An old term, after Wurtz (1885), preparation of higher alkanes
from alkylhalogenids:
MeI + 2Na -> MeNa+NaI
MeNa + MeI -> Me-Me + NaI
(it is also handy for a number of other things, as cooking
alkylosylanes). A variation of this is Corey-House, coupling
alkylohalogenids with lithiumdialkylcuprates (R_2CuLi), under
much milder conditions.
I guesstimate one can even substitute pretty specifically, e.g.
cooking substituted adamantanes by aforementioned Wagner-Meerwein,
oxidizing the sticking out alkane groups to COOH, then turning
them into -Br via Hunsdiecker.
> assume that you mean C-C bond formation by activated carbon
> nucleophiles (Grignard etc). This proceeds by back side attack
> with consequent steric inversion. This is of course impossible
> with a polycyclic diamondoid target.
You must be thinking of the alkohol synthesis from
aldehydes, via Grignard route. This is not a sole possible
version however, e.g. tin organyles and silver-halide (iodide)
reactions involve radicals, some things occur on metal surfaces
via SET. Metalloorganyles are a versatile lot.
> What other reactions might work for subunit linkage?
Not too many. One might consider carbenes (I once worked
briefly with these, they also tend to produce polymers when
adding to alkenes), radicals (e.g. via a good leaving group, as
diazo derivates), crosslinking of complementary groups. S-Hg-S
might also be considered.
> Diels-alder won't work because it requires highly unsaturated
> reactants, which are not permitted by the harsh conditions for
> rearrangement into diamondoid building-blocks. Radicals might
> work, though any easily removed radical group would probably
> also be destroyed in the building-block phase.
The creation of building blocks occurs in 2-3 stages, the
first of them very harsh, involving massive rearrangements
(hydride shifts, nonclassical carbocations), the second, e.g.
(electro)halogenation, being much milder. Metallization can be
arbitrarily mild. Once the halogenated diamondoid (notice that
chemists do not use this term) core is in place, it can be be
manipulated in many ways.
'gene