Overview — A fabulous chapter, with more new (and bizarre) chemistry than the rest of the book. I wonder what percentage of the ‘average’ total synthesis done today uses transition metal chemistry.
Even so the chapter is a disappointment. While there is stereochemistry of the organic moieties attached to the transition metal, the disposition of the ligands in space isn’t given for the most part. The exquisite dance of the orbitals to be found in fragmentation reactions (to give one example) is nowhere to be found. There is one picture of a d-orbital (p. 1073) in the discussion of back bonding.
To be sure they note (p. 1070) that a lot of work has gone into mechanism, but that the results ‘remain speculative’. So my disappointment may be with the state of our present knowledge rather than the way the chapter is written.
In the suggestions for further reading we find “Most textbooks of organometallic chemistry favor the inorganic approach of facts rather than explanation.” I’d say that is true of this chapter, where most of the chemistry is explained as sequences of the following 5 basic reactions
l. Oxidative addition
2. Reductive elimination
3. Migratory insertion
4. Beta hydride elimination
5. Cross coupling
The mechanisms of these 5 aren’t gone into (say the way Sn1 and Sn2 are explained in the rest of the book).
Perhaps the situation here is like the early days of quantum mechanics, when things were being calculated, and results obtained, with little introspection of what’s going on under the hood (although Bohr would say that isn’t a scientific question). Surely somewhere calculations have been done to show why coordination to a metal changes the reactivity of organic compounds so much (making alkenes coordinated to Pd++ electrophilic to take one example).
They do say Hartwig’s book does go into these things, but it’s quite expensive and doesn’t seem to be available in any of the local college libraries.
Can any of the readers out there send a link to a PDF answering some of these questions in a comment? I’d be grateful.
1070 — Why is the 4s orbital of lower energy than the 3d, the 5s than the 4d, the 6s lower than the 5d. The explanations I remember have always seemed like hand waving. Any comments or explanations>
The ‘explanation’ for the stability of 16 electrons in Ni, Pd, Pt is weak. ‘Adopting a square planar geometry’ — but as opposed to what other geometry?
1073 — ‘dsp’ orbital “derived from the vacant d, p and s orbitals of the metal” — why would the s orbital be empty?
In the terms oxidative and reductive, remember its the metal that’s being oxidized or reduced.
“You do not need to understand all the bonding properties of metal complexes” — OK, but how about a reference to a place where this is explained? Perhaps the reference to Hartwig at the end of the chapter is what I want.
1074 — I assume that the X in the second reaction sequence is halogen.
How do we know that the methyl iodide addition to the Iridium complex is trans. It’s nice to have a reference to stereochemistry (however small) in the first 5 pages of the chapter. Are transition metal complexes with 4 ligands always square planar? Can they be tetrahedral?
1075 — 4 coordinated Pd is shown to be square in the diagram. Is this always true? A statement to that effect would be good.
1076 — Very hard for me to see how the example in the top row of structures with Wilkinson’s catalyst is a migratory insertion (I guess the alkene inserts into the M-H bond — probably because I usually think of hydrogen as the moving atom). Carbonyation (2nd row) is much clearer.
1076 — In the carbonylation of Fe(CO)x, drawings of the complexes imply that they are trigonal bipyramidal or octagonal, but this is never stated explicitly.
1076– Why are alkyl groups poorer ligands than CO (lack of backbonding perhaps?).
p. 1078 — Beta hydride elimination contains a semantic trap — although hydride is eliminated from the carbon skeleton, it winds up bound (italics) to the metal. At last, some stereochemistry “In more complex structures, the metal and the hydride must be syn to each other on the carbon for the elimination to be possible”
1078 — “most syntheses of organic molecules of any complexity will now involve palladium chemistry in one or more key steps.” Wow ! ! That being the case, what is it doing in the last 9% of the book.
1079 — “The presence of hydrogen at an sp3 carbon in the beta position must be avoided”. It’s because beta-hydride elimination is quite exothermic
M-C (30 kiloCalories/mole) —> M-H (60 kiloCalories/mole)
C-H (100 kCal/M) —> C=C (148 kCal/M) — so 78 kCal/mole releasef as heat.
Things that release heat and gas are known as explosives.
p. 1080 — Watch out — the carbometallation step in the Heck cycle shown, encompasses a bunch of steps — see the carbopallidation reaction scheme on p. 1079.
p. 1081 — I found the mechanisms of Pd++ reduction at the top, extremely confusing and hard to follow.
p. 1081 — Some stereochemistry at last — “the C-Pd and C-H bonds have to eclipse one another for the Pd-H bond to form.
p. 1082 — More stereochemistry — Palladium is very sensitive to steric effects — well not the ion itself, but with all the junk hanging off it (triphenyl phosphines etc. etc.) it has to be bulky.
p. 1083 — the palladium couplings are so diverse. Does anyone use Grignards or silyl enol ethers etc. etc. anymore in synthesis?
I wrote the following to a practicing organic chemist involved in med chem drug development.
I’ve just finished the 32 pages of Ch. 40 of the new edition of Clayden’s textbook of organic chemistry concerning Organometallic chemistry. The number of new (and unusual) reactions is simply staggering, and this is only a 32 page account. Hartwig’s book (which I’ve not read or even seen) has some 1160 pages probably has even more novel reactions. To an old Woodward grad student, these reactions should have revolutionized synthetic organic chemistry.
My questions to you are
l. Is this true
2. If true, how often are they used in synthesis
a. academic type of stuff that’s never been done before
b. industrial and med chem type — e.g. day to day work making new drug candidates
I got the following back
The workhorse metal-catalyzed reactions are used a great deal, and it’s gone as far as affecting the kinds of molecules that even get made. But some of these reactions have a reputation for being very finicky about their substrates and conditions – they work on the examples in the paper, but can’t be extended so easily, so people are worried (after they’ve been burned) about trying some of them.
p. 1084 Coupling an alkyne to an alkene in the Stille reaction is truly magical.
Out of sequence, and rather delayed because of family events, but the hexahydro Diels Alder reation [ Nature vol. 490 pp. 208 – 212 ’12 ] is not to be missed, showing that there’s all sorts of new organic chemistry to be discovered.
p. 1086 — bottom row of reactions. The lack of steric hindrance in the coupling reaction might be due to the fact that the central Pd atom is large. The following web site http://www.webelements.com/palladium/atom_sizes.html gives a variety of radii for Pd. With a coordination number of 3 the single bond covalent radius is 1.2 Angstroms for Pd (almost as much as a whole C-C bond of 1.54 Angstroms, so the molecules bound to Pd have room to fit in. Because they are held to Pd there by the bonds, they are already in a position to react with each other, even though in solution, such a close approach would be improbable. The atomic radii of the transition metals are nowhere mentioned in this chapter.
1087 — Sonoshagira adds another Japanese name to an already impressive list of named transition metal chemistry reactions in the chapter — Suzuki, Kumada, Negishi. Was there one old Japanese master and are these his students?
1093 — there is a missing R on the benzene ring in the fourth benzene in the first reaction sequence at the top of the page.
1096 — At last, an explanation for one of the unusal reactivity patterns of transition metal chemistry — the drawing away of the pi electrons of an olefin toward the metal. Probably the partial filling of the pi* orbital by back bonding doesn’t hurt either. Do the cognoscenti have any thoughts on this one?
1096 — “CuCl2 oxidizes Pd(0) to Pd++ and is itself oxidized back to Cu++ by oxygen.” The itself should be the Cu(0).
1098 — The synthesis of claviciptic acid by Hegedus is elegance itself.