APPENDIX Degenerate orbitals In several of our model systems, including square H4 and triangular H„ some of the orbitals are degenerate, i.e., they have the same energy. We find degenerate orbitals such as these in geometries where there is a rotation axis of order higher than two. Obviously the triangle and the square have axes of order three and four respectively. Besides their energetic degeneracy these MOs have a number of other characteristics. (i) As a result of a symmetry operation an MO is usually transformed into itself (symmetric) or into minus itself (antisymmetric). So for H4 a rotation of 2ji/4 = n/2 around the z axis (this is the four-fold rotation axis perpendicular to the plane of 1 (4-19) and 4 into -c£4 (4-20). The situation the molecule) transforms 4>^ into 4-19 4-20 is different for the degenerate orbitals < other, namely 2 -» ->s (4-21) and 3 orbitals located at the center of the square (p and (j>3. They are transformed into each \ (4-22). The same holds for px and py ■ py and p, -+ -P»). In a very 4-21 4-22 2 general way certain symmetry operations transform degenerate MOs into linear combinations of the starting functions. Taking, for example, triangular H3, one can show that a rotation of 2w/3 around the z axis (C3) takes 4>2 to -fa + s/fa THE FRAGMENT ORBITAL METHOD; APPLICATION TO SOME MODEL SYSTEMS 127 (4-23) and <£3 to -jfa - fa (4-24). Degenerate orbitals must always therefore be treated as a pair and never individually. 4-23 4-24 (ii) Although the symmetry operation does not transform each degenerate orbital into itself or minus itself, the new orbital has the same energy as the starting one. In effect, if tj>; and (pt are two functions of the same energy then all normalized linear combinations of the type X(pt + jupj are equally good functions with the same energy (see Section 2.1.2c). (iii) This leads us to conclude that the degenerate orbitals which we obtained for systems such as square planar H4 (Figure 4.1) and triangular H3 (Figure 4.5) represent just one solution out of a whole host of possibilities. In general one can replace such pairs of orbitals with a pair of linear combinations of the form ', = i cos 9 + i sin 0 (J)) = - 4>t sin 0 + 4>} cos 0 We can easily show that the MOs ip] and cp) are normalized and orthogonal just like 4>t and 2) have radial nodes as shown in 2-6 for the 2s function, we shall ignore these in generating our orbital diagrams. Only the overlap with the outermost part of the orbital is chemically important at normal internuclear distances.) 5.1.1. Symmetry properties of the fragment orbitals Consider the collection of symmetry elements the two fragments have in common. This is effectively the collection of symmetry operations for the linear All, molecule. There are an infinite number of these. For example all planes which contain the z-axis are planes of symmetry for the two fragments (5-2). In the same way a rotation of any 5-2 angle around z leaves the positions of the nuclei unchanged. Other elements of symmetry include the xy plane perpendicular to z and containing the atom A, the inversion center, located at A, etc. A general treatment of the symmetry problem would study the behavior of the orbitals as a result of all of these symmetry operations but we will content ourselves here by making a judicious selection of just one symmetry element which will allow us to provide a symmetry classification good enough to be able to decide which pairs of orbitals may interact via non-zero values of their overlap integral. The px orbital on A (5-3) is antisymmetric (A) with respect to the yz plane, a nodal plane of this orbital. Contrarily the orbitals 2 and py (AS) is non-zero since the contributions from py-lsH overlap are both of the same sign (5-14). But now consider the case of four orbitals of SS symmetry. For the pairs (s, (pj and (px, fa) the overlaps involved are different from zero since all the contributions between the central orbital and each of the lsH orbitals are of the same sign (5-15 and 5-16). However, consider the overlap integrals between s and lt py and 2 as shown in 5-19. The pz orbital is not involved in any interaction. The actual p> —— - 5-19 « — form of the diagram will vary from one AH3 molecule to another since the energies of the central atom s and p orbitals depend upon the identity of A, but the interactions shown are the same irrespective of the identity of the system. As described before for the AH2 molecule wc usually attach group theoretical labels to describe the orbitals of the fragment. The s orbital is labeled a\. Here, a describes a non-degenerate level, just as a did in the linear molecule, and the single prime a function symmetric with respect to reflection in the plane perpendicular to the three-fold axis (z). The pz orbital is labeled <& antisymmetric with respect to reflection. The pairs of degenerate levels carry an e label, just like the label ji in linear molecules. Both px, p, and l (bonding) lies a little below and (j)2 and <£3 (antibonding) a little above. Their splitting is small since the interaction between the hydrogen Is orbitals is small as a result of the large H—H separation. First of all we readily see that the p, orbital (la'Q is unchanged in energy. The orbitals (j>1 and 2s, both of a', symmetry interact to give a bonding orbital (\a\) and an antibonding orbital (2a\). Similarly the orbital pairs, (j>3 and px (e'x), and 2 and py (e'y) interact to give a bonding pair (le'x and \e'y) and an antibonding pair (2e'x and 2e'y). The orbitals \e'x and \e'y are degenerate, as are 2e'x and 2e'y; as shown in exercise 5.1, since the overlap integrals associated with the x and y partners of a degenerate pair are equal the resultant molecular orbitals are degenerate. The origin of this degeneracy comes just as in triangular H3 from the presence of a three-fold rotation axis in the molecule. The MOs of trigonal planar molecules thus divide into three groups (5-21). (i) Three MOs bonding between the central atom and the hydrogen atoms. These are la'lt le'x and te'y, in-phase combinations of the fragment orbitals s), 138 BUILDING UP MOLECULAB ORBITALS AND ELECTRONIC STRUCTURE 5-21 ( BH4~, then we can readily see the origin or the Lewis acid properties of such a species. 