One-pot and catalyst-free amidation of ester: A matter of non-bonding interactions

Article abstract of DOI:10.1016/j.tetlet.2011.10.043

This paper describes a one-pot procedure for acetamide synthesis directly from amines and pyrimidine ester, without any catalyst or coupling agents. The inexpensive and simple reaction conditions are the most important features of this amidation. This rea

Full text of DOI:10.1016/j.tetlet.2011.10.043

Tetrahedron Letters 52 (2011) 6796–6799
Contents lists available at SciVerse ScienceDirect
Tetrahedron Letters
journal homepage: www.elsevier.com/locate/tetlet
One-pot and catalyst-free amidation of ester: a matter of
non-bonding interactions
Olivier Rezazgui a,b, Benjamin Boëns a, Karine Teste a, Julien Vergnaud a, Patrick Trouillas b,c
,
Rachida Zerrouki a,
⇑
a Laboratoire de Chimie des Substances Naturelles EA1069, Faculté des Sciences et Techniques, 123 Avenue Albert Thomas, F-87060 Limoges, France
b EA4021, Faculté de Pharmacie, 2 rue du Docteur Marcland, F-87025 Limoges, France
c Laboratoire de Chimie des Matériaux Nouveaux, Université de Mons Place du Parc 20, 7000 Mons, Belgium
a r t i c l e i n f o
a b s t r a c t
Article history:
This paper describes a one-pot procedure for acetamide synthesis directly from amines and pyrimidine
ester, without any catalyst or coupling agents. The inexpensive and simple reaction conditions are the
most important features of this amidation. This reaction was performed with various amines, showing
Received 15 September 2011
Revised 4 October 2011
Accepted 7 October 2011
Available online 15 October 2011
that long range stabilizing interactions (H-bonding and
chemoselectivity.
p-stacking) are the driving force for
Ó 2011 Elsevier Ltd. All rights reserved.
Keywords:
Amidation
Uracil acetate
DFT calculations
Non-bonding interactions
C–N bonds are widely present in natural (DNA, peptides, tetra-
pyrrolic macrocycles, etc.) as well as in synthetic products.1 Amide
bonds exemplified these bonds, being found in many synthetic
substances as intermediates or as active pharmaceutical products
or prodrugs.2 However its formation remains a challenging issue
of organic synthesis. The process for the formation of amide bonds
from esters involves hydrolysis of the ester followed by the conver-
sion of carboxylic acid to a more reactive functional group or in situ
activation by using coupling agents (BOP, DCC, NHS, HBTU, HOBt or
ByBOP).3 Both chemical and enzymatic methods for direct amida-
tion of esters were reported in the literature.4 In most of these
methods, complexes of amine derived from strong bases or acids
are used,5 such as Grignard or alkylaluminium reagents.6 Due to
their corrosive nature, expensive reagents, and formation of by-
products, these methods are not amenable for large scale reactions.
We recently developed a new and simple method for the ami-
dation of pyrimidine esters in the field of mustard derivatives.7 Di-
rectly from uracil acetate and diethanolamine, by refluxing and
without any catalyst, we obtained an acetamide derivative of the
nucleic acid. This process appears as an easy and ‘green’ procedure
providing the final product with high yields and selectivity, as no
esterification reaction was observed.
Scheme 1. Direct amidation of uracil ester with diethanolamine.
Here we propose to extend this methodology for the amidation
of uracil acetate with a series of amine derivatives and to deter-
mine the favoring factors.
A preliminary study using uracil ester and diethanolamine has
been conducted in ethyl alcohol (Scheme 1).
The influence of the concentration of diethanolamine (from
large excess to equimolar ratio) was first investigated (Table 1).
Large excess of amine gives the product in high yields and within
relatively short time in refluxing ethanol (entry 1). In order to re-
⇑
Corresponding author. Tel.: +33 5 55 45 72 24; fax: +33 5 55 45 72 02.
E-mail address: rachida.zerrouki@unilim.fr (R. Zerrouki).
0040-4039/$ - see front matter Ó 2011 Elsevier Ltd. All rights reserved.
doi:10.1016/j.tetlet.2011.10.043

