donors retained a similar stacking structure as compared to the single-crystal structures, thus, establishing a clear relationship between the molecular stacking structure and the device performance.

Silver assisted stereo-directed assembly of branched peptide nucleic acids into four-point nanostars.

Branched chiral peptide nucleic acids br(4S/R)-PNA with three arms of PNA-C4 strands were constructed on a central chiral core of 4(R/S)-aminoproline as the branching center. The addition of Ag+ triggered the self-assembly of branched PNAs through the formation of C-Ag+-C metallo base pairing of the three PNA C4 arms leading to non-covalent dendrimers, whose architecture is directed by the C4(R/S)-stereocenter of core 4-aminoproline. The 4S-aminoprolyl core enabled the precise formation of four-pointed nanostars that was not realised with 4R-aminoprolyl or acyclic, achiral aminoethyl glycyl PNA cores. The dendritic assembly of 4 pointed nanostars exhibited net chirality of base stacks in CD spectra, while the base stack assembly from br(4R)-PNA 2 was overall achiral. The results demonstrate that the silver assisted, 4S-aminoproline core stereo selective chiral assembly of branched PNAs manifests into nanostar morphology. The chiral branched PNAs open new vistas in the supramolecular organization of nucleic acids.

Effect of clamping pressure on liquid-cooled PEMFC stack performance considering inhomogeneous gas diffusion layer compression

Gas diffusion layer (GDL) is a critical component of fuel cells, since it directly determines the transfer rate of mass, charge and heat that in turn affects the cell performance and stability. Within a fuel cell stack, assembled by over hundreds of single cells, GDL plays a more dominant role since the GDL deformation caused by assembly pressure would result in significant changes in its thermal, electrical and transport properties, thus impacting the uniformity, consistency and overall performance of the stack. Here, a numerical study is performed to investigate the effect of clamping pressure on the performance of liquid-cooled PEMFC stacks, which can be utilized as a guidance for stack assembly process in practical application. In particular, inhomogeneous compression as well as nonlinear strain-stress behavior of GDL are taken into consideration in the fuel-cell stack model, thus the GDL at under-rib and under-channel regions owns different physical properties, i.e., thickness, porosity, effective diffusivity, permeability, thermal conductivity, electrical conductivity, and contact resistance. It is found that the stack power output can be maximized when clamping pressure ranging from 1.5 MPa and 3.5 MPa owing to the balanced mass transport and contact resistance, meanwhile, 1.5 MPa enables the lowest voltage difference among different single cells that is beneficial in cell-to-cell consistency. Moreover, the inhomogeneous compression of GDL is found to be capable of facilitating water removal via enhanced convection at both in-plane and through-plane directions, and avoiding local hotspots with the increased in-plane thermal conductivity.

Energetically significant antiparallel π-stacking contacts in Co(II), Ni(II) and Cu(II) coordination compounds of pyridine-2,6-dicarboxylates: Antiproliferative evaluation and theoretical studies

Three new coordination compounds of types [Co(py)(2,6-PDC)(H2O)2].H2O (1), [Ni(py)(2,6-PDC)(H2O)2].H2O (2) and [Cu(py)(2,6-PDC)(H2O)].2H2O (3) (py = pyridine, 2,6-PDC = Pyridine-2,6-dicarboxylate) have been synthesized from purely aqueous media and characterized using elemental analysis, spectroscopic (IR and electronic) and single crystal X-ray diffraction techniques. Several supramolecular contacts of types OH⋯O, CH⋯O, CH⋯π, CH⋯C and π–π stacking stabilize the crystal structures. The layers in the crystal structures stack in perpendicular direction resulting in 1D channel with enclathrated water molecules. The strength of the antiparallel π-stacking interactions involving the pyridine rings in the supramolecular dimer of the compounds have been evaluated using DFT calculations and the influence of the pyridine coordination to the strength of the stacking assembly have been confirmed. The anti-proliferative potential of the compounds has been studied in Dalton’s lymphoma (DL) cell line by using MTT cell viability assay and apoptosis assay. All the three complexes exhibit short term (24 h) cytotoxicity (∼20–30%) through apoptotic cell death with negligible cytotoxicity (∼5–10%) in normal cells. In silico docking simulation has been performed with apoptosis regulator protein BCL-2 for the identification of possible molecular mode of action of the synthesized complexes. The pharmacophore features based on structure activity relationship (SAR) of the complexes have been identified. The SAR results reveal that the molecular features such as, hydrophobic, aromatic, positive ionizable, negative ionizable, H-bond donor and acceptor and halogen bond donor associated with the structures of the compounds play important role in the biological activities. The ADMET features, viz., physicochemical properties, pharmacokinetics, drug-likeness and medicinal chemistry friendliness of the synthesized complexes have been investigated.

