BASIC CONCEPTS AND AN OVERVIEW - Foundations of Nanotechnology: Pore Size in Carbon-Based Nano-Adsorbents - page 155

can be derived by considering a micropore of width w exposed to adsorptive at a
fixed low pressure. We imagine that the pore walls resist expansion with a separa-
tion dependent potential w. We can write as a condition for dilation where n is the
mols of adsorptive transferred to the pore during dilation and m is the chemical
potential of the bulk adsorptive. The current widely used model for analyzing
micropore distribution in activated carbon assumes an array of semiinfinite, rigid
slits of distributed width whose walls are modeled as energetically uniform graph-
ite. Adsorption isotherms can be simulated for this system using GCMC or DFT.
Inversion of the integral equation of adsorption to determine micro pore size dis-
tribution from experimental isotherms using such models usually produces results
showing minima near 6 and 10 A˚ effective pore width, regard less of the simula-
tion method used. This is assumed to be a model induced artifact. The inclusion
of surface heterogeneity in the model, while more realistic, does not change this
observation significantly. The strong packing effects exhibited by a rigid parallel
wall model seems likely to be the dominant feature causing the double minima in
the derived pore size distributions [142].
Pore characteristics of the prepared porous samples in terms of PSD were
determined using some well-known models of DS, Stoeckli, HK, IHK, DA, DFT,
BET, etc. [121]. The effect of different parameters on PSD of porous carbon nano
structures samples was also investigated in recent works. For example, in recent
work increasing impregnation ratio increases the pore volume of activated sam-
ples, and ZnCl 2 resulted in ACs with more adsorption capacities than those of
KOH. This increase for KOH chemical agent will create more micropores with
insignificant variations of average pore size, while in the case of ZnCl 2 it creates
wider pores, and after a specific impregnation ratio (100%) it begins to create
mesopores in the carbonaceous structure. In this review three different meth-
ods: CPMD, MD and DFT is compared [129], which were adopted in nano-scale
researches. CPMD, which is a typical ab initio MD, still has the difficulties in
studying significantly larger systems. A novel ab initio MD method, suitable for
simulating more atoms, is desirable. Classical MD allows calculations on systems
containing significant numbers of atoms in a relatively long duration. However,
current empirical potential functions are not accurate enough to reproduce the
dynamics of molecular systems. DFT is expected to be applied in a larger system
in the further. Thus, a better separation can be obtained with bigger pore sizes and
relatively small distances. Therefore, an excellent separation effect can be prob-
ably obtained when varying the radius of the tube. As a result, choosing a proper
temperature can greatly improve the separation. It is obvious through this review
that pressures, temperatures and sorbent structures are all important factors for the
separation of gas mixtures. At last, we have to stress that our simulation results
depend on our choice of intermolecular potentials, but such potentials seems to
be a reasonable estimate of the interactions. REBO-based MD simulations consti-
tute a dead-end for the simulation methods. Both chief tasks of the metal catalyst