Isolation of 5-Hydroxy-γ-lactams from a Classical 2-Aminopyrrole Synthesis

Article abstract of DOI:10.1002/jhet.3156

5-hydroxy-γ-lactams have been isolated as major byproducts from a classical 2-aminopyrrole synthesis involving condensation of an in situ prepared α-aminoketone with methyl cyanoacetate. The classical 2-aminopyrrole was obtained in very low yield, or not

Full text of DOI:10.1002/jhet.3156

Month 2018
Isolation of 5-Hydroxy-γ-lactams from a Classical 2-Aminopyrrole
Synthesis
*
Nisaraporn Suthiwangcharoen, Andrew C. Bean, Minna Hassan, Cheryl K. Eidell, and Chad E. Stephens
Department of Chemistry and Physics, Augusta University, Augusta, GA 30904, USA
*
E-mail: cstephe7@augusta.edu
Received July 27, 2017
DOI 10.1002/jhet.3156
Published online 00 Month 2018 in Wiley Online Library (wileyonlinelibrary.com).
Dedicated to Dr. J. Walter Sowell (mentor to CES) on occasion of his retirement from academia.
5-hydroxy-γ-lactams have been isolated as major byproducts from a classical 2-aminopyrrole synthesis
involving condensation of an in situ prepared α-aminoketone with methyl cyanoacetate. The classical 2-
aminopyrrole was obtained in very low yield, or not at all. One 5-hydroxy-γ-lactam was dehydrated to the
known 5-methylene-γ-lactam in good yield using thionyl chloride.
J. Heterocyclic Chem., 00, 00 (2018).
INTRODUCTION
formation of this previously unreported byproduct helps
explain why yields for these 2-aminopyrroles are relatively
low compared with yields for the analogous pyrroles
containing tert-butyl esters or other functional groups.
In 1975, Roth and Eger first described the now classical
synthesis of N-1-substituted-2-aminopyrroles containing a
cyano group at the 3-position [1]. Later, Sowell and
coworkers extended this reaction to synthesis of similar
2-aminopyrroles with esters [2] or sulfones [3] at the
3-position. This multicomponent reaction typically
involves condensation of an in situ formed α-aminoketone
(prepared by reaction of an α-hydroxyketone with a
primary amine) with an activated CH2CN compound, such
as malononitrile, an alkyl cyanoacetate, or a (cyanomethyl)
sulfonyl derivative. Although the 4,5-dimethylpyrrole
analogues are most commonly prepared (using acetoin as a
starting material), other substituent patterns can also be
obtained, such as the 4,5-diphenyl analogues (using
benzoin) [1,4]. These functionalized 2-aminopyrroles have
since served as starting point for synthesis of a variety of
pyrrole-containing derivatives, including annulated
bicyclic and tricyclic ring systems [4–12].
Yields for the 2-aminopyrroles by this classical synthesis
are generally moderate to good, including for the tert-butyl
esters (prepared using tert-butyl cyanoacetate). However,
yields for the methyl esters (prepared using methyl
cyanoacetate) are generally lower [2]. To our knowledge,
there has been no report of the formation of byproducts
from this classical pyrrole synthesis, nor has an explanation
for the lower yield for the methyl esters been offered.
Herein, we describe the isolation of a unique 5-hydroxy-γ-
lactam product from this pyrrole synthesis when using
methyl cyanoacetate as the CH2CN compound. The
RESULTS AND DISCUSSION
The 2-aminopyrrole 1a (Scheme 1) was recently needed
in our laboratory for an ongoing project. We were able to
prepare this known pyrrole [2] in ~15% purified yield
using our modified version of the classical synthesis and
following standard recrystallization of the crude product.
Our modified version involves heating the acetoin and
primary amine briefly without solvent to produce the α-
aminoketone, followed by reaction of this intermediate
with the CH2CN compound in refluxing EtOH. This
procedure, which eliminates the use of an organic solvent
in the first step and also allows for performing this step
in just 1–2 min, has given good yields of the pyrroles
containing sulfones at the 3-position [13]. However, our
yield for pyrrole 1a using this procedure was lower than
that originally reported by Sowell and coworkers (33%,
using Dean-Stark conditions and cyclohexane as solvent
for each step). Interestingly, TLC analysis of our crude
product mixture showed that a second major product was
formed in our reaction in addition to pyrrole 1a. This
other major product was thus being removed during the
recrystallization process to give pure samples of the more
crystalline 2-aminopyrrole, which is perhaps why the
byproduct has not been detected previously. To isolate
© 2018 Wiley Periodicals, Inc.

