A new method for the acylation of pyrroles

Article abstract of DOI:10.1016/j.tetlet.2004.10.133

N-Tosylpyrroles can be very efficiently acylated by carboxylic acids in the presence of trifluoroacetic anhydride to give only the 2-acyl derivatives. N-Tosylpyrroles can be very efficiently converted into the corresponding 2-acylpyrroles by reaction with carboxylic acids and trifluoroacetic anhydride; little or none of the isomeric 3-acyl derivatives are formed.

Full text of DOI:10.1016/j.tetlet.2004.10.133

Tetrahedron
Letters
Tetrahedron Letters 45 (2004) 9573–9576
A new method for the acylation of pyrroles
*
Chuanjun Song, David W. Knight and Maria A. Whatton (ne e´ Fagan)
Cardiff School of Chemistry, Cardiff University, PO Box 912, Cardiff CF10 3TB, UK
Received 15 September 2004; accepted 22 October 2004
Available online 11 November 2004
Abstract—N-Tosylpyrroles can be very efficiently converted into the corresponding 2-acylpyrroles by reaction with carboxylic acids
and trifluoroacetic anhydride; little or none of the isomeric 3-acyl derivatives are formed.
ꢀ
2004 Elsevier Ltd. All rights reserved.
6
Because of their very high reactivity towards electrophi-
lic substitution and general sensitivity to acid-catalysed
polymerisation, there are few very efficient methods
available for the selective acylation of pyrroles 1; mix-
tures of 2- and 3-acylated product 2 and 3 are often ob-
tained (Scheme 1). Perhaps the best and most proven
method is by Vilsmeier reaction with phosphorus oxy-
chloride and N,N-dimethyl acetamide, which leads to
dation level. Less direct but nevertheless efficient alter-
natives, which do not appear to suffer from formation of
3-acyl isomers, include condensations of pyrrole with a
1,3-benzoxathiolium tetrafluoroborate followed by
mercury(II) oxide-induced hydrolysis or an N-methyl-
nitrilium salt, obtained by N-methylation of the
7
corresponding nitrile using Meerweinꢀs reagent, fol-
lowed by hydrolysis of the resulting imine. Clearly, an
8
1
respectable yields of 2-acetylpyrrole. The same method,
when applied to the general series of N,N-dimethylcarb-
oxamides, is also viable and delivers good yields (75–
obvious way to moderate the excessive reactivity of a
ꢁfreeꢀ pyrrole is to derivatise it by placing an electron
withdrawing groupon the nitrogen. In this res pe ct, sulf-
onyl derivatives have been the most widely studied.
Unfortunately, the problem of mixed product formation
(Scheme 1) has been found to persist in acylations of
8
amounts of the corresponding 3-acyl isomers or bis-acyl
0%) of 2-acylpyrroles, uncontaminated by significant
2
derivatives. This is certainly not the case with direct
acetylation with acetic anhydride, which gives mixtures
such derivatives (R = SO Ar), at least under Friedel–
2
3
of 2- and 3-acetylpyrrole in moderate overall yields.
A similar but even less efficient result has been obtained
Crafts conditions using aluminium trichloride as the cata-
9
lyst and various aroyl chlorides, although not when
the electrophile is an alkanoyl chloride and the catalyst
4
using triethyl orthoacetate and BF ÆOEt as catalyst.
3
2
1
0,11
12
The formation of both possible isomers (Scheme 1) also
plagues otherwise reasonably efficient methodology
based on the thermal rearrangement of N-acetylpyr-
is BF ÆOEt .
A more recent report has outlined a
general method for the selective 2-acylation of a range
3
2
of pyrrole derivatives (N-Boc; N-SO Ar) as well as of
2
5
roles or on the acylation of pyrryl Grignard reagents
with a variety of electrophiles at the carboxylic acid oxi-
pyrrole itself, by exposure to an acid chloride and zinc
powder in toluene. Yields are generally 80% or better
from these highly regioselective acylations. It was
against this background that we felt somewhat uncertain
as to the best method to use for the preparation of a ser-
ies of deactivated (i.e., N-derivatised; preferably N-tosyl)
O
R1
1
R CO.X
2
-acylpyrroles, which were required for a separate
R1
+
13
study. Clearly, the use of anhydrides would be rather
wasteful, especially when the precursor acid is relatively
precious. However, we were intrigued by the contrast
between the foregoing results and those reported by
N
R
N
R
N
R
O
1
2
3
Scheme 1.
