We also wondered if 3,6-disubstituted 1,2,4-triazine (17)
would react with enolate 2 to give a cyclopenta[b]pyrrole
(e.g., 18; pathway A, Scheme 3) or an azocinone (e.g., 19;
Intrigued by the unexpected formation of cyclopenta[b]-
pyrrole analogues 16a and 16b (Scheme 2), we set out to
examine the possible mechanisms by which these products
are formed. The first question regards the role of pyrrolidine.
One possibility is that this amine reacts with cyclobutanone
to form the corresponding enamine (2′ in Scheme 2). This
species could then participate as an electron-rich dienophile
in a cycloaddition with triazine 9 to form intermediate 10
and, by loss of N2, 11 (X ) N-pyrrolidinyl).10 This scenario
is supported by Taylor’s observation of 22 in a closely related
reaction.2b Alternatively, pyrrolidine may simply function
as a base and form small quantities of enolate 2 (protonated
pyrrolidine is more acidic than cyclobutanone by ∼10 pKa
units), which through reaction with 9 would give 11 (X )
O-). Previously 2, generated using KOH/methanol, was
shown to react in a similar manner with a 1,2,4,5-tetra-
zene.4-6 In fact, employing KOH/methanol, while not
effective in preparative reactions, does result in the reaction
of cyclobutanone with 9 to give 12 (X ) OH) and/or 15 as
evidenced by LC/MS of the crude reaction mixture. However,
no reaction was observed when DBU was used as base.
Scheme 3. Reactions of 3,6-Disubstituted 1,2,4-Triazines
pathway B, Scheme 3). Analysis of the reaction mixture from
1,2,4-triazine 17a by LC/MS revealed that unreacted starting
material was a major component in the mixture, accompanied
by a trace amount of what appeared to be an addition product
according to its molecular ion ([M + H+] ) 304). However,
when a 1,2,4-triazine with an electron-withdrawing 2-pyridyl
at the C3 position (17b) was reacted with cyclobutanone
enolate 2, a mixture of tautomeric adducts 20 and 20′ resulted
(pathway C, Scheme 3). The tautomeric nature of these
addition products is evidenced by two sets of doublets (Ha,
5.43 and 5.35 ppm) and two sets of broad singlets (Hb, 10.21
Whether intermediate 11 is formed with X ) O- or X )
N-pyrrolidinyl, there are at least two possible paths by which
it can be converted to 15. As depicted in Scheme 2, 11 may
undergo a substituent-accelerated six-electron, electrocyclic
ring-opening reaction to form the eight-membered system
12,5 which could then contract into 14. Alternatively, a direct
1,2-alkyl shift may occur, as predicted previously for related
systems,5 to produce 13, a tautomer of 15. In order to gain
additional insight into these possibilities, density functional
theory calculations (B3LYP/6-31+G(d,p); see Supporting
Information for details) were employed to probe the key steps
outlined in Scheme 2. As shown in Figure 2 (see Supporting
Information for additional structures), transition state struc-
tures corresponding to the 11 f 12 f 14 pathway for species
derived from both 2 and 2′ were located. However, transition
state structures for direct 1,2-alkyl shifts for either the
enamine or enolate systems could not be located.11 These
observations lead us to conclude that the reaction likely
occurs via a stepwise ring-opening/ ring-closing sequence.12
Our calculations suggest that the enolate version of 12 would
face lower barriers to form 14 than would the enamine
version (Figure 2). On balance, although neither the enolate
nor enamine pathways can be definitively ruled out, we favor
the enolate mechanism.
1
and 10.06 ppm) in the H NMR of these isolated products
as well as their co-eluting LC/MS peak with [M + H+] )
305. Indeed, nucleophilic addition to 17b is precedented.9
An analogous addition product was observed upon reaction
of thietanone 7 with triazine 17b in the presence of KOH
(Scheme 3). Adduct 21 was fully charaterized by 1H NMR,
13C NMR, LC/MS, and X-ray crystallography (see Support-
ing Information). When LHMDS was used as base instead
of KOH, the yield of 21 increased from 36% to 50%;
presumably because the non-nucleophilic LHMDS is less
likely to add to 17b.
(8) After the relative amount of triazine starting material no longer
decreased as indicated by LC/MS, the reaction mixture was cooled to room
temperature and concentrated under reduced pressure. The crude product
was redissolved in ethanol and mixed with hydrazine or O-Bn hydroxyamine
hydrochloride in a 90% ethanol/10% water solution. This mixture was
refluxed for 2 h, the solvent was then removed, and the crude product was
dried in DCM over Na2SO4, followed by chromatographic purification (silica
gel with gradient hexane/ethyl acetate elution).
(10) On the basis of calculations on a related system, it is possible that
10 is not a true minimum on the potential energy surface for this reaction.5
(11) Note here that additional calculations (see Supporting Information)
revealed that structure 15 (X ) O) is at least 20 kcal/mol lower in energy
than neutral tautomers of structure 12.
(9) Konno, S.; Ohba, S.; Sagi, M.; Yamanka, M. Chem. Pharm. Bull.
1987, 35, 1378.
(12) See Supporting Information for discussion related to this same
possibility for the tetrazine-derived systems studied in ref 5.
166
Org. Lett., Vol. 12, No. 1, 2010