Mass Transport of Multicomponents Solute in Bentonite Clay

  • PDF / 907,491 Bytes
  • 14 Pages / 612 x 792 pts (letter) Page_size
  • 3 Downloads / 197 Views

DOWNLOAD

REPORT


(a) Original image

(b) Enlarged height image

Figure 1. Confocal scanning laser photomicrographs of compacted bentonite (ρdry=1.6).

Since the crystalline structure of montmorillonite determines the fundamental properties of the montmorillonite hydrate, we need to analyze its molecular behavior. Thus by applying a molecular dynamics (MD) simulation we inquire into the physicochemical properties of the montmorillonite hydrate such as diffusivity of chemical species. For extending the microscopic characteristics of constituent materials to a macroscopic diffusion behavior of the micro-inhomogeneous material we apply a homogenization analysis (HA; Sanchez-Palencia 1980). Note that before applying HA we summarize the existing diffusion models for soil. Then we present our model, which can treat the micro-inhomogeneous properties with adsorption behavior on the surface of clay minerals. A CLASSICAL DIFFUSION THEORY OF MULTICOMPONENT SOLUTION IN POROUS MEDIA First we treat a classical diffusion theory of chemical species in a dilute solution and the diffusion in a porous media saturated with solution, because it gives the fundamental idea why we need to introduce the micro/macro analysis based on the molecular simulation and the homogenization method. Diffusion of multicomponent chemical species in solution Let us think a dilute solution with N-components, in which (1,2,…,N-1)-th components are solutes (i.e., diffusing chemical species), and the N-th component is solvent (i.e., water in our case). We introduce the mass-percent concentration of the α-th species, cα , by cα = mα/m where mα is the mass of the α-th species, and m the total mass. Let ρα = mα/V (V the volume) be the mass

density of the α-th species, and ρ = m/V = ΣN ρα the average mass density. Due to the mass conservation law of the α-th species we have ∂ ( ρ cα ) ∂ d α ρα & + fα = + ρ cα viα + f&α = 0 dt ∂t ∂xi

(

)

(1)

where vαi is the velocity of the α-th component particles, d α φ / dt = ∂φ / ∂t + ∂ (φ viα ) / ∂xi implies the material time derivative of a function φ with respect to the α-th component, and f&α gives the source term per volume per unit time because of chemical reactions, etc. The average velocity v is defined by v=

1

N

N

∑ ρ vα = ∑ cα vα

ρ α =1

(2)

α =1

A diffusing mass flux of the α-th species, jα, is introduced as jα = ρα ( vα − v) = ρ cα ( vα − v) , and by using the Fick’s first law jα = − ρ ∑ β Dαβ ∇cβ , Eqn(1) can be written as ∂cβ  & * ∂cα ∂c ∂  N + vi α −  ∑ Dαβ  + fα = 0 ∂t ∂xi ∂xi  β =1 ∂xi 

(3)

where we set as f&α* = f&α / ρ . The source term f&α is determined by an inter-species relation either at equilibrium or on reaction process. Here we treat only on the equilibrium state. Let us consider a stoichiometric (reversible) process given by

ν

f f A1 + ν A2 + ... 1 2

ν

b b B1 + ν B2 + ... 1 2

or

να Mα ∑ α

=0

(4)

where ν αf and ν αb are the forward and reverse stoichiometric coefficients of α-th species, respectively. Note that the stoichiometric coefficients, ν α , of Eqn(4)2 corresponding to the le