N. Suthiwangcharoen, A. C. Bean, M. Hassan, C. K. Eidell, and C. E. Stephens
Vol 000
Scheme 1. Synthesis of 2-aminopyrroles (1a–f), 5-hydroxy-γ-lactams (3a–f), and 5-methylene-γ-lactam (4e).
both products, we instead purified a crude product mixture
by column chromatography. This gave 2-aminopyrrole 1a,
eluting from the column first, in 10% yield, followed by
the second more polar product in a higher 30% yield. As
discussed in the succeeding text, this second/major
product was eventually determined to be 5-hydroxylactam
3a. We next repeated the reaction with other primary
amines, and these reactions also gave the corresponding 5-
hydroxylactam (3b–f) as the major product (29–34% yield),
with the classical 2-aminopyrroles (1b–d) also obtained in
lower 10–19% yield. Curiously, we were only able to obtain
pure 2-aminopyrrole when using benzylamines in the
synthesis. When phenethylamine or butylamine was used,
the 2-aminopyrrole (1e–f) was formed in even lower
amounts and we were unable to obtain pure samples via
chromatography (as pronounced darkening of the silica gel
column typically occurred during the chromatography, it is
possible that some decomposition of these 2-aminopyrroles
may have taken place during the purification process).
The 5-hydroxylactams (3a–f) were identified by IR and
NMR spectroscopy and combustion analysis. A key
indicator of structure was the presence of both a carbonyl
and OH absorbance in the IR spectrum. Also,
hydroxylactam 3e is a known compound, having been
previously prepared by a different cyclization route [14].
Our product assignment for hydroxylactams 3 is thus
confirmed by comparison with literature data for 3e. The
other hydroxylactams (3a–d, 3f) and 2-aminopyrroles
(1b–e) prepared here are each novel.
As might be expected, hydroxylactam 3a was formed in
only negligible/trace amounts when tert-butyl cyanoacetate
was used in our synthesis instead of the methyl ester (using
our modified conditions). In this case, the known 2-
aminopyrrole [2] with the tert-butyl ester at the 3-position
was the only significant product formed based on our
TLC analysis of the crude reaction mixture (benzylamine
was used as the amine), with that product being isolated
in 42% yield following chromatography (pronounced
darkening also occurred during the chromatography,
again suggesting some decomposition of the 2-
aminopyrrole on the column). The lack of formation of
the 5-hydroxylactam product 3a in this reaction indicates
that the cyclization predominantly occurred via attack of
the cyano group by the amine, with very little to no
attack of the tert-butyl ester. The greater selectivity for
the cyclization step when using tert-butyl cyanoacetate
(or malononitrile, cyanomethylsulfonyls, etc) helps
explain why yields for those 2-aminopyrroles are usually
higher than for those containing a methyl ester.
Finally, as the known 5-hydroxylactam 3e has
previously been dehydrated to the 5-methylene-γ-lactam
using formic acid [14], we attempted this reaction in
order to further confirm our structure assignment for
lactams 3. However, we used thionyl chloride as reagent
(in refluxing acetonitrile) instead of formic acid. This
reaction readily gave the dehydrated 5-methylene-γ-
lactam product 4e in 78% yield (Scheme 1).
Hydroxylactam 3 is apparently formed by air oxidation
of intermediate lactam 2, which, in turn, is most likely
formed by cyclization of the amino group of the α-
aminoketone onto the ester of methyl cyanoacetate
(instead of cyclization onto the cyano group to give
aminopyrrole 1). Although we have not observed lactam
2 by any means, such by TLC, its formation is
mechanistically sound and the air oxidation of such an
intermediate to give 3 is consistent with the oxidation of
other pyrrole type compounds at the 5-position to give 5-
hydroxy-γ-lactams [15], with oxidation to similar
hydroxylated pyrroles by air also known [16,17].
CONCLUSION
In conclusion, we have shown that a classical,
multicomponent 2-aminopyrrole synthesis yields 5-
hydroxy-γ-lactams as the major product when methyl
cyanoacetate is used as the activated CH2CN compound.
