Jerry March, "Advanced Organic Chemistry", 3rd ed. John Wiley &
Sons (1985), pp. 961-962 says (copious references omitted): "It
was mentioned above that even alkanes undergo Wagner-Meerwein
rearragements if treated with with Lewis acids and a small
amount of initiator. An interesting application of this reaction
is the conversion of trycyclic molecules to adamantane and its
derivatives. It has been found that _all_ tricyclic alkanes
containing 10 carbons are converted to adamantane by treatment
with a Lewis acids such as AlCl_3. If the substrate contains
more than 10 carbons, alkyl-substituted adamantanes are
produced. Some examples are [ Fig: tetramethyladamantane from
norbornene + cyclic polyenes Diels-Alder adducts and diamantane
from hydrogenated annelated aromates ]. If 14 or more carbons
are present the product may be diamantane or a substituted
diamantane. These reactions are successful because of the high
thermodynamic stability of adamantane, diamantane, and similiar
diamond-like molecules. The most stable of a set of C_nH_m
isomers (called the stabilomer) will be the end product if the
reaction reaches equilibrium. Best yields are obtained by the
use of "sludge" catalysts (i.e. a mixture of AlX_3 and t-butyl
bromide or sec-butyl bromide) or by passing the substrate in the
gas phase through a tube containing a chlorinated
platinum-alumina catalyst. Though it is certain that these
adamantane-forming reactions take place by nucleophilic 1,2
shifts, the exact pathways are not easy to unravel because of
their complexity. Treatment of adamantane-2-14^C with AlCl_3
results in total carbon scrambling on a statistical basis".
Obviously, small carbon diamondoids are trivial to obtain in
good yields (the almost-forgotten adamantane rush preceded that
of the buckball craze decades before). Through this is not
mentioned in literature, I guess large branched/cycled
carbohydrogens rearrange to (defect-rich, but who cares)
diamondoids easily enough. A statistical small carbon cluster
synthesis is therefore easy to accomplish. A random CAS
(Chemical AbstracS) inquiry from years ago said that
quantitative polyfluorination of adamantane is easy to
accomplish by an electrochemical process (anodical fluorination
of adamantane in anhydrous HF).
Carbon-carbon bond formation via Wurtz-class metalloorganyles
(Grignard, lithium, copper, or whatever) reactions is a
time-honoured technique of organic synthesis.
Manipulative proximal probe spectroscopy and GA solution of the
knapsack problem are the new pungent spices in the statistical
pot.
A possible scenario would involve cooking a wild mixture of
different polycyclic alkanes at high temperatures
(alternatively, in the gas phase), then a flash chromatography
of the quenched mess to get rid of the coarsest muck. Using de
novo designed antibody affinity columns, an arbitrarily precise
separation stage could ensue. Alternatively, halogenated and/or
metallated adamantanes could be spread out on HOPG tablets at
cryo UHV conditions, and be mapped via a STM/SPM scan. Those 3d
puzzle pieces best fitting the virtual machinery mold (which
requires a _hefty_ (GA) crunch run), and then be assembled via
manipulative proximal probe microscopy. Once in place, they
crosslink by multiple C-C instantly, irreversibly. (Using
phlegmatized surface groups, this should be worthwhile to try
even in solvated phase, xref autoassembly via complementary
surfaces).
By this means, a first generation of diamondoid devices could be
assembled. If they are stiff enough to avoid progressive defect
runaway at autoreplication, the bootstrap bottleneck is thus
shattered to pieces.
ciao,
'gene
P.S. Please observe that PPM, biochemistry and numerical
modeling star in above speculative scenario, synthetic chemistry
is, once again, the Assipattle here.