Page 218 - Adsorbents fundamentals and applications
P. 218

MOLECULAR ORBITAL THEORY CALCULATIONS  203

            where H is the Hamiltonian operator, E is the energy of the particle (an electron
            or nucleus), and ψ is the wave function. The product of ψ with its complex con-
                                         2
            jugate (ψ ∗ ψ, often written as |ψ| ) is interpreted as the probability distribution
            of the particle (or electron).
              Electronic structure methods are characterized by their various mathematical
            approximations to its solution, since exact solutions to the Schr¨ odinger equation
            are not computationally practical. There are three classes of electronic structure
            methods: semi-empirical methods, density functional theory (DFT) methods, and
            ab initio methods.


            8.2.2. Semi-Empirical Methods
            Semi-empirical methods are parameters derived from experimental data to sim-
            plify the computation. They are used to solve an approximate form of the
            Schr¨ odinger equation that depends on having appropriate parameters available
            for the type of chemical system under investigation. Different semi-empirical
            methods are largely characterized by their differing parameter sets. One of the
            most commonly used semi-empirical computer programs is MOPAC, developed
            by Stewart (1990). A commercial package based on MOPAC is available from
            Fujitsu Company of Japan. MOPAC includes the semi-empirical Hamiltonians
            MNDO (modified neglect of diatomic overlap), MINDO/3 (modified intermediate
            neglect of differential overlap), AM1 (Austin model 1), PM3 (parametric method
            3), MNDO-d, and PM5. These methods have been calibrated by using experimen-
            tal data for thermodynamic properties such as heats of formation (Pople et al.,
            1965; Baird and Dewar, 1969; Bodor et al., 1970; Murrell and Harget, 1972).
              The advantage of semi-empirical methods is in the economy of computation.
            These programs can be performed with a personal computer. However, the bond
            energies calculated from these programs are substantially higher than the actual
            values. In many instances, the bond energies are divided by an empirical factor
            of 5 to give an indication of the real values. MOPAC provides the most accurate
            energy calculations among these methods, while the results in bond energies are
            still substantially over-estimated (e.g., Chen and Yang, 1997).


            8.2.3. Density Functional Theory Methods
            The approach of density functional theory (DFT) was developed in the 1960s by
            using mathematical functions, called functionals, to describe the electron density
            (Hohenberg and Kohn, 1964; Kohn and Sham, 1965). A summary of the theory is
            given by Foresman and Frisch (1996), and a detailed description is available from
            Parr and Yang (1989). DFT methods are attractive because they include the effects
            of electron correlation (i.e., electron–electron interaction) in the energy. Devel-
            opments in DFT have led to nonlocal (gradient-corrected) functionals — BLYP
            (Becke-Lee-Yang-Parr), and to hybrid functionals — B3LYP. The nonlocal func-
            tionals account for the non-uniformity of the overall electron distribution. The
            hybrid functionals use a linear combination of the Hartree/Fock (HF) and the
   213   214   215   216   217   218   219   220   221   222   223