Page 102 - Partition & Adsorption of Organic Contaminants in Environmental Systems

IMPROPER SURFACE-AREA MEASUREMENT 93

Most adsorbents are highly porous bodies with tremendously large internal
surfaces. The external surface, even that visible under the best microscope, only
constitutes a small fraction of the total surface.
However, as long as the gas does not penetrate into the field of force that exists
between the atoms, ions, or molecules inside the solid, it is considered to be on the
outside, even it is adsorbed on the internal surface of the adsorbent.
If the gas enters inside of a solid, two things may happen: either the gas merely dis-
solves in it, forming a solid solution, or it reacts with the solid and forms a new
compound.

The comments above clearly convey the point that the large internal sur-
faces of the porous solids (e.g., activated carbon) are internal only when
viewed with respect to the outer granule boundary, but are nevertheless exter-
nal to the bulk solid and accessible to an inert gas, as later emphasized by
Chiou et al. (1992). Therefore, if the vapor penetrates significantly into a solid
and if such vapor data are used for surface-area calculations,a mistakenly large
total surface and hence an erroneous internal surface will result. As a reduc-
tio ad absurdem, if one were to measure the uptake of, say, butane vapor by
bulk hexane liquid, the resulting data could easily be misinterpreted as result-
ing from a presumed very large “internal surface” of the hexane.
A comparison of the surface areas obtained by the BET method by use of
an inert vapor (N 2) and a polar vapor (e.g., EGME) for solids of different
makeups helps to pinpoint the effect of solvent penetration on the determined
surface areas. To meet this objective, let us consider the vapor uptakes of N 2 at
liquid N 2 temperature (77K) and EGME at room temperature on a series
of natural and synthetic solids (Chiou et al., 1993), as shown in Table 6.2.
The isotherm data on kaolinite (KGa-2), alumina (Quanta Chrome Co.), Ca-
montmorillonite (Ca-SAz-1, unpurified), illite (Fifthian, Illinois), a mineral
(Woodburn) soil, and a peat soil (Everglades, Florida) are shown, respectively,
in Figures 6.4 to 6.6. The N 2 isotherms on all the solids are notably nonlinear
with a type II shape.At liquid N 2 temperature,the N 2 vapor uptake by any solid
takes place virtually all by surface adsorption, as there is little possibility for
N 2 penetration into the solid. The respective EGME isotherms on all solids at
room temperature are similarly nonlinear, with a generally sharper curvature
except for the peat, the latter exhibiting instead a practically linear isotherm.
As noted in Figure 6.4, the EGME uptake capacities on alumina and kaoli-
nite are only moderately higher than the corresponding N 2 capacities. The
same applies to the N 2 and EGME data with sand, hematite, and synthetic
hydrous iron oxide (Chiou et al., 1993). This result is indicative of the surface
coverage, since EGME has a greater molecular weight (MW = 88) and a some-
what higher density (d l = 0.93g/mL) than N 2 (d l = 0.808g/mL) and since,
according to Eq. (6.1), the adsorbate mass per unit area is proportional to
1/3
2/3
d l M . On the other hand, the EGME uptake capacities on highly expand-
ing Ca-SAz-1, nonexpanding illite (0.4% organic matter),Woodburn soil (21%