.
Angewandte
Communications
methyl-transfer agent is the tetrahydrofolate 14. An impor-
transition state of this reaction. To study the novel cleavage
reactions further, two other substrates were prepared, the
pyrrolidine 18b and the piperidine 18c.
tant question concerns the level of activation that could be
[12,13]
ꢀ
achieved for the Me N bond in tertiary amine 14,
ꢀ
a moiety that would normally be expected to be completely
unreactive to methyl transfer. The current hypothesis is that
cation 14’ (a conjugate acid of 14 with protonation at N5) is
attacked by a zinc-bound thiolate of homocysteine.[13] How-
ever, double protonation, as in 14’’, might lead to significantly
enhanced reactivity, thus allowing attack by thiol 11 rather
than by the corresponding thiolate. But how reactive could
If these substrates underwent analogous C N bond
cleavage reactions with triflate ion, then a product containing
a triflate ester should be formed. The reaction of 18b and 18c
with triflic anhydride in anhydrous dichloromethane again did
not lead to the respective salts 15b and 15c, but instead gave
the alkyl triflates 21 (73%) and 22 (90%) in very good yield,
thus confirming the hypothesized substitution reaction
involving triflate anion. These reactions must proceed by
SN2 mechanisms (see below) because the carbon atom at
which substitution occurs in 15a–c is a methylene or a methyl
carbon atom and because the reactions afford a single product
in high yield (no alkene resulting from elimination from 15b
and 15c was observed). The increased reactivity of these
systems, relative to 5, is remarkable. The formation of an alkyl
triflate in essentially quantitative yield, as described herein,
means that the amidine cation 16 is approximately 100-fold
better as a leaving group than triflate ion.[14b,c] To determine
the mechanism of demethylation, we carried out a computa-
tional investigation on the reactive species outlined in
Scheme 3.
ꢀ
a Me Nsp3 bond be?
We envisioned that amidine dications[8] could be used for
exploring the limits of synthetic, as opposed to enzymatic,
activation toward alkylation.[14,15] An important driving force
for the methyl transfers from 5 and 6, may be that the
demethylation reveals a nitrogen lone pair, which can deloc-
alize over the heteroaromatic ring, in the respective products
7 and 8. Herein, we introduce a new type of amidine dication
which was designed to be more reactive in demethylation
reactions (Scheme 3). Salt 15a should undergo demethylation
to afford amidinium salt 16. The lone pair of the demethylated
The initial reaction of 18a with triflic anhydride to form
salt 19 (R = R’ = Me), which contains both a triflate substitu-
ent and a triflate counterion, was calculated as being
exothermic (DG = ꢀ3.5 kcalmolꢀ1). Calculations of the sub-
sequent steps in the reaction show that they can occur either
in the presence or in the absence of the triflate counterion
(blue curve and red curve in Figure 1, respectively). The
intermediate, 20, which is formed in a facile reaction, is
strongly favored thermodynamically relative to 19. The
counterion has little effect on the energetics of the step that
forms intermediate 20, which contains a tetrahedral triflate-
bearing carbon atom. Calculations on a reaction involving
direct abstraction of the methyl group by the sulfonyl oxygen
atom of the triflate substituent via a 6-membered cyclic
transition state derived from 20 suggest that it is not feasible.
Instead, the triflate substituent spontaneously dissociates
from the central carbon atom to be placed in the alignment
ꢀ
necessary to cleave the C N bond (see below).
However, in the subsequent step, the dissociation of the
triflate moiety to form the planar dicationic species (15a), the
counterion plays a stabilizing role; the presence of the
counterion leads to a slight lowering of the barrier to
dissociation as well as to a decrease in the endothermicity
of the reaction to 5.8 kcalmolꢀ1 (compared with 6.4 kcal
molꢀ1, which is the value calculated when no counterion is
present, see Figure 1). Despite the presence of the counterion,
the formation of 15a is thermodynamically disfavored, and
the reverse reaction (that is, 15a!20) occurs with a very
small barrier (0.5 kcalmolꢀ1). Therefore, the lifetime of 15a is
extremely short and, consistent with the experimental results,
is unlikely to be observed. Conversely, the transformation of
20 into 16 is a strongly exothermic reaction (ꢀ21.9 kcalmolꢀ1)
with an accessible barrier (20.9 kcalmolꢀ1; see Figure 1). For
the transformation of 20 into 16, the inclusion of the triflate
counterion was found to play an important role in lowering
the barrier to demethylation. However, the demethylation
Scheme 3. Preparation and reactions of amidine dications 15.
nitrogen atom is appropriately placed for extensive delocal-
ization, which can be reflected in both the reaction kinetics
and thermodynamics. The initial challenge was to explore the
reactivity of 15. In the preparation of target salt 15a,
formylation of 17a afforded 18a (Scheme 3). Treatment of
the resulting formamide with triflic anhydride[15] did not give
salt 15a, but instead gave the expected product of demethy-
lation of salt 15a, that is, 16, exclusively in 84% yield upon
isolation. The completely selective formation of 16 and the
absence of product arising from demethylation of the sp2-
hybridized nitrogen atom in 15a supported our thinking that
ꢀ
the electrons in the scissile C N bond would be stabilized
through their conjugation with the adjoining p system in the
2
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2012, 51, 1 – 5
These are not the final page numbers!