Matrix element between hybrids

Bond orbitals are constructed ft om s/r hybrids for the simple covalent tetrahedral structure energies are written in terms of a eovalent energy V2 and a polar energy K3. There are matrix elements between bond orbitals that broaden the electron levels into bands. In a preliminary study of the bands for perfect crystals, the energies for all bands at k = 0 arc written in terms of matrix elements from the Solid State Tabic. For calculation of other properties, a Bond Orbital Approximation eliminates the need to find the bands themselves and permits the description of bonds in imperfect and noncrystalline solids. Errors in the Bond Orbital Approximation can be corrected by using perturbation theory to construct extended bond orbitals. Two major trends in covalent bonds over the periodic table, polarity and metallicity, arc both defined in terms of parameters from the Solid State Table. This representation of the electronic structure extends to covalent planar and filamentary structures. 

A particularly important matrix element is that between hybrids pointed at each other, or into the bond, from two neighboring atoms. We call the magnitude of this matrix element the hybrid covalent energy,... 

Eq, (4-12), as applied to hybrid states, is used in the last step it shows also that = 0. The remarkable result is that the matrix elements are not expected to vary with polarity in an isoelectronic scries. This feature also applies to the intra-atomic terms in the matrix element between bonding and antibonding orbitals on adjacent sites. A similar treatment of the matrix elements ofx indicates that they vary inversely with covalcncy 1 isoelectronic... 

This still leaves an important error in the matrix elements. It can be shown by symmetry that matrix elements between certain wave functions vanish c.g., the matrix clement of d/dx between two band states of T i symmetry is zero just as it is for d/dx between two s states in the free atom. Such dependence on band and wave number could also be readily incorporated by using the decomposition of each state into hybrids or atomic orbitals, again, if one desired, with a single-parameter description of the remaining matrix elements. It would be interesting to see how adequate a description of the spectrum, made entirely in terms of the parameters of the Solid State Table, this would produce. Certainly the main features of the spectrum would be predicted, and a more complete calculation, such as that which is illustrated in Fig. 4-1, can very accurately produce the real spectrum. 

The picture that emerges is remarkable and applies to all tetrahedral semiconductors. The principal peak in the optical absorption comes at an energy determined by matrix elements between p orbitals, rather than between hybrids. To be sure, the s orbitals are necessary for any reasonable description of the energy bands or for calculating the full absorption spectrum, but the strongest features in X2(< ) and, by Eq. (4-4), in Xi(w), and perhaps in all dielectric properties, are dominated by p orbitals. 

Notice from Fig. 6-4 that there are other contributions, designated by - V, to the matrix elements between nearest-neighbor bonds. Two of these contribute for each pair of bonds, each weighted by the coefficient for the anion hybrid and the cation hybrid (Eq, 3-13). The two values are not required to be equal by... 

The formulation proceeds exactly as it did for bond orbitals in Section 6-B, but the matrix elements and now refer to matrix elements between antibond orbitals. We take the convention of letting the coefficient of the hybrid on the anion be positive. (See Fig. 6-7.) Then, when we write these matrix elements in terms of interhybrid matrix elements, we use the coefficients from Eq. (3-16) rather than (3-13), so that Eq. (6-13) is modified for antibonds by changing the sign before p in each expression ... 

The hybrid energy and Vj were obtained in Problem 3-2, where a value of the bond energy Cb was found. A value for the matrix element between two hybrids on the same atom, — Ki, was also obtained, and only its contribution, — V, jl, to the interbond matrix element need be retained. 

As in the simple tetrahedral solids, the bond orbitals are not eigenstates because matrix elements exist between them and bond orbitals on adjacent sites. We can in fact use the values for the matrix elements between two hybrids on the same silicon atom, — Vy—see Eq. (3-5)—to obtain the matrix elements between adjacent bond orbitals in Si02 ... 

To calculate the properties in terms of the perovskite gliosts, we will need a covalent energy, equal to the matrix element between a (x-oriented [Pg.457]

For analysis of the transition metals themselves, the use of free-electron bands and LCAO d states is preferable. The analysis based upon transition-metal pseudopotential theory has shown that the interatomic matrix elements between d states, the hybridization between the free-electron and d bands, and the resulting effective mass for the free-electron bands can all be written in terms of the d-state radius r, and values for have been listed in the Solid State Table. 

Fig. 26. Calculated 4f contribution and the energy dependent hybridization matrix element between Ce 4f and chalcogen p states for Ce chalcogenides. The 4f contribution is labeled spectral density and the matrix element is drawn dotted. (After Takeshige et al. 1985.)...

This model can be applied to the ideal rans-polyacetylene chain assuming sp2 hybridization of the carbon px and py orbitals with the carbon 2s electron to form the fully occupied a bands of the polymer. The pz orbital extends perpendicular to the plane of the polymer chain. If the carbon atoms were uniformly spaced, the ideal tight-binding band with transfer integral t given by the matrix element between adjacent Pz orbitals would apply. As there is only one pz electron per carbon atom, the resulting n band would be only one-half filled with band structure and density of states shown in Figure 2. 

---