The classical 2-aminopyrrole is also formed, but as a minor
product (in negligible amount in some cases). One 5-
hydroxy-γ-lactam was subsequently dehydrated to the
known 5-methylene-γ-lactam using thionyl chloride in
good yield.
Journal of Heterocyclic Chemistry
DOI 10.1002/jhet

Month 2018
Isolation of 5-Hydroxy-γ-lactams from a Classical 2-Aminopyrrole Synthesis
EXPERIMENTAL J = 3.0 Hz), 147.3, 161.7 (d, J = 240 Hz), 166.6. Anal.
Calcd. for C15H17FN2O2 (276.31): C, 65.20; H, 6.20; N,
10.14. Found: C, 65.20; H, 6.32; N, 10.09.
General. Melting points were obtained using a digital
Mel-Temp apparatus and are uncorrected. IR spectra were
obtained using attenuated total reflection. NMR were
recorded on a Bruker 300 Avance instrument using
DMSO-d6 as solvent, with signals referenced to the
residual DMSO-d6 peaks (2.49 ppm for 1H NMR,
39.5 ppm for 13C NMR). Elemental analyses were
performed by Atlantic Microlab in Norcross, GA.
Distilled hexanes was used for chromatography.
2-Amino-1-(2-chlorobenzyl)-3-methoxycarbonyl-4,5-
dimethylpyrrole (1c). This compound was obtained as a
pale yellow fluffy solid, mp 159–161°C; IR (cmÀ1):
3421, 3321, 2926, 1645, 1611, 1523, 1499, 1439, 1300,
1
1250, 1106, 1070, 781, 749; H NMR: 1.80 (s, 3H), 2.03
(s, 3H), 3.64 (s, 3H), 5.01 (s, 2H), 6.02 (s, 2H), 6.33–
6.36 (m, 1H), 7.25–7.29 (m, 2H), 7.46–7.49 (m, 1H); 13C
NMR: 8.6, 11.4, 42.7, 49.6, 90.9, 111.1, 117.1, 126.4,
127.5, 128.6, 129.2, 131.0, 135.0, 147.1, 166.1. Anal.
Calcd. for C15H17ClN2O2 (292.76): C, 61.54; H, 5.85; N,
9.57. Found: C, 61.37; H, 5.87; N, 9.42.
General procedure for synthesis of pyrroles (1a–d) and
5-hydroxylactams (3a–f). Acetoin (3-hydroxy-2-butanone,
finely ground) (20 mmol) and the primary amine
(20 mmol) were combined in a 50 mL round-bottom
flask and heated on a hotplate until the mixture dissolved
and a homogenous yellow liquid was obtained (~1–2 min
of heating; the water vapor that condensed in the neck of
the flask was removed using a chemwipe). Absolute
EtOH (7–8 mL) was then added followed by methyl
cyanoacetate (20 mmol) and the solution was heated at
reflux (open condenser) for 2 h. The darkened solution
was then diluted with water and extracted with EtOAc.
The extract was washed with brine, dried over Na2SO4,
and then evaporated onto silica gel (high vacuum).
Column chromatography (SiO2) of the mixture eluting
with a gradient of hexanes:EtOAc (~10:1 to 1:1) gave,
after concentration of pure fractions in vacuo, the 2-
aminopyrrole (usually as a solid) followed by the 5-
hydroxylactam (usually as an oil), with yields given in
Scheme 1 (yields for pyrroles 1e–f were negligible and
these could not be obtained in pure form). Analytical
samples of the isolated pyrroles 1a–d were obtained by
recrystallization from MeOH. Crystalline samples of the
5-hydroxylactams 3a–f were generally obtained by slow
evaporation from diethyl ether, diethyl ether/hexanes, or
toluene/cyclohexane.