14
Kakushima et al. These authors report that either an
acid chloride or the corresponding anhydride, in combin-
ation with boron trifluoride etherate, will acylate an
N-tosylpyrrole slowly and largely in the 2-position
*
Corresponding author. Tel./fax: +44 2920 874 210; e-mail:
knightdw@cf.ac.uk
0
040-4039/$ - see front matter ꢀ 2004 Elsevier Ltd. All rights reserved.
doi:10.1016/j.tetlet.2004.10.133

9574
C. Song et al. / Tetrahedron Letters 45 (2004) 9573–9576
(42–83% yields) with 5–15% of the 3-acyl derivative also
being formed. By contrast, the use of the stronger Lewis
rather than the derived acid chlorides, anhydrides or
amides.
acid, AlCl , was reported to give only 3-acyl derivatives
3
(
generally >95%) quite rapidly at ambient temperature.
In our hands, the former method, when applied to
-methyl-N-tosylpyrrole using BF ÆOEt , gave only
Initial experiments established that 2-methyl-N-tosyl-
pyrrole could indeed be acetylated using a large excess
(>10equiv) of acetic acid in a 1:1 mixture of TFAA
and dichloromethane at ambient temperature. Whilst
none of the corresponding 3-acetyl isomer was detected,
we were a little surprised to observe that a by-product
was rather the 2,4-diacetyl derivative. The large excess
of reagents is evidently sufficient to overcome the deacti-
vating effect of the first acetyl group. At least, a 2,4-
disubstitution pattern is what would be expected from
a second acetylation. Subsequent optimisation experi-
ments defined a more efficient procedure, in which acetyl-
ation was achieved using 4equiv or less of the carboxylic
acid. Under these conditions, no 2,4-diacyl products
were detected. The results using this method with a
range of N-tosylpyrroles and carboxylic acids are col-
2
3
2
moderate yields of the desired 2-acyl-N-tosyl-5-methyl-
pyrroles, together with substantial amounts of the 3-acyl
isomer from various alkanoyl chlorides. Attempts to
secure useful yields using other Lewis acids (e.g., ZnCl2,
SnCl , TiCl ) were also not successful. However, we
noted that these authors had commented that 2-acetyl-
N-tosylpyrrole could also be obtained, but in excellent
yield, using a ꢁlarge excessꢀof acetic acid and trifluoro-
acetic anhydride. The fact that a somewhat related
mixture of trifluoroacetic anhydride (TFAA) and
phosphoric acid had been used to induce thiophene acyl-
4
4
1
5
ation led us to investigate this method in more detail.
We were also attracted by the additional favourable
prospect of being able to use carboxylic acids directly,
1
6
lected in Table 1.
Table 1. Acylation of N-tosylpyrroles by carboxylic acids and trifluoroacetic acid
1
R CO2H
R1
R
N
R
N
TFAA, CH2Cl2
Ts
Ts
O
16
16
Entry; acid [equiv] conditions Yield (%) Product
Entry; acid [equiv] conditions Yield (%) Product
1
; R = H.
HOAc [4.0];
0ꢁC, 30h,
CH Cl
94
90
72
73
84
84
83
8; R = Me.
Me CH(CH
20ꢁC, 4h,
CH Cl
87
96
74
57
N
2
2
)
2
CO
2
H [1.5];
N
Ts
Ts
2
O
O
O
2
2
.
2
2
.
2
; R = H.
BuCO H [4.0];
0ꢁC, 16h then reflux, 7h
CH Cl
9; R = Me.
PhCO H [4.0];
20ꢁC, 42h,
CH Cl
Bu
2
2
N
Ts
N
Ts
2
O
O
2
2
.
2
2
.
3
; R = H.
t-BuCO H [4.0];
reflux, 119h,
Cl(CH Cl.
10; R = Ph.
HOAc [2.0];
20ꢁC, 43h,
N
2
Ts
N
Ts
O
O
2
)
2
2 2
CH Cl .
CO2Me
)7
4; R = H.
MeO
11; R = m-O
HOAc [2.0];
reflux, 48h,
2 6 4
NC H
N
Ts
2
C(CH
2
)
8
CO
2
H [1.5]
(
reflux, 48h,
CH Cl
N
Ts
O
O
NO2
2
2
.
