Pore multi-layered

Fabrication processing of these materials is highly complex, particularly for materials created to have interfaces in morphology or a microstructure [4—5], for example in co-fired multi-layer ceramics. In addition, there is both a scientific and a practical interest in studying the influence of a particular pore microstructure on the motional behavior of fluids imbibed into these materials [6-9]. This is due to the fact that the actual use of functionalized ceramics in industrial and biomedical applications often involves the movement of one or more fluids through the material. Research in this area is therefore bi-directional one must characterize both how the spatial microstructure (e.g., pore size, surface chemistry, surface area, connectivity) of the material evolves during processing, and how this microstructure affects the motional properties (e.g., molecular diffusion, adsorption coefficients, thermodynamic constants) of fluids contained within it. 

Homogenous (multi) layers in the pores Plugs in the pores (constrictions) Plugs/layers on top of the pores... 

Multi-layered high-quality ceramic membranes do not have any appreciable pore blockage by the precursor particles from the adjacent layers. This is evidenced by the sharp and clean interface between layers shown in Figure 4.2. It should also be noted that the smoothness of the underlying layer is critical to forming a smooth layer on top of iL... 

Multi-layered pore size distributions. The multi-layered structure of the membrane/support composites seen by the scanning electron microscopes as shown earlier is reflected in multiple "plateaus of the cumulative pore size distributions (Figure 4.12). The sharp drops in the distribution indicate that the pore size in each layer is quite uniform. The drop near 10 pm indicates the bulk support while the drop near 4 nm represents the selective membrane layer. The two drops in between are related to the two intermediate support layers. 

Many commercial ceramic membranes have two three or even four layers in structure and their pore size distributions are similar to that shown in Figure 4.12. It should be noted, however, that determination of those multi-layered, broad pore size distributions is not suaightforward. The major reason for this is the overwhelmingly small pore volume of the thin, fine pore membrane layer compared to those of the support layer(s) of the structure. It is possible, although very tedious, to remove most of the bulk support layer to increase the relative percentage of the pore volumes of the membrane and other thin support layers. Provided the amount of bulk support layer removed is known and the mercury porosimeu data of the "shaved" membrane/support sample is determined, it is feasible to construct a composite pore size distribution such as the one shown in Figure 4.12. 

For multi-layered asymmetric or composite membranes where the pore sizes between layers are widely different, the analysis gives the pore size distribution of the densest layer (membrane) even if they may be only a small volume or weight fraction of the total membrane/support structure. Shown in Figures 4.14 (a) and (b) are the pore size distributions of two very similar multi-layered alumina membranes prepared by the sol-gel process and two distinctively different alumina membranes made by the anodic oxidation process. The comparison of pore size distributions in each case reflects the similarities and differences of the membrane samples. 

Thus Eq. (4-13) implies that Knudsen diffusion is practical only when those gases with large differences in their molecular weights are to be separated. For applications where this mechanism poses as a severe limitation, other more effective separation mechanisms would be necessary. Two such possibilities are surface diffusion (and multi-layer diffusion) and capillary condensation, both of which are dependent on the chemical nature of both the membrane material and pore size and the species to be separated. 

The type II isotherm is associated with solids with no apparent porosity or macropores (pore size > 50 nm). The adsorption phenomenon involved is interpreted in terms of single-layer adsorption up to an inversion point B, followed by a multi-layer type adsorption. The type IV isotherm is characteristic of solids with mesopores (2 nm < pore size < 50 nm). It has a hysteresis loop reflecting a capillary condensation type phenomenon. A phase transition occurs during which, under the eflcct of interactions with the surface of the solid, the gas phase abruptly condenses in the pore, accompanied by the formation of a meniscus at the liquid-gas interface. Modelling of this phenomenon, in the form of semi-empirical equations (BJH, Kelvin), can be used to ascertain the pore size distribution (cf. Paragr. 1.1.3.2). 

The nitrogen adsorption isotherms are characteristic of microporous solids. The aging treatments cause a clear decrease in the micropore volume and in the microporous surface. The decreases are nevertheless smaller for the Cu-ZSM-5 solid (variation A = 0.04 after aging at 1173 K) than for the parent zeolite (A= 0.07) (Table 2). An apparent BET surface area has been reported though the BET theory is not applicable to microporous materials since the pore condensation isotherm is interfering with the multi-layer adsorption isotherm. 

The PS layer formed on a Si substrate is the most frequently used as a PS host. Sometimes, partial or complete modification of the PS layer (by annealing, oxidation, carbonization, etc.) is performed prior to pore filling or coating process [2]. Multi-layered PS structures (with different porosity) [8], self-supported PS films [9], and PS grains (PS layer separated from Si substrate and divided onto grains) [10] are used as well. 

Different multi-layered structures are formed using the PS host. These stmctures consist of upper (sometimes, epitaxial) layer, PS layer with filled or not-filled pores, and single crystal Si substrate. 

The Crystal FT filters are made of 100% self-bonded, re crystallised silicon carbide wuth multi-layer SiC membranes. The filter carrier is characterised by high porosity and an open three-dimensional pore structure, Crystar says modules with 4CFT and 5CFT filters showed time and energy savings of 36%. 

We are aware that these assumptions are simplifications of the high complexity of real GDL structures. However, as will be shown in Section 24.5, the fitted model coincides fairly well with real data. Note that the characteristics for the model validation are mainly focused on properties of the pore phase, that is, the multi-layer model seems adequate to investigate processes in the pore phase such as gas and water transport. 

In this section, the structural model for paper-type GDLs is compared with real 3D data gained by means of synchrotron tomography [23, 24]. This comparison is based on the above-described structural characteristics for porous media and can be seen as a model validation. In particular, characteristics are applied which are related to transport properties of the pore phase, since the investigation of transport processes is the main area of application of the multi-layer model see also the results and discussions in Thiedmann et al. [5]. 

---