(695e) Density Functional Theory Calculations of the Interaction of Water with Forsterite (100) Surface

Density functional theory
calculations of the interaction of water with forsterite (100) surface

Valentina Prigiobbe¹, Dong-Hee Lim², Ana
Suarez-Negreira², Jennifer Wilcox²

¹Department of Petroleum and Geosystems
Engineering, University of Texas at Austin, USA

²Department of Energy Resources
Engineering, Stanford University, USA

The interaction of water (H₂O) with forsterite (Mg₂SiO₄),
the magnesium-end member of the silicate mineral olivine ((Mg,Fe)₂SiO₄),
is a fundamental mechanism in the weathering process investigated in the
geosciences and engineering such as hydration of the upper Earth's mantle,
transport of contaminants in the subsurface, and mineral carbonation for carbon
dioxide (CO₂) storage.

Weathering of Mg₂SiO₄ in aqueous systems consists
of a solid-liquid interfacial reaction where H₂O and CO₂
react with the mineral. In particular, Mg₂SiO₄ undergoes
hydration (adsorption of H₂O molecules) and dissolution (release of
Mg²⁺ cations), which are described by the overall chemical reaction:

Mg₂SiO_4(s) + 4H⁺ó 2Mg²⁺_(aq) + H₄SiO_4(aq)

As suitable supersaturated conditions are achieved in solution the
carbonation reaction (precipitation of magnesite (MgCO₃)) takes
place:

Mg²⁺_(aq) + CO^2-_3(aq)  óMgCO_3(s)

The kinetics of the weathering reaction are controlled by the hydration
and the dissolution of Mg₂SiO₄ and may be catalyzed by
organic molecules such as oxalate and citrate (Olsen and Rimstidt, 2008; Krevor
and Lackner, 2011; Prigiobbe and Mazzotti, 2011).  The interaction of the organic molecule with
the Mg₂SiO₄ surface depends on the atomic and electronic
structure of its interface, the bulk structure and composition and the impact
of thermodynamic conditions such as temperature and pressure. In case the
selection or the design of the best catalyzing organic compound to enhance the
dissolution process is required a deep understanding of the interaction is
needed.

In this work, we present the results of density functional theory
(DFT)-based electronic structure calculations to investigate the interaction of
water (H₂O) with forsterite (Mg₂SiO₄) at the
most stable stoichiometric (100) surface (Watson et al., 1997).

Periodic ab initio DFT calculations were
performed using the Vienna Ab initio Simulation Package (VASP) (Kresse and
Hafner, 1993) applying the Perdew-Burke-Enzerhoff (PBE) generalized gradient
approximation (GGA) with pseudopotential based upon projector-augmented waves
(PAW)-type. The morphology and the energetics of the bulk as well as of the
gaseous H₂O molecule are in reasonable agreement with the
experimental data using a cutoff energy for the planewave basis set to 460 eV
and a grid of 4x2x4 for the bulk (Figure 1.a). A slab with four unit cells of
bulk was created with a vacuum space of 20 Å to minimize the interaction between
periodic images and the stoichiometric (100) surface reacted with H₂O.
The surface comprises 14 reactive sites of which eight are oxygen atoms, namely
O1, O2, O3, four are magnesium atoms, namely, Mg1 and Mg2, and two are silicon
atoms, Si (Figure 1.b).

DFT calculations were performed together with bond valence and density
of state (DOS) analysis at 0 K to investigate the morphology and the electronic
states of the (100) surface upon adsorption of dissociated and molecular H₂O.
Ab initio thermodynamics was used to
extend the DFT calculations to 422 K and 100 bar, to provide predictions of the
changes in surface reactivity as a function of thermodynamic variables.

Figure 1. Structure of the
stoichiometric (100) Mg₂SiO₄ surface. (a) Lateral view of
the four layer slab and (b) top view of the surface reactive sites. In red are
the O atoms, in yellow the Mg atoms, and in blue the Si atoms.

We have investigated H₂O adsorption on the (100) Mg₂SiO₄
surface by placing one molecule vertically to the surface and tilted around its
axis with its reactive site (i.e., O or H) 1.9 Å above the reactive surface sites.
The effect of surface coverage has been investigated up to one monolayer of H₂O
molecules. Preliminary results of the adsorption energy (E_ads)
of H₂O defined as

E_ads = E_H -
E_S - E_H2O(gas),

where E_H is the minimum free
energy of the surface with adsorbed molecular or dissociated water and E_H2O corresponds to the minimum free energy
gaseous water molecule, indicate that the hydration of the (100) Mg₂SiO₄
is favored. The surface becomes more stable upon the adsorption of molecular
H2O under different coverage except in one case when H2O reacts with the Mg
site located at Mg2a (Figure 1.b). With the exception of this case, E_adsvaries between -3.596 eV/H₂O
mole and -0.851 eV/H₂O with an average value of -1.345 eV/H₂O,
which is in the range of previously calculated E_ads
by de Leeuw et al. (2000), Stimpfl et al. (2006), and King et al. (2010).
The minimum E_adscorresponds
to the adsorption onto the oxygen atom type O3 located at the O3c indicating
that this is the most reactive site as a matter of fact under this condition
the normalized dislocation at the O3 atom site is 1.43 times the initial
position. The high reactivity of the O3 site was already suggested by molecular
dynamics simulations performed by Liu et al. (2009)
for the study of the dissolution of forsterite with water and oxalate ions
suggesting that the adsorption of water onto this atom was critical for the
release into solution of Mg2 atom.

The effect of temperature, pressure, and entropic effect are
taken into account through ab initio
thermodynamic calculations and the type of adsorption mechanism (i.e, chemisorption and physisorption) at the reactive site and the
subsequent change of its electronic structure are determined by analysis of the bond valence and DOS.

References 

Prigiobbe, V., Mazzotti, M. (2011) Chem Eng Sci 66 24 6544-6554.

Olsen, A. A., Rimstidt, J. D. (2008) Geoch Cosmochim Acta 72 1758-1766.

Krevor, S.C.M., Lackner, K.S. (2011) Int J Greenhouse Gas Control 5
1073-1080.

Watson, G.W., Oliver, P.M., Parker, S.C. (1997) Phys Chem Minerals 25
70-78.

Kresse G., Hafner J. (1993) Phys
Rev B 48, 3115-13118.

Liu, Y., Olsen, A.A., Rimstidt,
J.D. (2006) Am Mineral  91 455-458.

de Leeuw, N.H., Parker, S.C.,
Catlow, C.R.A., Price, G.D. (2000) Phys Chem Minerals 27 332-341.

Stimpfl, M., Walker, A.M., Drake, M.J., de Leeuw, N.H., Deymier, P. (2006)
J Crys Growth 294 83-95.

King, H.E., Stimpfl, M., Deymier, P., Drake, M.J., Catlow, C.R.A.,
Putnis, A., de Leeuw, N.H. (2010) Earth Planet Sci Lett 300 11-18.