2 2
Cl(CH ) Cl.
5; R = H.
Ph
PhCO
2
H [2.0];
a
N
Ts
12; R = p-MeOC H .
40
6
4
reflux, 70h,
Cl(CH Cl.
HOAc [2.0];
0ꢁC, 35h,
CH Cl .
N
2 2
)
Ts
2
O
MeO
OMe
2
2
6
; R = H.
p-MeOC
0ꢁC, 48h,
CH Cl
6
4 2
H CO H [2.0];
2
N
Ts
O
2
2
.
7
; R = Me.
HOAc [3.6];
ꢁC, 2h,
CH Cl
13; R = 2-naphthyl.
HOAc [2.0];
79
N
Ts
O
N
Ts
0
20ꢁC, 95h,
CH Cl .
2 2
O
2
2
.
a
The 5-trifluoroacetyl derivative was also isolated in 11% yield.

C. Song et al. / Tetrahedron Letters 45 (2004) 9573–9576
9575
As expected (entries 1–6), N-tosylpyrrole itself proved to
be somewhat less reactive than its 2-methyl derivative
used in the initial optimisations. While acetic acid re-
acted smoothly if slowly in dichloromethane at ambient
temperature, in order to secure a similarly excellent yield
from pentanoic acid, it was necessary to reflux the reac-
tion mixture (entry 2). The much more hindered and
electron-rich pivalic acid required prolonged reflux in
dichloroethane before a respectable yield (72%; entry
glutaric acid (0.5equiv). Further oxidation then gave a
moderate but unoptimised return of ca. 40% of the pyr-
one 9 (Scheme 4). Evidently, a second acylation of N-
tosylpyrrole does not occur as fast as intramolecular
cyclisation of the remaining carboxylic acid grouponto
the new ketone function, followed by dehydration. Stud-
ies of the chemistry of these potentially useful structures
are underway.
3
) of the corresponding acylpyrrole was obtained. Benz-
Thirdly, by reduction of the new carbonyl group, this
present method could represent part of an efficient meth-
od for overall pyrrole a-alkylation. We initially tried
various classical methods for selectively reducing the ke-
tone groupin the keto-ester shown in entry 4 of Table 1
(Scheme 5). Unfortunately, both Clemmensen and vari-
ous versions of the Wolff-Kishner reduction methods
failed. However, a more recent method featuring the
borane-t-butylamine complex as reductant, assisted by
aluminium trichloride, did produce the desired prod-
uct 10 but was somewhat impractical in its original ver-
sion, due to the claimed requirement for a considerable
excess of the reagents. Fortunately, an optimisation
study showed this to be unnecessary and delivered
cleanly an 85% yield of the ester 10 as shown.
oic acid (entry 5) also only reacted slowly under these
more vigorous conditions; despite the lengthy reaction
times however, no 3-acylated derivatives were evident
in H NMR spectra of the crude products. In contrast,
4
temperature (entry 6). That the presence of an a-methyl
substituent provides significant additional activation is
shown by the relatively milder conditions required to
acylate 2-methyl-N-tosylpyrrole (entries 7–9). Acyl-
1
-methoxybenzoic acid reacted smoothly at ambient
1
8
1
7
ations of representative 2-aryl-N-tosylpyrroles (entries
0–13) were also similarly selective for the a-pyrryl posi-
1
tion and best carried out at ambient temperature for ex-
tended periods. Some limitations which were discovered
were perhaps not too surprising: direct formylation
using formic acid failed to deliver any pyrrole-2-carbox-
aldehyde derivatives and 2- and 3-furoic acids also failed
to act as acylating agents. Under the present conditions,
it is also proved impossible to acetylate the alkenylpyr-
role 4; polymers and a variety of decomposition prod-
ucts were formed and not the ketone 5 (Scheme 2).
Finally, ꢁfreeꢀ pyrroles can also be obtained by this
route, followed by subsequent N-detosylation. Although
we have not examined any alternatives, simply exposing
the 2-acyl-N-tosylpyrroles to potassium hydroxide in
aqueous methanol (3h, 60ꢁC) delivers >80% yield of
the NH-pyrroles, also uncontaminated by any migration
1
9
In addition to these limitations there are, however, three
additional and positive caveats to this work. Firstly,
such acylations can be conducted in an intramolecular
fashion. To demonstrate this, we prepared the substi-
tuted sulfonamide 6 from 2-methoxycarbonylbenzene-
sulfonyl chloride and the sodium salt of pyrrole
followed by ester saponification. We were delighted to
observe that, albeit under relatively vigorous conditions,
this was smoothly converted into the hoped-for product
products.
