Journal of the American Chemical Society
Communication
sulfilimine formation. In contrast to a preliminary study relying
on high temperatures (130 °C, two examples),15 subsequent
6π- electrocyclization/ring-contraction/elimination16 of 7
proceeded spontaneously at 23 °C in an open flask to give
pyrrole 9 (Scheme 1C).
Employing Sharpless’ conditions for the synthesis of N-tosyl
sulfilimines (chloramine-T trihydrate, acetonitrile, 23 °C),17
we observed rapid conversion of 1,3-diene 5a to pyrrole 9a in
53% yield (Scheme 2, entry 1). The 2,5-dihydropyrrole 10 was
single run. Modifications of R1 (highlighted in red) allowed for
the implementation of electronically enriched arenes and a
thiophene to give 9b−d in constantly good yields (72−79%).
The presence of a strongly electron withdrawing substituent
such as a nitro group (9e) or a trifluoromethyl group (9f) was
well tolerated (63−64%). Different aryl halides were also
shown to effectively undergo pyrrole formation to deliver
chloride 9g, fluoride 9h, and bromide 9i in high yields between
69 and 78%. In addition, tertiary amide 9j and aldehyde 9k
were accessible from the reaction (59−65%). As shown for the
synthesis of the alkyl (R1 = Me, n-Bu)- and allyl-substituted
pyrroles 9l−n (52−76%), an aryl residue was not required at
the C3 position. Only alkyne 9o and pivalate 9p were obtained
in lower yields (28−30%). Lactone 9q (42%) was also
accessible, thus expanding the synthetic utility to annelated
ring systems. When the ester was changed to amides (R2,
highlighted in blue), the primary and secondary amides 12a,b
were isolated in 56 and 81% yields, respectively. The latter
bears the 3,4-substitution pattern as found in atorvastatin (2).
Additionally, the Weinreb amide 12c was synthesized in 33%
yield. Ketones also participated in the transformation and gave
the di- and trisubstituted pyrroles 13a−c in good yields (55−
77%). The presence of nitriles was also tolerated under the
reaction conditions but required the absence of m-CPBA
during the workup process. This allowed for the isolation of
pyrrole 14a in 51% yield (18% in the presence of m-CPBA).
Consequently, we were able to prepare pyrrole 14b (42%),
which was quantitively converted to the fungicide fludioxonil
(15, Pestanal)1c,18 through N-tosyl cleavage under basic
conditions (NaOH, MeOH). Application of O-mesitylenesul-
fonyl hydroxylamine (MSH) and sodium carbonate19 allowed
for the direct conversion of 1,3-diene 5a to the unprotected
pyrrole 16 (30%), which was produced in higher yields via
deprotection of 9a (Cs2CO3, MeOH, 84%). To conclude the
synthetic scope, we explored the productivity of other
chloramines to trigger the pyrrole formation of 5a.
Commercially available chloramine-B monohydrate allowed
for the construction of pyrrole 17a in 88% yield. When its p-
nitrophenyl (chloramine-N), p-methoxyphenyl (chloramine-P)
and methyl (chloramine-M) derivatives were applied, pyrroles
17b−d were also accessible in yields up to 75%.
By changing to sterically encumbered 1,3-dienes such as 18,
we were able to isolate the reactive sulfilimine 19 (61% yield,
step A) under the standard reaction conditions (Scheme 4A).
To our delight, thermal activation (toluene, reflux) allowed for
the smooth initiation of the subsequent cascade to deliver
pyrrole 20 in decent yield (76%, step B). When this two-step
protocol was applied, trisubstituted pyrrole 21 (78% and 49%)
and tetrasubstituted pyrrole 22 (61% and 99%) were formed.
In addition, trisubstituted pyrrole 23 was obtained in good
yields (62%), provided that benzonitrile was employed as the
solvent.20 As exemplified by 24, we found that the absence of
an electron-withdrawing group (EWG) also allows for the
isolation of its corresponding sulfilimines (99% yield, step A)
under the standard reaction conditions. After this, thermal
activation resulted in the formation of pyrrole 24 in
quantitative yield. It is worth noting that, when sulfilimine
25 was subjected to thermal conditions (111 °C), a complete
reaction was observed within 20 min. However, the main
product was identified as the 2,5-dihydropyrrole 26 (44%)
accompanied by small quantities of its cis-fused diastereomer
(not shown, 10%) and pyrrole 27 (10%). Resubjecting 26 to
refluxing toluene led to full conversion (28 h) to 27 in
a
Scheme 2. Optimization Studies
a
Legend: (1) yield determined by 1H NMR analysis using
nitromethane as internal standard; (2) isolated yield, 0.2 mmol
scale of 5a−c. Abbreviations: Ts = p-toluenesulfonyl, DMF = N,N-
dimethylformamide, HFIP = hexafluoroisopropyl alcohol, m-CPBA =
m-chloroperbenzoic acid.
isolated as the second product together with traces of
trisubstituted pyrrole 11, which might originate from 10 via
a competing oxidation pathway. Further screening revealed
slightly lower yields for the solvents N,N-dimethylformamide,
methanol, and water (32−49%, entries 2−4). In the presence
of 1 equiv of p-toluenesulfonic acid monohydrate (p-TsOH·
H2O, entry 5), the yield was increased to 70%. The use of
hexafluoroisopropyl alcohol (HFIP) as the cosolvent allowed
for the removal of p-TsOH·H2O and further improved the
yield of 9a to 84% (entry 6). The use of 1.5 equiv of
chloramine-T trihydrate or anhydrous chloramine-T (2 equiv)
led to decreased yields (41−65%, entries 7 and 8). Dichlor-
amine-T (TsNCl2) led to rapid consumption of the substrate,
but pyrrole formation was accompanied by decomposition to
give 9a in only 23% yield. Variation of the vinyl sulfide revealed
diene 5a (R = Me) to be superior to 5b (R = Et, 68%) and 5c
(R = Ph, 59%), delivering pyrrole 9a in an 83% isolated yield.
The addition of m-chloroperbenzoic acid (m-CPBA) after full
conversion of the starting material allowed for selective sulfur
oxidation of 11 and facilitated the isolation of pure 9a.
With our optimized conditions in hand, we investigated the
robustness and compatibility of the protocol for a panel of 1,3-
dienes (Scheme 3). The scalability was demonstrated by the
rapid synthesis of more than 1.5 g (78%) of pyrrole 9a in a
9003
J. Am. Chem. Soc. 2021, 143, 9002−9008