5.3. Tetrahedral AH4 molecules The natural decomposition for a tetrahedral AH4 molecule is into a central A atom and a tetrahedron of hydrogen atoms. The levels of the latter, a tetrahedral H4 unit were studied in the previous chapter. The fragment orbitals for the atom A are just its valence s and p orbitals. 5.3.1. Symmetry properties of the fragment orbitals The fragments A and H4 have many symmetry elements in common, among them (5-23) six planes of symmetry containing two A—H bonds (xz and xy are two 5-23 140 BUILDING UP MOLECULAR ORBITALS AND ELECTRONIC STRUCTURE examples), three C2 axes which bisect opposite pairs of H—A—H angles (x for example) and four C3 axes coliinear with the A—H bonds. As before we will just retain two planes (xz and xy) in order to distinguish between the orbitals concerned. There are two SA orbitals (tf>4 and pj symmetric with respect to xz and antisymmetric with respect to xy, two AS orbitals (3 and py) and four SS orbitals (t>2, s and pj as shown in 5-24. The fragment orbitals thus separate into three SA AS 5-24 groups. Recall that orbitals belonging to different groups may not interact, but we still have to look carefully at the overlaps between orbitals within each group. We can see that overlap between the pz and (fiA orbitals is non-zero since the individual overlaps between pz and \sH orbitals arc of the same sign (5-25). The same is true for p}, and 3 (5-26). For the orbitals of SS symmetry we need to consider the two pairs (s, 2) and (px and i)- -fust as we showed for the related AH3 case, the overlap integrals between these pairs are identically zero (5-29 and 5-30). In both cases the two positive overlap integrals are exactly cancelled by the two negative overlap integrals. Also, as in AH3, these zero overlap integrals between orbitals of the 'same symmetry' come about because of the reduction of the tetrahedral symmetry to just the two planes xz and yz. Use of the full symmetry removes this problem. In conclusion, just as in all of the preceding examples, only the orbitals of the same symmetry, with non-zero overlap may interact. The construction of the molecular orbital diagram for tetrahedral AH4 thus reduces to a question of the four pairs of interactions of 5-31, the variation from one molecule to another being set by the central atom s and p orbital energies dependent upon the identity of A. We will use in what follows the group theoretical labels for these orbitals (5-32). Both s and 2), t2y (py and tp3) and tu (pz and 04). Notice that here (and in the AH 3 molecule too) there is no center of symmetry unlike the situation in linear AH2. Accordingly the subscript g or u which described the behavior with respect to inversion is inappropriate here. 5.3.2. MOs of tetrahedral AH4 molecules The fragment orbital interaction diagram of Figure 5.3 corresponds to the case of CH4 where c, ■19.4eV andc. • 10.7 eV. The H4 fragment levels lie just below {2, 2 and px, between 3 and py and between 2, 3 and 4 on H4 with respectively px, py and pz on A, are equal. This triple degeneracy comes about because or the high symmetry or the tetrahedral molecule. The molecular orbitals of tetrahedral AH4 molecules divide into two sets (5-33). 2a, 1a, 5-33 - 9^ (i) Four MOs bonding between the central atom and the hydrogens. These are la,, \t2x, U3y\ \t2z, in-phase combinations of the fragment orbitals (0,,s), ( (^3. Py) and (04, pz) respectively. The deepest lying orbital, la,, arises via interaction with the deepest lying fragment orbitals. (The three lr2 orbitals are degenerate.) (ii) Four MOs antibonding between the central atom and the hydrogens. These are 2a,, 2t2x, 2tly and 2t2z out-of-phase combinations of the same fragment orbitals. (The three 2t2 orbitals are degenerate.) 5.3.3. Application to the electronic structure of CH4 In the methane molecule, with a total of eight valence electrons, the lowest four molecular orbitals are doubly occupied in the electronic ground state to give the electronic configuration \a\ Iff, \t\y \t\z or la? \t\ as in 5-34. These four occupied bonding orbitals correspond to the four C—H bonds of the Lewis structure. The central atom uses one s and three p orbitals to form these bonds. Just as in our earlier AH2 and AH3 examples it is not possible to make a one-to-one correspondence between a single delocalized molecular orbital and a particular C—H bond. In the la, orbital the bonding character is the same between the central atom and each of the hydrogens since the hydrogen coefficients are all equal. The same is true for the ti2x orbital. Here all of the coefficients are equal in absolute magnitude and each of the A—H bonds make the same angle (one half or the 'tetrahedral' angle, 109.5°/2) 1 144 BUILDING UP MOLECULAR ORBU ALS AND ELECTRONIC STRUCTURE 14» 4h 8*4 ^< C ) 1a, S-34 with the axis (x) of the px orbital. On the other hand the U2y orbital is bonding only between carbon and H, and HbJ and the 112. orbital is bonding in the same way between carbon and Hc and Hd. It is the collection of four occupied MOs taken together which lead to four equivalent C—H bonds. Finally, all of the occupied MOs are not of the same energy, lat lying deeper than It,, This prediction from molecular orbital theory is confirmed experimentally via the photoelectron spectrum. There are two ionization energies which differ by about 10 eV. To conclude, just as in the earlier examples it is necessary to distinguish between the equivalence of the four C—H bonds and the non-equivalence of the four occupied molecular orbitals (split into the two sets la, and 1t2). EXERCISES 5-1 Overlap integrals between fragment orbitals in AH3 We will consider the orbitals and S3 = <^3 | px) are equal. (Call S0 the overlap integral between a p orbital and a 1sH orbital lying along the p orbital axis at a distance d = A—H.) Overlap integrals between fragment orbitals in AH^ Show in the same way that the overlap integrals between the pairs of fragment orbitals s2 = (