Hg intrusion porosimetry

The microstructure of the green body as observed by SEM is shown in Fig. 4. After consolidation, powders in green body compact closely and homogeneously. However, the polymer network was not observed via SEM. The pore diameter distribution, obtained by Hg intrusion porosimetry, showed a monomodal distribution type. The relative density, porosity, and the median pore diameter of green samples were 57.67%, 35.02%, and 10.6nm, respectively. 

A very detailed fuel cell-related TEM study of CB and CB corrosion can be found in Eiu et al. [15]. Gas adsorption, Hg intrusion porosimetry, and small-angle X-ray scattering (SAXS) provide information on porosity and pore size distribution, each of them having its own advantages and limitations. Methods that can be used to probe the functional groups at the carbon surface are X-ray photoelectron spectroscopy (XPS) and near-edge X-ray absorption spectroscopy (NEXAES). Table 7.1 gives an overview of characterization techniques applied... 

Mercury intrusion porosimetry is suited for pores as low as 2 nm and greater than 350 tun. N2 intrusion porosimetry measures pores smaller than 2 nm but this technique is unsuited for determining the inter-particle void fraction its upper range is smaller than for Hg intrusion. Pore sizes are classified into three ranges ... 

Fig. 5 exhibits the Hg porosimetry curves simulated for the diverse types of porous networks. Crudely, Hg intrusion is the simile of N2 desorption both are capillary processes in which percolating, pore blocking or entr ping effects can arise. These effects are highly influenced by the conn tivity of the porous network. The shapes of the Hg intrusion curves are a function of the bond size distribution. Comparatively, the Hg intrusion curve of N-I has an intrusion zone that is located at relatively high pressures, whilst N-III depicts an intrusion process at feirly low pressures. Finally, the Hg intrusion course corresponding to the N-II substrate occurs at intermediate pressures. 

Another characterization method that can be included with those based on the Laplace equation is mereury intrusion porosimetry. The method (also proposed by Washburn [101]) was developed by Ritter and Drake [52] and was applied for the first time to the characterization of membrane filters by Honold and Skau [102]. It has been shown to be a reliable method for the characterization of pore size distributions, pore strueture, and speeific surface areas. Here, a Hg-air interface appears inside each pore. Thus, Eq. (4) is also followed. However, in this case, Hg does not wet practically any kind of sample (the corresponding contact angles ranging from 112° to 150°) [52, 101]. 

Froment discusses pore network influences in deactivation. He concludes that, "Evidently, the parameters associated with the pore and network structure should be determined from independent physical measurements adsorption, mercury porosimetry, electron microscopy..." Unfortunately, no experimentally based studies have been published that have employed one or a combination of techniques to determine the pore network structure and its changes during deactivation. The few experimental determinations of pore structure are limited to determination of pore dimensions usually from the intrusion data in Hg porosimetry or the desorption data from nitrogen physical sorption (often incorrectly referred to as BET analyses). 

In Figure 8.8, the distribution section below 2 nm has been calculated from Nj adsorption and Brunauer s model, the section between 2 nm and 7.5 nm has been calculated from Nj adsorption and Broekhoff and de Boer s model, the section between 7.5 nm and 94 nm has been calculated from Hg porosimetry and Washburn s model and the section beyond 94 nm has been calculated from Hg porosimetry and Pirard s model. The size of 94 nm corresponds to the transition from the collapse mechanism in pores larger than 94 nm to the intrusion mechanism in pores smaller than 94 nm (P = 16 MPa). The total cumulative porous volume is obtained by summing up the cumulative volumes corresponding those successive ranges. 

The surface areas calculated from the porosimetry intrusion curves (Hg Area), presented in column 4 were calculated using a cylindrical nonintersecting pore model, and only represent the surface area of pores down to 3.5 nm pore radius [10]. Discrepancies between the surface area results obtained from this method and that of gas adsorption, presented in column 5, were due to the presence of pores of less than 3.5 nm which remained undetected. These differencies were greatest with the lower heat treatments but as the thermal treatment was increased the shifts in the pore size distributions to wider mesopores and eventually to macropores, brought the two measurements into closer agreement. 

Figure 6.8 Incremental intrusion spectra of Hg porosimetry of (open squares) nonporous LaFe03 prepared without PMMA template and (open circles) 3DOM LaFe03 prepared using PMMA (diameter 268 nm).

Liquid intrusion/extrusion method should apply ideally to membranes containing cylindrical pores. When this is not the case, the pore size distributions obtained refer actually to equivalent or effective pore sizes. In Fig. 12 the air-liquid displacement distribution is presented along with the Hg-porosimetry one for the AOl membrane. A Ti02 /a-Al2 O3 layer on a stainless steel woven wire cloth made by Synthesechemie, whose pores are far from being cylindrical, as shown in Fig. 13, gives a different pore size distribution by air-liquid displacement and Hg-porosimetry distributions, as shown in Fig. 14. 

---