Boss Tower and Baseplate Flange Optimization

The swaging process is commonly used in head stack assembly, which is a process that is undertaken to either reduce or increase the diameter of tubes or rods. In the head stack assembly process, a swage boss engages with an arm hole of E-Block. Current hard disk drives have very small tolerances, requiring the swage effect of new hard disk drives to be reduced to as little as possible. In this study, finite element analysis (FEA) was used to reduce the swage effect. At present, FEA and investigative tests indicate that a significant portion of gram changing through swaging is caused by E-block tip deformation due to swaging. Deformation of the E-block tip is the result of unbalanced swage forces between the tension and compression suspension. Achieving optimization concepts leads to a new swage plate design to reduce the gramload change of the target and to equalize the swage forces. The FEA results found that gramload change of the target decreased by 35% on the tension side, and by 65% on the compression side through swage by boss tower. The HGA torque out was not different. The suspension flange was also studied. Subsequently, suspension manufacturing is able to bias the suspension as incoming HGA. The flange deformation might be closer to zero after the swaging process.

Overcoming the Challenges for a Mass Manufacturing Machine for the Assembly of PEMFC Stacks

One of the major obstacles standing in the way of a break-through in fuel cell technology is its relatively high costs compared to well established fossil-based technologies. The reasons for these high costs predominantly lie in the use of non-standardized components, complex system components, and non-automated production of fuel cells. This problem can be identified at multiple levels, for example, the electrochemically active components of the fuel cell stack, peripheral components of the fuel cell system, and eventually on the level of stack and system assembly. This article focused on the industrialization of polymer electrolyte membrane fuel cell (PEMFC) stack components and assembly. To achieve this, the first step is the formulation of the requirement specifications for the automated PEMFC stack production. The developed mass manufacturing machine (MMM) enables a reduction of the assembly time of a cell fuel cell stack to 15 minutes. Furthermore the targeted automation level is theoretically capable of producing up to 10,000 fuel cell stacks per year. This will result in a ~50% stack cost reduction through economies of scale and increased automation. The modular concept is scalable to meet increasing future demand which is essential for the market ramp-up and success of this technology.

Measurements of Blowdown Thrust Loading in Natural Gas Facilities

Abstract The thrust loading on a vertical blowdown stack during a natural gas blowdown was investigated using a combined experimental and modeling approach. A gravimetric vessel initially at 4000 kPa-g was blown down through two geometrically different stack assemblies. Thrust loads were measured using a dynamic weigh scale typically used for gravimetric calibration of gas flow meters. A one-dimensional (1D) compressible flow model, calibrated using the experimental data, revealed stagnation pressure losses at the entrance to the riser, resulting in lower thrust loads. A comparison between thrust loading obtained from the measurements and the 1D compressible flow model is presented. This work shows that the analytical flow model predicts the blowdown thrust loads within ±30%.

A systematic error modeling and separation method for the special cylindrical profile measurement based on 2-dimension laser displacement sensor

The aim of this study is to improve aeroengine multistage rotor performance and to reduce the incidence of failures. Measuring the form error of seam allowance connecting cylindrical surface accurately is critical to achieve rotor optimal stack assembly stage-by-stage. In this paper, compared to the traditional cylindrical profile measurement model, a more comprehensive measurement model was built based on a 2-dimensional line laser sensor measurement technique, a model in which the component eccentricity error e, the sensor lateral offset error d, the sensor forward tilt error θx, the sensor lateral tilt error θy, and the rotary table tilt error γ were gradually modeled and separated. The particle swarm optimization algorithm was adopted to solve the model error parameters. The residual error simulation was performed to observe the effect of different levels of offset errors on the measurement results. The aeroengine rotor seam allowance surface measurement experiment was performed to verify the validity of the method model. We can conclude that the sensor lateral offset error d was 3.214 µm, sensor forward tilt error θx was 12.754″, sensor lateral tilt error θy was 10.365″, and rotary table tilt error γ was 2.146″. The cylindricity error value was 3.701 µm. Compared with the traditional cylindrical profile measurement method, the measurement accuracy of cylindricity error was improved by 1.768 µm. The proposed method can improve the measurement accuracy significantly of multistage rotors in the aeroengine measurement process; besides, it can also be extended to the measurement of other geometric form errors.

