Organic Letters
Letter
that sense, leading to the conclusion that it was best to excise
the serine portion altogether. Compound 14 was thus advanced
to acid 16, in preparation for an eliminative Curtius−Schmidt
Scheme 2. Oxidative Cyclization of 9 and Stereoselective
Formation of 11
rearrangement a
̀
la Bermejo−Gonzalez (16 → 17).21 However,
the reaction of 16 with diphenyl phosphorazidate (DPPA)22
under the reported conditions (toluene, Et3N, 90 °C)21
afforded only some of the desired 17, the major product
being isocyanate 19 (Scheme 4). VanNieuwenhze et al., also
observed low yields of enamide under such conditions.23
Scheme 4. Likely Mechanism for the Formation of 17
securing the correct (R)-configuration of the spiro carbon
(confirmed by X-ray diffractometry, Figure 2).19 Stereorelay
It should be noted that the DPPA protocol is more attractive
than alternative methods that effect the same transformation, in
that it delivers the enamide in one step and avoids the use of
transition- or heavy metals. Indeed, such alternatives require
two steps from the acid and involve oxidative decarboxylation
by electrochemical means,24 by reaction with Pb(IV)25 or
I(III)26 reagents, or by photolysis of N-nitroso derivatives,27 as
well as Ni- or Pd-mediated dehydrocarbonylation of
thioesters.28,29 We thus sought to improve the published
DPPA procedure.
In accord with VanNieuwenhze,23 operation in dioxane,
instead of toluene, afforded a slightly greater amount of 17,
although the major product remained 19. The addition of DBU
to a mixture of 19 and 17 thus obtained, and continued
refluxing, induced conversion of 19 into 17 (reaction
Figure 2. X-ray structure of lactam 11 (·H2O).
from the serine fragment to the spirocenter thus occurred with
perfect fidelity. The stereochemical outcome of this step may
be understood by recognizing that groups adjacent to the N
atom in the N-acylmorpholine unit of 11 favor an axial position,
so as to minimize nonbonding interactions with the N-acyl
group.20 Assuming that cyclization occurs via a chairlike
transition state, then 11 forms via conformer 12, where the
COOMe (depicted in 12 and 13 as E) is pseudoaxial in the
developing morpholine. The diastereomer of 11 would result
via 13, wherein the pseudoequatorial COOMe is compressed
against the N-acyl group.
Catalytic reduction of 11 (Scheme 3) returned 14 and
denied the molecule any opportunity for loss of configuration at
the spiro center. In principle, the five-membered ring of the
target alkaloids could be created from two of the serine carbons
in 14. However, numerous difficulties arose during attempts in
19b
1
monitored by H NMR).
Thus, enamide 17 forms from
19. It is unlikely that isocyanate ion acts as a leaving group in
E2 reactions.30 We presume that thermal activation of 19
reversibly forms acyliminium ion 20, deprotonation of which
then gives 17. Notice that the axial NCO group can depart
with assistance from the lactam N atom. Furthermore, the use
of more polar dioxane in lieu of toluene is likely to favor this
dissociative step. On the basis of the foregoing, acid 16 was
elaborated to 17 in 59% yield by reaction with DDPA and Et3N
in dioxane (rt to reflux, 2 h), followed by the addition of DBU
and further refluxing for 2 h.
Upjohn31 dihydroxylation of 17 produced lactam 18
(Scheme 5), probably through release of the etheno bridge as
glyoxal.32 The amide carbonyl in 18 should facilitate access to
Scheme 3. Preparation of Enamide 17
Scheme 5. Preparation of Ketones 22 and 24
B
Org. Lett. XXXX, XXX, XXX−XXX