In summary, it would appear that the combination of a
carboxylic acid and trifluoroacetic anhydride is a simple
and practical method for the selective 2-acylation of N-
tosylpyrroles. We assume that the mechanism involves
mixed anhydride formation between TFAA and the
7, in 78% isolated yield (Scheme 3). To the best of our
knowledge, compound 7 represents the parent of a novel
heterocyclic ring system.
HO2C
CO2H
N
TFAA, CH Cl
N
2
2
Ts
Ts
4
8 h, 20˚C
then reflux, 24 h.
9%
O
O
Secondly, we found that the dihydropyrone 8 was ob-
tained when 2-methyl-N-tosylpyrrole was exposed to
8
9
O
O
5
HOAc
DDQ
Bu
N
Bu
N
TFAA, CH2Cl2
Ts
Ts
O
toluene, reflux
7 h, ~ 40%
N
20˚C
1
Ts
4
5
Scheme 2.
Scheme 4.
O
S
HO2C
TFAA
( )6 CO Me
2
( )6 CO Me
2
t
2
eq. BuNH2.BH3
N
S
O O
N
Cl(CH2)2Cl
reflux, 50 h
N
1 eq. AlCl , CH Cl ,
N
3
2
2
Ts
O
Ts
O O
20˚C, 3 h
78%
85%
6
7
[entry 4]
10
Scheme 3.
Scheme 5.

9576
C. Song et al. / Tetrahedron Letters 45 (2004) 9573–9576
reacting carboxylic acid. It is perhaps surprising there-
fore that only in the example involving acetylation of
2-(4-methoxyphenyl)-N-tosylpyrrole (entry 12) was an
isolable amount of a 2-trifluoroacetyl derivative formed.
delivers moderate yields (37–45%) of 2-acylpyrroles:
Hasan, I.; Marinelli, E. R.; Lin, L.-C. C.; Fowler, F. W.;
Levy, A. B. J. Org. Chem. 1981, 46, 157.
1
1
1
2. Yadav, J. S.; Reddy, B. V. S.; Kondaji, G.; Rao, R. S.;
Kumar, S. P. Tetrahedron Lett. 2002, 43, 8133.
3. Cf. Fagan, M. A.; Knight, D. W. Tetrahedron Lett. 1999,
4
0, 6117.
Acknowledgements
4. Kakushima, M.; Hamel, O.; Frenette, R.; Rojach, J. J.
Org. Chem. 1983, 48, 3214.
We thank the EPSRC Mass Spectrometry Service, Uni-
versity College, Swansea for the provision of high reso-
lution mass spectrometric data and the EPSRC and
Cardiff University for financial support and the provi-
sion of high field NMR facilities.
15. Galli, C.; Illuminate, G.; Mandoli, L. J. Org. Chem. 1980,
45, 311.
1
6. A typical procedure is as follows: 5-Methyl-2-(4-methyl-
pentanoyl)-N-(4-toluenesulfonyl)-pyrrole. To stirred
solution of 2-methyl-N-tosylpyrrole (7.173g, 33mmol),
a
trifluoroacetic anhydride (50ml) and dichloromethane
(
75ml) maintained at 0ꢁC was added 4-methylpentanoic
acid (6.2ml, 49mmol). The resulting solution was stirred
without further cooling until TLC analysis showed com-
plete conversion (ca. 4h) and then the volatiles were
removed by rotary evaporation. The residue was stirred
briefly with dichloromethane (30ml) and 10% aqueous
sodium carbonate (50ml) then the organic layer was
separated and the aqueous layer extracted with dichloro-
methane (2 · 20ml). The combined organic solutions were
References and notes
1
. Alia, I.; Smith, G. F. J. Chem. Soc. 1954, 3852; Silverstein,
R. M.; Ryskiewiez, E. E.; Willard, C.; Koehler, R. C. J.
Org. Chem. 1955, 20, 668; Anthony, W. C. J. Org. Chem.
960, 25, 2048; Cooper, G. H. J. Org. Chem. 1971, 36,
2897; Kurada, Y.; Murase, H.; Suzuki, Y.; Ogoshi, H.