O. Rezazgui et al. / Tetrahedron Letters 52 (2011) 6796–6799
6797
Table 1
Influence of diethanolamine amount and activation
Entry
Diethanolamine (equiv)
Activation
Reaction time (h)
Isolated yield (%)
1
2
3
4
5
6
10
1.5
1.2
1.02
1.02
1.02
D
D
D
D
24
96
96
144
10
13
94
80
75
90
70
55
Ultrasound
Microwave
duce such large quantities of amine (especially when it is not com-
mercial), and because the further purification is awkward with a
large reactant concentration, the amidation reaction was tested
with lower and more reasonable excesses of amine (entries 2 and
3). These conditions appear as adequate compromises as the reac-
tion also provided the final product in high yields. An equimolar
amount of amine (entry 4) also gives product 1 in a similar yield
with respect to entry 1. Nonetheless for entries 2, 3, and 4, the
reaction time was significantly increased.
The influence of two activation processes was investigated un-
der equimolar conditions. Compared to classical heating, ultra-
sound activation (entry 5) and microwave irradiations (entry 6)
dramatically reduced the reaction times, but also the yields, partic-
ularly in the case of microwave activation.
To rationalize chemoselectivity, 3D conformations of reactants
and products, and Gibbs energy of reaction were calculated at
the quantum level.10 The best stability observed with compound
1 is explained by the existence of two H-bonds, giving a stabilizing
Gibbs energy of À6.8 kcal molÀ1, rationalizing a yield of 90%. The
hydroxyl ethyl group can form two H-bonds, a weak H-bond with
the uracil carbonyl moiety (2.07 Å) and a stronger bond with the
carbonyl group of the amide function (1.83 Å) (Fig. 1a).
For compound 6, only one H-bond can be formed preferentially
between the hydroxyethyl group and the carbonyl group of the
amide bond (Fig. 1b), as the H-bond is weaker with the uracil car-
bonyl moiety (1.83 Å vs 2.06 Å, respectively). This compound is
slightly less stable than compound 1. Nonetheless due to the pres-
ence of
a strong H-bond it still exhibits high stability
This preliminary study confirms that a stoichiometric amount
of diethanolamine (1.02 equiv) is sufficient to afford the final prod-
uct in high yield and reasonable reaction times in refluxing ethyl
alcohol after only recrystallization from ethanol.
(D
G = À6.2 kcal molÀ1). No H-bond can be formed in compound 4
which explains non-product formation (
D
G° = À2.4 kcal molÀ1).
Surprisingly the formation of compound 8 is favored (60% yield).
In this case no H-bond is expected and may explain stabilization.
To validate this method of catalyst-free amidation in refluxing
ethanol, a series of primary and secondary amines were tested un-
der these experimental conditions (Table 2).
This easy method can be used for a range of amines providing
relatively high yields (Table 2, entries 1, 3, 5, 6 and 8). The process
surprisingly appears chemoselective. The presence of one to two
hydroxyethyl groups on the nitrogen of amine (compounds 1, 3
and 6) significantly enhances yields and decreases reaction time
with respect to the compounds without this moiety (compounds
4, 5, 7, 8 and 9).
Nevertheless the compound is stabilized by p-stacking interaction
between the aromatic rings (Fig. 1c). The flexibility of the molecule
allows an optimized parallel displaced stack between both aromatic
rings (Fig. 1c), thus stabilizing the final compound
(
D
G° = À4.4 kcal molÀ1) by dispersive interactions. Such stabilizing
interactions are also observed in compound
3
(DG° =
À6.0 kcal molÀ1), for whichtheconcomitantpresenceofone H-bond
with the carbonyl group of the amide bond (Fig. 1d) explains the effi-
cient 70% yield.9 For compound 8,
p-stacking is the only driving force
for amidation. This influences the reaction time which is signifi-
Table 2
Direct amidation of different amines8,11
Entry
Amine
Reaction time (days)
Compound
Isolated yields (%)
D )
G°10 (kcal molÀ1
NH
NH
1
2
6
1
2
90
10
À6.8
À3.2
HO
OH
OH
22
3
4
5
9
34
11
3
4
5
70
À6.0
À2.4
À3.9
NH
OH
a
NH
NH
—
O
46
H2N
6
7
6
6
7
70
À6.2
À2.6
OH
NH2
a
34
—
8
9
20
15
8
9
60
À4.4
NH2
NH2
a
—
2.5
a
TLC did not show any evolution of reaction.

