I had really looked forward to this chapter as organometallic chemistry has exploded since I left the field in ’62. The explanation of the basics of the bonds formed between the metal and the organic substituents seem rather skimpy and at odds with much of the preceding 704 pages. Instead of orbital drawings, there are lines drawn between atoms, something typical of English and Cassidy’s book (copyright 1956) which we used as undergraduates in ’58. I had looked forward to seeing d-orbitals (or hybridizations of them with s and p orbitals) in action, with electrons swooping like swallows between them as they changed, similar to earlier diagrams.
D orbitals are a form of the spherical harmonics for angular momentum with quantum # 2. The 5 of them aren’t even shown in this chapter, just a referral back to Figure 1.26 on p. 60. If they are hybridized with s and p orbitals the results aren’t drawn (as they are over and over for sp1, sp2 and sp3). Why not? Too are they hard to draw? Don’t we know what such orbitals look like?
Then on p. 712 d orbital energy level splitting is introduced. Crystal field theory is mentioned, said to predict the splitting, but not further discussed. A referral to ‘a text book focused on organic or organometallic chemistry’ is advised. Well, I never studied the stuff, so I tried to look it up in Cotton’s “Advanced Inorganic Chemistry” — nothing in there, surprisingly, even under ligand field theory. I decided to plunge ahead, even though I never like to read past something unclear to me (in chemistry at least. In mathematics, it happens all the time).
It was the next statement that totally blew me away: ”Suffice it to say, a general rule for predicting the splitting is that d orbitals that lie along the bonds are raised in energy”. Hello? Everything up to now has implied that orbital overlap is a good thing and lowers molecular energy — remember strain energy.
So it was definitely time to read about crystal field theory. The first thing I read was pp. 288 -293 of Berry’s PChem book, “Electronic Structure of Complexes”. The description of crystal field theory (developed for crystals, hence the name) was quite clear. It starts from the distribution of ions in space (never mind how they got there). This is strikingly reminiscent of VSEPR (Valence Shell Electron Pair Repulsion).
As they note, an atom by itself with its electrons is inherently spherically symmetrical. Only the orbitals with an angular momentum (e.g. everything other than an s orbital) will be affected by the disposition of ions around the central atom. Even if we have as symmetric a disposition of surrounding ions as we can get, the electric field they set up is not the same in all directions (e.g. not spherically symmetric), so orbitals with angular momentum will be affected by the field.
A bit of thought shows you that if you want to put 4 negative charges an equal distance from a central positive charge, and put all 4 as far away from each other that they can get (e. g. on the surface of a sphere), you wind up with a tetrahedron. Similarly, 6 negative charges, given the same constraints, produce an octahedron.
So if we have the positive metal nucleus surrounded by negative ions, the filled orbitals of the metal pointing toward the ions will experience repulsion by the electrons of the ions, raising their energy. Orbitals not pointing this way will be unchanged. This I understand clearly, and it’s how previously energetically identical orbitals (say px, py and pz) are split into orbitals of different energies. But what holds things together in this situation? Where’s the bonding? I was taught to think orbital overlap is a good thing.
The discussion on p. 290 shows that (if you shut your eyes and somehow pretend that bonding is still present) in an octahedral complex dz2 and dx2-y2 will be raised in energy, while dxy, dyz, and dxz will be lowered. Similarly in a tetrahedral field exactly the opposite occurred. This is quite clear when you look at the pictures.
Then on p. 292 — the denouement. How bonding actually occurs. It’s covalent, and somehow molecular orbitals make everything OK. No pictures are given. All valence shell orbitals of the metal (s, p, d) and the ligands are used in their construction.
Perhaps bonding doesn’t matter. Somewhere it is said that the organic ligands in transition metal complexes act differently, with unstable ligands (like carbenes) being stabilized and stable ones (like C-H bonds) labilized. Maybe the central metal atom acts like the alcohol used by pimply adolescents everywhere, trying to get their date a little high and doing things they ordinarily wouldn’t.
Any help the cognoscenti out there can give will be appreciated. I plan to look at a variety of books on the subject — Crabtree, Hegedus, Mathey and Hill, which I’ve checked out of the local college library. Unfortunately, what appears to be the latest and greatest, Hartwig has been checked out, so I may have to buy it.
This post is going to be work in progress, with more added (and dated) as clarity (hopefully) emerges. Now back to chapter 12 of Anslyn and Dougherty.