2-Amino-1-(2-bromobenzyl)-3-methoxycarbonyl-4,5-
dimethylpyrrole (1d). This compound was obtained as a
pale yellow fluffy solid, mp 159–160°C; IR (cmÀ1):
3424, 3326, 2926, 1646, 1607, 1501, 1438, 1300, 1251,
1
1105, 1028, 781, 748; H NMR: 1.79 (s, 3H), 2.03 (s,
3H), 3.64 (s, 3H), 4.95 (s, 2H), 6.00 (s, 2H), 6.31 (d,
J = 8.0 Hz, 1H), 7.18–7.21 (t, 1H), 7.28–7.31 (t, 1H),
7.63 (d, J = 7.5 Hz, 1H); 13C NMR: 8.6, 11.4, 45.2, 49.7,
90.9, 111.2, 117.0, 121.1, 126.5, 128.1, 129.0, 132.4,
136.4, 147.1, 166.1. Anal. Calcd. for C15H17BrN2O2
(337.21): C, 53.43; H, 5.08; N, 8.31. Found: C, 53.40; H,
5.11; N, 8.18.
DATA FOR 5-HYDROXY-γ-LACTAMS (3A–F)
1-Benzyl-5-hydroxy-4,5-dimethyl-2-oxo-2,5-dihydro-1H-
pyrrole-3-carbonitrile (3a). This compound was obtained
as pale pink/colorless crystals, mp 96–97°C (toluene/
cyclohexane); IR (cmÀ1): 3397, 3063, 3036, 2985, 2933,
2235, 1684, 1414, 1154, 737, 698; 1H NMR: 1.27 (s,
3H), 2.20 (s, 3H), 4.35 (d, J = 16 Hz, 1H), 4.59 (d,
J = 16 Hz, 1H), 6.62 (s, 1H), 7.21–7.31 (m, 5H); 13C
NMR: 12.7, 22.3, 41.6, 89.8, 107.0, 112.4, 126.9, 127.3,
128.2, 138.1, 162.8, 175.9. Anal. Calcd. for C14H14N2O2
(242.27): C, 69.41; H, 5.82; N, 11.56. Found: C, 69.51;
H, 5.83; N, 11.67.
DATA FOR PYRROLES (1A–D):
1-(4-Fluorobenzyl)-5-hydroxy-4,5-dimethyl-2-oxo-2,5-
dihydro-1H-pyrrole-3-carbonitrile (3b).
This compound
2-Amino-1-benzyl-3-methoxycarbonyl-4,5-dimethylpyrrole
(1a). This compound was obtained as a tan fluffy solid,
mp 138–140°C (Lit mp: 138–139°C) [2].
was obtained as colorless crystals (diethyl ether), mp
100–101°C; IR (cmÀ1): 3352, 3007, 2991, 2940, 2933,
2231, 1687, 1672, 1604, 1509, 1404, 1225, 1151, 1096,
2-Amino-1-(4-fluorobenzyl)-3-methoxycarbonyl-4,5-
dimethylpyrrole (1b). This compound was obtained as a
1
856, 775, 760; H NMR: 1.26 (s, 3H), 2.17 (s, 3H), 4.32
colorless fine crystalline solid, mp 138–139°C; IR
(d, J = 16 Hz, 1H), 4.52 (d, J = 16 Hz, 1H), 6.63 (s, 1H),
7.08–7.13 (m, 2H), 7.30–7.34 (m, 2H); 13C NMR: 12.7,
22.3, 41.0, 89.9, 107.0, 112.5, 115.0 (d, J = 22 Hz),
129.5 (d, J = 8.3 Hz), 134.3 (d, J = 2.3 Hz), 161.3 (d,
J = 241 Hz), 162.8, 176.0. Anal. Calcd. for C14H13FN2O2
(260.27): C, 64.61; H, 5.03; N, 10.76. Found: C, 64.50;
H, 5.06; N, 10.72.
(cmÀ1): 3428, 3322, 2919, 1641, 1606, 1501, 1447,
1
1359, 1306, 1255, 1227, 1108, 1094, 810, 781; H NMR:
1.84 (s, 3H), 1.99 (s, 3H), 3.62 (s, 3H), 4.96 (s, 2H), 6.00
(s, 2H), 6.99–7.04 (m, 2H), 7.10–7.17 (m, 2H); 13C
NMR: 9.4, 11.8, 44.3, 50.1, 91.4, 111.4, 115.7 (d,
J = 21 Hz), 117.7, 128.6 (d, J = 8.3 Hz), 134.5 (d,
Journal of Heterocyclic Chemistry
DOI 10.1002/jhet

N. Suthiwangcharoen, A. C. Bean, M. Hassan, C. K. Eidell, and C. E. Stephens
Vol 000
1-(2-Chlorobenzyl)-5-hydroxy-4,5-dimethyl-2-oxo-2,5-
dried (Na2SO4) and evaporated onto silica gel (high
vacuum). Column chromatography (SiO2) eluting with
hexanes:EtOAc (2:1) gave, after concentration of pure
dihydro-1H-pyrrole-3-carbonitrile (3c).
