Abstract
| Original language | English |
|---|---|
| Pages (from-to) | 352-363 |
| Number of pages | 12 |
| Journal | International Journal of Hydrogen Energy |
| Volume | 99 |
| Early online date | 18 Dec 2024 |
| DOIs | |
| Publication status | Published - 20 Jan 2025 |
Funding
The pore size distribution of the prepared supports was measured using Capillary Flow Porometry (CFP) in a tailor-made setup [31]. The technique relied on the correlation between the pressure needed to displace a wetting liquid from the sample's pores, and the permeating flow of N2 gas through the sample.Due to the simultaneous coexistence of multiple transport mechanisms, the dusty-gas model (DGM) is considered as the most suitable approach to describe hydrogen permeation through the membrane's support [35]. Specifically, the single component dusty-gas model used in this work is shown in Equation (1) [35]. This model requires only the determination of characteristic parameters K0 and B0 which are to be fitted based on flux data obtained through permeation tests. In Figure S.3 (supplementary information), the parity plot of the fitted/modelled flux and of the experimental flux is shown. For the fitted values of K0 and B0 are 3.850\u00D710\u22128m and 1.693\u00D710\u221214m2 respectively with an R2 of 86.81%. Through these parameters, the contribution of Knudsen diffusion and viscous flow to the support's permeation mechanism can be evaluated comparing the terms of Equation (1) that are related to the two different transport mechanisms. Thus, in Equation (5) and Equation (6) the contribution given by Knudsen diffusion (i.e., Knudsen diffusivity, Dk,i) and the contribution of viscous flow are calculated for the case in which the support is exposed to hydrogen permeation at 400 \u00B0C and 2 bar. A comparison between these terms leads to the observation that both Knudsen diffusion and viscous flow contribute to the support's overall permeation behavior, thus, from a modelling perspective, neither of the two terms can be neglected. This is indeed in line with the outcomes of the analysis of the pore size distribution and of the calculated Knudsen numbers depicted in Fig. 5.The pore morphology visualized by laser-optical microscopy of the support equipped with interdiffusion barrier (configuration S-IDB) is shown in Figure S.2 (supplementary information). The asymmetric \u03B1-Al2O3 filler particles are visible underneath a superficial transparent layer, which is clearly distinguishable above the pore mouth in the height mapping view. The asymmetric filler and the interdiffusion barrier were proven to increase suitability for PdAg deposition and prevent intermetallic diffusion between Pd and support metal [27].To identify the changes in terms of governing gas transport mechanisms arising from the asymmetric filling of the support and from the deposition of the interdiffusion barrier layer, pure hydrogen and nitrogen permeation tests were performed for the support equipped with interdiffusion barrier (configuration S-IDB) at similar conditions to those investigated for the support (configuration S). In agreement with the results obtained for configuration S, Fig. 7, in which the H2 and N2 fluxes through the membrane are represented as a function of the trans-membrane pressure difference at different temperatures, shows that the flux increases linearly with an increasing trans-membrane pressure difference. However, from a comparison with Figs. 4 and 7 indicates that the asymmetric filling and deposition of interdiffusion barrier result in a significant decrease of membrane's permeance as well as in a less remarkable dependence of the H2 and N2 fluxes through the membrane on temperature. Fig. 8, in which the pore size distribution of the support equipped with interdiffusion barrier is depicted alongside the calculated Knudsen number for hydrogen at 400 \u00B0C and 3 bar(a) as a function of pore size, shows that the pore size distribution, with the majority of pores in the range of 20\u2013100 nm, is shifted significantly towards smaller pores compared to what is observed for the support (S) in Fig. 5. Moreover, it illustrates that the calculated Knudsen numbers of the most abundant pores are higher than unity, which makes it possible to infer that Knudsen transport mechanism plays a significant role in these pores. Consequently, a possible reason for the decrease in permeation flux after asymmetric filling and deposition of the interdiffusion barrier may be attributed to the increased influence of Knudsen diffusion on the overall transport behavior (see Fig. 9).In Figure S.4 (supplementary information) the parity plot of the experimental flux data obtained from the permeation tests with S-IDB and of the flux through the support equipped with interdiffusion barrier obtained from the dual-layer mass transport model is represented. The fitted values of K0 and B0 for the IDB layer are 1.436\u00D710\u221211m and 0m2 respectively with an R2 of 92.64%. The data fitting resulted into a value of B0 approaching zero, indicating that viscous transport has no significant contribution to the overall mass transport behavior. This is in line with the outcomes of the analysis of the pore size distribution and of the calculated Knudsen numbers depicted in Fig. 10, thus confirms that Knudsen diffusion can be considered the dominant transport mechanism governing permeation through the membrane when the support is equipped with interdiffusion barrier.[Figure presented]This work has received funding from the European Union's Horizon 2020 Research and Innovation Program under grant agreement No 869896 (MACBETH). Image 1 This work has received funding from the European Union\u2019s Horizon 2020 Research and Innovation Program under grant agreement No 869896 (MACBETH)
Keywords
- Hydrogen
- Mass transfer
- Membrane
- Metallic support
- Modelling
- Palladium
- Separation
