K. R. Law, C. S. P. McErlean / Tetrahedron Letters 57 (2016) 3113–3116
3115
of the reaction may in fact be the samarium(II) reduction of the
unsaturated system to give 16 or 17 (Scheme 4),18 which could
undergo radical or anionic addition to the pendant aldehyde.
Protonation of 17 prior to ring-closure would give the samarium
enolate 18, which can undergo elimination. If this proved to be
the case, inverse addition would exacerbate the problem, and N-
acylpyrroles would not be suitable substrates for SmI2-mediated
cyclizations.
Pleasingly, the slow addition of aldehyde 11 into the SmI2 mix-
ture and competitive protonation with a tenfold excess of MeOH,
gave the desired compound 12 in excellent yields as a single
stereoisomer (Table 1, entry 8). The yield of the transformation
could be further increased at higher dilution (entry 9). NMR anal-
ysis of the isolated compound 12 demonstrated key through-space
interactions that are diagnostic of the trans, syn, trans stereochem-
ical arrangement of the newly formed tetrahydropyran (Fig. 2).
Additionally, we hypothesized that if the mechanism depicted
in Scheme 4 was operational, then the aldehyde unit of compounds
16 and 17 could be replaced with a functional group that has sim-
ilar reactivity toward nucleophiles, i.e., with an N-acylpyrrole, and
the SmI2-mediated cyclization would still operate. The redox
potentials of SmI2-additive mixtures vary from À0.89 to À1.79 V
depending on the identity of the additive.19–23 We therefore mea-
sured the reduction potentials of the N-acylpyrroles 9 and 19–22
shown in Figure 3. This confirmed that SmI2–MeOH mixtures
Scheme 4. Plausible pathways to compounds 12, 14 and 15.
would have sufficient reducing power to react with
a,b-unsatu-
rated N-acylpyrroles (9, 19–22), while saturated N-acylpyrroles
(21 and 22) would require a more powerful reductant.
As shown in Scheme 5, the product of the previous SmI2-medi-
ated ring-closure 12 underwent Michael addition onto N-propi-
olpyrrole 9 to give compound 23 in high yields. This molecule
contained both an
a,b-unsaturated N-acylpyrrole and a saturated
N-acylpyrrole. Chemoselective reduction of the unsaturated N-
acylpyrrole would produce 24, which could undergo cyclization
onto the pendant saturated N-acylpyrrole.24
Figure 2. Stereochemistry of cyclized product 12.
When compound 21 was subjected to the action of SmI2 and
MeOH, no ring-closed product was observed. Isolation of the elim-
inated product 12 confirmed that the
a,b-unsaturated N-acylpyr-
role unit was reduced under the reaction conditions, but did not
undergo intramolecular radical or anionic addition to the pendant
N-acylpyrrole. This outcome is consistent with the SmI2-mediated
carbonyl-ene cyclization onto
proceeding by initial reduction of the aldehyde unit.
In summary, we have shown that an ,b-unsaturated N-acyl-
a,b-unsaturated N-acylpyrrole 11
a
pyrrole is a suitable substrate for SmI2-mediated reductive cycliza-
tion to give a trans, syn, trans-fused tetrahydropyran. In contrast to
previously reported reductive cyclizations using unsaturated
esters and ketones, the intermediate samarium enolate 13 readily
engages in aldol-type reactions, so inverse addition is mandatory.
The application of this transformation to the construction of poly-
cyclic ethers is underway in our laboratory.
Figure 3. Reduction potentials of N-acylpyrroles 9 and 19–22.
Scheme 5. Attempted ring-closure of compound 23.