This compound
was obtained as pale golden crystals, mp 143–144°C
(diethyl ether); IR (cmÀ1): 3235, 2988, 2232, 1678, 1432,
fractions, the product as
a
yellow/brown solid
1
1394, 1163, 1050, 1038, 949, 746; H NMR: 1.30 (s, 3H),
(0.185 g, 78%), mp 128–129°C. Recrystallization from
EtOH/water gave an olive green solid, mp 130–131°C
(Lit mp: 130–131°C) [14]. A second recrystallization
from a larger volume of hexanes gave fine yellow
crystals, mp 131–133°C; IR (cmÀ1): 3058, 3009,
2968, 2939, 2225, 1715, 1633, 1612, 1354, 1190,
1129, 865, 760, 706, 640; 1H NMR: 2.32 (s, 3H),
2.78 (t, 2H), 3.79 (t, 2H), 5.48 (d, J = 2.9 Hz, 1H),
5.51 (d, J = 2.9 Hz, 1H), 7.18–7.26 (m, 5H); 13C
NMR: 12.0, 33.5, 40.6, 101.8, 105.8, 112.4, 126.3,
128.2, 128.7, 138.1, 143.4, 158.9, 163.2. This reaction
could also be accomplished in comparable yield using
POCl3.
2.22 (s, 3H), 4.51 (s, 2H), 6.65 (s, 1H), 7.23–7.30 (m, 3H),
7.41–7.45 (m, 1H); 13C NMR: 12.7, 21.9, 39.3, 89.8,
107.0, 112.4, 127.2, 128.5, 128.7, 129.1, 131.4, 134.7,
162.9, 176.0. Anal. Calcd. for C14H13ClN2O2 (276.72): C,
60.77; H, 4.74; N, 10.12. Found: C, 60.60; H, 4.87; N, 10.04.
1-(2-Bromobenzyl)-5-hydroxy-4,5-dimethyl-2-oxo-2,5-
dihydro-1H-pyrrole-3-carbonitrile (3d).
This compound
was obtained as a fine golden crystalline solid, mp 154–
156°C (diethyl ether); IR (cmÀ1): 3253, 2988, 2232,
1
1674, 1436, 1393, 1162, 1028, 950, 744; H NMR: 1.31
(s, 3H), 2.24 (s, 3H), 4.49 (s, 2H), 6.68 (s, 1H), 7.19–
7.25 (m, 2H), 7.32–7.37 (m, 1H), 7.61 (d, J = 7.8 Hz,
1H); 13C NMR: 12.7, 21.9, 41.9, 89.7, 106.9, 112.4,
121.6, 127.7, 128.4, 128.9, 132.3, 136.2, 162.9, 176.0.
Anal. Calcd. for C14H13BrN2O2 (321.17): C, 52.36; H,
4.08; N, 8.72. Found: C, 52.34; H, 4.05; N, 8.62.
Acknowledgments. We thank undergraduate students Reshma
Benny and Allyson Knapp for technical lab assistance. We also
thank our department for partial funding of this work.
5-Hydroxy-4,5-dimethyl-2-oxo-1-phenethyl-2,5-dihydro-1H-
pyrrole-3-carbonitrile (3e). This compound was obtained as
a pale orange crystalline solid, mp 136–137°C (diethyl ether);
REFERENCES AND NOTES
IR (cmÀ1): 3262, 3020, 2996, 2940, 2232, 1681, 1416, 1158,
1
1103, 753, 701; H NMR: 1.32 (s, 3H), 2.17 (s, 3H), 2.88
[1] Roth, H. J.; Eger, K. Arch Pharm 1975, 308, 179.
(t, 2H), 3.49 (t, 2H), 6.53 (s, 1H), 7.17–7.33 (m, 5H); 13C
NMR: 12.6, 21.7, 34.5, 89.7, 106.9, 112.4, 126.2, 128.4,
128.6, 139.0, 162.3, 175.6 (one signal obscured by DMSO
peaks). Anal. Calcd. for C15H16N2O2 (256.30): C, 70.29; H,
6.29; N, 10.93. Found: C, 70.36; H, 6.19; N, 10.94.