Acoustic analysis of ultrasonic assisted soldering for enhanced adhesion

Ultrasonic soldering utilizes high-intensity acoustic fields to induce cavitation in the solder melt in order to (i) bond dissimilar materials and (ii) improve solder joint strength. The acoustic energy transfer from the piezoelectric transducer (PZT) into the liquid solder pool is critical in both understanding and optimizing this process. We use finite element analysis of the acoustics and compare with experiment. Our finite-element modeling approach is two-pronged; (i) we develop a one-dimensional model that is used as a design tool to optimize the solder stack geometry to match the transducer frequency for maximal acoustic energy transfer and (ii) we use a three-dimensional model to compute the frequency response in the solder stack assembly (solid acoustics) and the acoustic pressure in the liquid solder pool (solid-fluid interaction). The acoustic pressure is a proxy for cavitation and therefore bond strength. Our simulations show the acoustic pressure rapidly decreases as the height of the solder tip above the substrate surface increases, which correlates with controlled experiments that show the solder bond quality also decreases with increasing tip height.

Delamination of Reinforced Membranes in Vanadium Redox Battery Cells

Redox-flow batteries (RFBs) possess compelling attributes for stationary energy storage. In particular, the decoupling of energy and power in RFBs enables the cost effective use of redox active materials with inherently low energy density1. This decoupling is especially advantageous for long-duration applications, i.e., discharge times ≥ 5hours at rated power2. All-vanadium redox battery (VRB) cells typically consist of a proton-conducting membrane between two porous carbon-fiber-based electrodes. The VII/VIII and VIV/VV redox couples, dissolved in strong sulfuric acid, flow through the negative and positive electrodes, respectively. The primary reactions during charge and discharge are the redox reactions of these two vanadium couples. Protons in an ion-exchange membrane (IEM) typically carry charge between the two electrodes. Movement of reactant species from one electrode to the other through the membrane results in inefficiency and reversible capacity losses and is undesirable3. Therefore, for a given membrane material, there is an optimal membrane thickness to achieve the desired performance with respect to both voltage efficiency (i.e., membrane ohmic loss) and coulombic efficiency (i.e., membrane crossover)4. State-of-the-art VRB cells employ relatively thin IEMs to enable higher ionic conductance (e.g., ≤ 25 μm thick3), which enables high power densities and reduces VRB stack costs2, 5. Perfluorinated sulfonic-acid (PFSA) membranes are typically used in VRBs due to their outstanding stability in the highly oxidizing positive electrolyte (containing VV)6. Thin PFSA membranes are often reinforced with a porous matrix to improve the mechanical robustness and reduce dimensional changes when exposed to varying levels of humidity. In addition to reducing swelling, some reinforced membranes (e.g., GORE-SELECT) can significantly decrease V permeability (33% to 47%) compared to non-reinforced membranes with the same equivalent weight (EW), while the corresponding increase in area-specific resistance (ASR) with reinforcement is smaller (13% to 18%)7. Therefore, reinforced membranes can potentially enable improved VRB cell performance, as well as being easier to handle during stack assembly. Unfortunately, reinforced membranes can undergo delamination in VRB cells. An example of this is shown in Fig. 1, which is a picture of a reinforced membrane from 3M Corporation that was removed from a stack after cycling for more than 2,000 hours. Vionx Energy’s cells use interdigitated flow fields (IDFF), which enable high performance with low pressure drops8, 9. The clear (darker) areas in Fig. 1, where the membrane remains relatively intact, coincide with the ribs of the IDFF. The white areas are where the membrane has separated into layers. These areas coincide with the IDFF channels. There is residual electrolyte between the delaminated layers. This talk shall describe the mechanisms that we hypothesize to be the causes of this delamination of reinforced membranes in VRB cells. We shall also describe accelerated test protocols that can be used to assess the likelihood for a membrane to delaminate in a RFB cell and discuss various mitigation strategies, including reinforced-membrane architectures that are less prone to delamination in RFB cells.

Circular economy in membrane technology: Using end-of-life reverse osmosis modules for preparation of recycled anion exchange membranes and validation in electrodialysis

In the current study, the indirect recycling of discarded reverse osmosis modules as i) membrane support for the preparation of anion exchange membranes and ii) polypropylene components for the assembly of an electrodialysis stack have been investigated for the first time. In such a way, 51% of the discarded module could be recycled into an electrodialysis system composed of 54% of recycled components. Recycled anion exchange membranes have been prepared by casting and phase inversion methods using discarded reverse osmosis membranes, preconditioned as support. The influence of casting thickness and solvent evaporation time in membrane properties has been studied. Besides, a complete membrane characterization has been carried out (SEM, water content, ion exchange capacity, permselectivity, electrical resistance, diffusion coefficients and mechanical properties) to select the optimal membrane preparation conditions, obtaining recycled anion exchange membranes with a high permselectivity (87%, similar to the commercial membranes). Finally, the technical viability of the recycled membranes has been tested by brackish water desalination experiments in the assembled electrodialysis stack, achieving 84.5% of salt removal. This study could open an alternative within the recycling of discarded reverse osmosis membranes, avoiding their disposal in landfills and moving membrane technology into a circular economy.