Tetrahedron Lett. 1989, 30, 411.
1
then washed with brine (50ml), dried (MgSO ), filtered
4
2
. Garrido, D. O. A.; Buldain, G.; Frydman, B. J. Org.
Chem. 1984, 49, 2619.
. Ciamician, G. L.; Silber, P. Chem. Ber. 1885, 18, 881;
Anderson, A. G., Jr.; Exner, M. M. J. Org. Chem. 1977,
and evaporated. Column chromatography of the residue
(SiO
acylpyrrole (Table 1, entry 8) (9.56g, 87%) as a yellow oil,
2
, 10% EtOAc–40–60ꢁ petrol) then separated the
3
which showed R 0.75 (80% Et O–40–60ꢁ petrol), mmax/
f
2
ꢀ
1
42, 3952.
cm (film) 1679, 1597, 1488, 1366, 1177, 1105 and 812, d
H
4. Pindur, U.; Flo, C. J. Heterocycl. Chem. 1989, 26, 1563.
5. Patterson, J. M.; Soadigo, S. J. Org. Chem. 1968, 33, 2057.
6. Chelintzev, V.; Terentiev, A. Chem. Ber. 1914, 47, 2647
(400MHz, CDCl
0
3
) 0.82 (6H, d, J 6.3Hz, 2 · Me), 1.41–
1.58 (3H, m, 4 -H and 3-CH ), 2.34 (3H, s, Ar-Me), 2.46
0
2
(3H, s, 5-Me), 2.58–2.69 (2H, m, 2 -CH ), 5.91 (1H, d, J
2
(
(
5
RCO
RCO.Cl); Baltazzi, E.; Krimen, L. I. Chem. Rev. 1963, 63,
11; Nicolaou, K. C.; Claremon, D. A.; Papahatjis, D. P.
2
Et); Beon, G. P. J. Heterocycl. Chem. 1965, 2, 473
3.5Hz, 4-H), 6.69 (1H, d, J 3.5Hz, 3-H), 7.25 (2H, d, J
(100MHz, CDCl
8.2Hz) and 7.90 (2H, d, J 8.2Hz), d
C
3
)
0
16.3 (5-Me), 2.21 (Ar-Me), 22.8 (2 · Me), 28.2 (4 -CH),
0
0
Tetrahedron Lett. 1981, 22, 4647 (RCOÆSpy); Kozikowski,
A. P.; Ames, A. J. Am. Chem. Soc. 1980, 102, 860
(
. Barbero, M.; Cadamuro, S.; Degani, I.; Fochi, R.; Gatti,
A.; Regondi, V. J. Org. Chem. 1988, 53, 2245.
. Eyley, S. C.; Giles, R. G.; Heany, H. Tetrahedron Lett.
1
. Lainton, J. A. H.; Huffman, J. W.; Martin, B. R.;
Crompton, D. R. Tetrahedron Lett. 1995, 36, 1401;
Kimbaris, A.; Varvounis, G. Tetrahedron 2000, 56, 9675.
34.2 (3 -CH ), 39.3 (2 -CH ), 112.6 (4-CH), 121.1 (3-CH),
128.1 (2 · ArCH), 129.9 (2 · ArCH), 136.7, 137.3, 141.0,
2
2
1
RCOÆSeR ).
145.1 (all ArC) and 192.0 (C@O), m/z (APCI) 334
+
+
7
8
9
(M + H 100%) (Found: M + H , 334.1474. Calcd for
S, 334.1477).
18 3
C H24NO
17. The 2-aryl substrates used in entries 10–13 were most
conveniently prepared by Suzuki couplings between 2-
bromo-N-tosylpyrrole and the corresponding arylboronic
acid: Knight, L. W.; Huffman, J. W.; Isherwood, M. L.
Synlett 2003, 1993.
985, 26, 4649.
1
1
0. Muratake, H.; Natsume, M. Tetrahedron Lett. 1987, 28,
265.
1. A completely different strategy, featuring lithiation of N-
18. Ketcha, D. M.; Carpenter, K. P.; Atkinson, S. T.;
Rajagopalan, H. R. Synth. Commun. 1990, 20, 1647.
19. Full analytical and spectroscopic data have been obtained
which support all of the structures reported herein.
2
Boc-pyrrole followed by acylation with an acid chloride,