[2] Laks, J. S.; Ross, J. R.; Bayomi, S. M.; Sowell, J. W. Synthesis
1985, 1985, 291.
[3] Mattson, R. J.; Wang, L. C.; Sowell, J. W. J Heterocyclic
Chem 1980, 17, 1793.
[4] Vorob’ev, E. V.; Kurbatov, E. S.; Krasnikov, V. V.;
Mezheritskii, V. V.; Usova, E. V. Russ Chem Bull Int Ed 2006, 55, 1492.
1-Butyl-5-hydroxy-4,5-dimethyl-2-oxo-2,5-dihydro-1H-
pyrrole-3-carbonitrile (3f). This compound was obtained
[5] Ross, J. R.; Laks, J. S.; Wang, D. L.; Sowell, J. W. Synthesis
1985, 1985, 796.
as golden micro-needles, mp 76–78°C (diethyl
[6] Vishwakarma, L. C.; Sowell, J. W. J Heterocyclic Chem 1985,
22, 1429.
ether/hexanes); IR (cmÀ1): 3324, 2961 2875, 2231, 1688,
1
[7] Ross, J. R.; Vishwakarma, L. C.; Sowell, J. W. J Heterocyclic
Chem 1987, 24, 661.
1449, 1409, 1375, 1153, 1095, 766; H NMR: 0.87 (t,
3H), 1.27 (m, 2H), 1.39 (s, 3H), 1.52 (m, 2H), 3.14 (m,
1H), 3.26 (m, 1H), 6.44 (s, 1H); 13C NMR: 12.6, 13.7,
19.8, 21.9, 30.7, 38.3, 89.8, 107.0, 112.6, 162.2, 175.4.
Anal. Calcd. for C11H16N2O2 (208.26): C, 63.44; H, 7.74;
N, 13.45. Found: C, 63.28; H, 7.61; N, 13.18.
[8] Player, M. R.; Sowell, J. W. J Heterocyclic Chem 1992, 30, 125.
[9] Stephens, C. E.; Sowell, J. W. J Heterocyclic Chem 1996,
33, 1615.
[10] Muller, C. E.; Geis, U.; Grahner, B.; Lanzner, W.; Eger, K.
J Med Chem 1996, 39, 2482.
[11] Hassan, H. M.; Bayomi, S. M.; Ali, M. M. Indian J Chem
2000, 39B, 764.
[12] Willeman, C.; Grunert, R.; Bednarski, P. J.; Troschutz, R.
Bioorg Med Chem 2009, 17, 4406.
DEHYDRATION OF LACTAM 3E TO
5-METHYLENE-γ-LACTAM 4E
[13] Suthiwangcharoen, N.; Pochini, S. M.; Sweat, D. P.; Stephens,
C. E. J Heterocyclic Chem 2011, 48, 706.
4-Methyl-5-methylene-2-oxo-1-phenethyl-2,5-dihydro-1H-
[14] Adhikari, R.; Jones, D. A.; Liepa, A. J.; Nearn, R. H. Aust J
Chem 2005, 58, 882.
pyrrole-3-carbonitrile (4e).
To
a
solution of 5-
[15] Howard, J. K.; Rihak, K. J.; Bissember, A. C.; Smith, J. A.
Chem Asian J 2016, 11, 155.
hydroxylactam 3e (0.256 g, 1.0 mmol) in dry acetonitrile
(5 mL) was added thionyl chloride (0.1 mL, 1.37 mmol)
and the mixture was refluxed for 40 min. The darkened
solution was then concentrated in vacuo to a brown oil,
which was dissolved in EtOAc and washed with
sodium bicarbonate solution. The solution was then
[16] Cirrincione, G.; Dattolo, G.; Almerico, A. M.; Aiello, E.;
Jones, R. A.; Dawes, H. M.; Hursthouse, M. B. J Chem Soc Perkin Trans
1959, 1, 1987.
[17] Hawkins, R. A.; Stephens, C. E. Tetrahedron Lett 2010,
51, 6129.
Journal of Heterocyclic Chemistry
DOI 10.1002/jhet