800
B. S. Gourlay et al. / Tetrahedron Letters 47 (2006) 799–801
products the yields were not consistently high due to
decomposition by contact with acid or heat or both.
These methods are not sufficiently robust for the prepa-
ration of the types of compounds that we require as
starting points for an asymmetric synthesis. With this
as a major stumbling block in our synthetic path we
embarked on an investigation of the requirements to
effect the Clauson-Kaas reaction. While acid or heat
are required to promote hydrolysis of 2,5-dimethoxy-
tetrahydrofuran to the more reactive dialdehyde,
which exists as the dihydroxy furan 4 under the aqueous
conditions, we reasoned that these conditions may not
be necessary for the condensation of 4 with the amine
to yield the pyrrole6 (i.e., Paal-Knorr reaction). There-
fore, we proposed to decouple these two processes for
our current synthesis of N-substituted pyrroles such that
exposure of the pyrroles to heat or strong acid could be
avoided altogether.
crude products were usually obtained as colourless oils,
which contrasted with the dark brown to black tarry
material (indicative of decomposition of the pyrrole
nucleus) often obtained by other literature methods.
The products could then be purified simply by passing
through a short plug of silica to yield the desired com-
pound in a pure state. We have applied this procedure
to both achiral and chiral substrates including amines
as the free base or the hydrochloride salt. The results
of these experiments are shown in Table 1.
In general, the yields of isolated products were excellent
and consistently higher than those cited most recently
(59–81% for amino esters).5 The importance of generat-
ing 4 as an intermediate is highlighted by the fact that
the yield of the product in entry 1 was only ꢀ10%
when the initial refluxing of 3 in water was omitted.
The pH of the reaction was also shown to be important
as a 31% yield of pyrrole product was obtained from
R-(À)-phenylethylamine when the amine was added to
4 in the absence of acetate buffer compared to a 91%
yield obtained under buffered reaction conditions (entry
4). Scale up of the procedure was readily achieved
with the product from L-alanine methyl ester being syn-
thesised on a multigram scale without decline in yield
or resulting difficulty in isolation of the pure product.
The mild reaction conditions result in no detectable
racemisation of the stereogenic centre of any of the
chiral substrates under the conditions with eeÕs of all
substrates being >99% as determined by chiral gas chro-
matography. Previously, most attempts to increase the
product yield led to compromised eeÕs due to epimerisa-
tion under the reaction conditions that were employed.
For example, the recent report of the synthesis of pyr-
roles from the methyl esters of alanine and norvaline
gave high eeÕs; however, the yields were relatively mod-
erate (ꢀ70%).8 In addition, the mild reaction conditions
are highlighted by the fact that sensitive products, which
are prone to elimination reactions such as the pyrroles
obtained from serine, b-alanine and ( )-methyl 4-ami-
no-3-hydroxybutyrate were isolated in high yield.9 In
our hands, the latter product could not be isolated under
the conventional Clauson-Kaas conditions due to rapid
decomposition and formation of black tarry material.
Similarly, the serine derivative (entry 3) is prone to
dehydration under acidic conditions and was reported
in only a 19% yield employing standard conditions.10
The hydrolysis of 3 is typically performed using mineral
acids such as hydrochloric or sulfuric acid.7 As we
wished to avoid the use of strong acids, we investigated
the use of aqueous acetic acid. The use of acetic acid
would also mean that after hydrolysis sodium acetate
could be added to generate a buffer of ꢀpH 5 for the pyr-
1
role formation. H NMR studies of the hydrolysis of 3
were undertaken and showed that heating a solution of
the furan in D2O with 1 equiv of acetic acid resulted in
the complete hydrolysis in 1 h and clearly demonstrated
that strongly acidic conditions were not required. Sur-
prisingly, it was also observed that simple heating in
D2O without acid resulted in the hydrolysis of the acetal
within 2 h indicating that acid was not necessary for the
hydrolysis step (Scheme 3). In the latter case, addition of
ammonia to the NMR tube resulted in only a trace of
pyrrole. In contrast, when ammonium chloride and
sodium acetate were added to the solution of 4 in D2O,
formation of pyrrole was observed. Therefore, while
acid is not required for the activation or hydrolysis of
2,5-dimethoxytetrahydrofuran, a solution of pH <7 is
required to promote formation of the pyrrole. These
observations were then considered in designing the syn-
thetic method.
The synthetic method was then carried out by heating
2,5-dimethoxytetrahydrofuran in water at reflux for
2 h under nitrogen to form dihydroxytetrahydrofuran
4. The solution was cooled to room temperature, dichloro-
methane was added followed by the amine and 1 equiv
of both acetic acid and sodium acetate to form a buffer
of ꢀpH 5. After stirring at room temperature overnight,
a high yield of the pyrrole was observed. Both free
amines and hydrochloride salts can be used with this
procedure; however, for hydrochloride salts of amines
no acetic acid was required and 2 equiv of sodium
acetate were added to maintain a buffer of ꢀpH 5. The
In summary, we have shown that a one-pot two-step
method involving hydrolysis of 2,5-dimethoxytetrahydro-
furan followed by the addition of the amine in a buffered
solution gives pyrroles cleanly, in excellent and repro-
ducible yield. The method has allowed the preparation
of acid- or heat-sensitive pyrroles for the first time in
high yield, high ee and high purity. Couple these points
with the fact that the method can readily be carried out
on a multigram scale with easy purification and the pro-
cedure has the potential to be a widely adopted method
for the synthesis of pyrrole derivatives.
i or ii
H3CO
OCH3
HO
OH
O
O
3
4
Representative procedure: 2,5-Dimethoxytetrahydro-
furan (0.200 mL, 1.544 mmol) was added to a stirred
solution of water (2.0 mL) and the solution was refluxed
Scheme 3. Reagents and conditions: (i) D2O, 1 equiv AcOH, reflux
1 h; (ii) D2O, reflux 2 h.