Page 270 - Adsorption by Powders and Porous Solids

2.58 ADSORPTION BY POWDERS AND POROUS SOLIDS

micropores (i.e. the ultramicropores) occurs at very low p/pO, the change in amount
adsorbed at P/pO > 0.01 is quite small (Atkinson et al., 1987; Kenny et al., 1993) and
therefore it is not surprising to find that the DR plot is virtually linear over a fairly
wide range of p/pO.
In some investigations (Femandez-Colinas et al., 1989a; Kakei et al., 1991;
Kenny et al., 1993; Kaneko, 1996) it has been found that the shapes of high-resol"-
tion as-plots and DR plots provide strong evidence for a sequential filling of several
groups of micropores. For example, the nitrogen isotherms in Figure 9.12a were
determined on a series of activated pine wood charcoals (Femandez-Colinas et a/.,
1989a). The change in isotherm shape is the first indication that pore widening has
occurred as a result of progressive activation in steam. The as-plots in Figure 9.12b
confirm this interpretation and indicate that this was mainly due to the development
of a supermicropore structure. It appears that the initial stage of primary micropore
filling at very low p/pO (i.e. p/pO < 0.01) was followed by the more gradual filling of
supermicropores.
Surface coverage of the supermicropore walls is indicated by the appearance of a
short linear section at as > 0.5. Extrapolation of this section to a, = 0 provides an
approximate evaluation of the effective ultramicropore volume, vp(u, rnic).
The second linear section extends over the multilayer range of each as-plot in
Figure 9.12b. Backward extrapolation of this branch gives the total effective micro-
pore volume, v,(mic), from the intercept on the vs axis. It follows that the effective
supermicropore volume, v,(sup, mic), can be regarded as the difference vp(mic) -
v,(u, mic). It is of interest that after an initial small change in v,(u, rnic), this has
remained constant during further activation while the magnitude of vp(sup, mic) has
increased steadily.
As discussed in Chapter 8, the pre-adsorption of n-nonane can be used as a means
of blocking narrow micropore entrances (see Section 8.2.3). Thus, in the case of an
ultramicroporous adsorbent such as Carbosieve, the pre-adsorption of nonane leads
to complete blockage of the pore structure. The effect of progressively removing the
pre-adsorbed nonane from a supermicroporous carbon is shown in Figure 9.13. The
adsorbent used in this work was a well-characterized carbon cloth with the following
properties: a(BET), 1330 m2 g-l; a(ext), 25 m2 g-l; vp(mic), 0.44 cm3 g-l; w,,
0.6-2.0 nm (Carrott et ad., 1989).
Inspection of the as-plots in Figure 9.13 is instructive. It can be seen that there are
two linear sections, back-extrapolation of the first giving a zero intercept when part
of the pre-adsorbed nonane was removed by prolonged outgassing at 50°C. We may
conclude that the initial stage of nitrogen adsorption is by monolayer adsorption on
the walls of the supermicropores. This is followed by the formation of a quasi-multi-
layer until pore filling is complete at = 0.4. Increase of outgassing temperature
leads to the progressive removal of nonane from the ultramicropores and narrow
entrances and the original nitrogen isotherm is restored after prolonged outgassing at
200°C. Thus, we may compare the change in the extent of total micropore capacity,
v,(mic) (obtained as described earlier by back-extrapolation of the linear multilayer
branch), with the magnitude of the restoration of the ultramicropore capacity,
v,(u, mic) (as defined above). The fact that at each stage vp(mic) > vp(u, mic) is a