6798
O. Rezazgui et al. / Tetrahedron Letters 52 (2011) 6796–6799
Figure 1. 3D conformation of compounds 1 (a), 6 (b), 8 (c), 3 (d) and 2 (e) exhibiting non-bonding interactions. For compound 8 (c) the view is perpendicular to the plane
defined by both aromatic rings, the upper ring (uracil moiety) being drawn in balls & cylinders while the lower ring is in sticks.
The absence or non-efficiency of the amidation reaction in these
cases is mainly attributed to the dramatic increase of steric hin-
drance due to the alkyne group, preventing molecular flexibility
0
(Fig. 1e) hence producing less of stable compounds (
D
G° = À2.6
-1
-2
-3
-4
-5
-6
-7
-8
and À3.2 kcal molÀ1 for both compounds, respectively).
R2 = 0.92
Reaction with the secondary cyclic amine, morpholine (entry 5),
exhibited 46% yield. This efficient yield cannot be explained by H-
bonding and
mation of a relatively stable compound (
p-stacking, however the calculations confirm the for-
D
G° = À3.9 kcal molÀ1). In
this particular case, nucleophilicity of this cyclic amine is more
probably the major descriptor explaining reaction feasibility.
A series of pyrimidine esters have been synthesized by a simple
and very convenient amidation in refluxing ethanol. The yields are
very promising thus highlighting the relevancy of this method to
obtain amides easily and with a variety of amines. The avoidance
of corrosive reactants or coupling agents is the main benefit of this
procedure. All final products were obtained by a simple recrystal-
lization from ethyl alcohol. A surprising chemoselectivity has been
observed, which has been rationalized by the presence of intra-
molecular stabilizing effects attributed to long range interactions,
0
20
40
60
80
100
Yield (%)
Figure 2. Correlation between Gibbs energies of reaction (
yields. The regression coefficient is obtained without taking compound 9 into
account (see footnote 12).
D
G°) and experimental
H-bonding, and
p-stacking. We have also shown that the method
of calculation must be carefully chosen in order to fit with experi-
mental data, the use of recent functionals including dispersive ef-
fects is mandatory.10 The correlation between the calculated
DG°
cantly higher than for compounds having also H-bond as a driving
and the experimental yields is particularly noteworthy. Taking all
nine compounds into account, the regression coefficient between
both parameters is around 0.7; excluding compound 912 it is higher
than 0.9 (Fig. 2).
force (e.g., 20 vs 9 days for compounds 8 and 3, respectively).
For compound 9, three reasons may explain the positive
(Table 2): (i) the flexibility of the aniline moiety does not allow
the formation of -stacking complexes, (ii) no H-bond can be
DG°
p
formed, (iii) the doublet of the amine is delocalized over the aro-
matic ring, being less available for the reaction.
Supplementary data
The amidation reaction with propargylamine did not occur after
34 days of reaction (entry 7); while it is completed with N-propar-
gylethanolamine (entry 2) within 22 days, but at a low 10% yield.
Supplementary data (Gibbs energies of reaction obtained with a
series of DFT functionals including more or less HF (Hartree-Fock)

O. Rezazgui et al. / Tetrahedron Letters 52 (2011) 6796–6799
6799
intermolecular interaction (between solute and solvent). It is known that
such solvent effect can influence mainly kinetics of the reactions involving the
H-atom engaged in these intermolecular H-bonds (Litwinienko, G.; Ingold, K. U.
Acc. Chem. Res. 2007, 40, 222–230) according to the bH2 Abraham’s coefficient,
characteristic of HBA capacities of solvents (around 0.4 for ethanol). In our
case, thermodynamics would probably be corrected by specific H-bond
interactions with solvent. However the general trend would be the same and
exchanges and non-bonding corrections) associated with this arti-
cle can be found, in the online version, at doi:10.1016/
j.tetlet.2011.10.043.
References and notes
taking explicitly the solvent into account would have required
computational effort. Geometries, energies, and Gibbs energies (G) at 298 K
of the reactants and products were determined at the PCM-wB97XD/6–
a huge
1. (a) Trost, B. M.; Fleming, I.; Heathcock, C. H. In Comprehensive Organic Synthesis;
Pergamon Press: New York, 1991; Vol. 6,; (b) Larock, R C. Comprehensive
Organic Transformations; Wiley-VCH: New York, 1973.
2. (a) Cupido, T.; Tulla-Puche, J.; Spengler, J.; Alberico, F. Curr. Opin. Drug Discov.
Devel. 2007, 10, 768–783; (b) Bode, J. W. Curr. Opin. Drug Discov. Devel. 2006, 9,
765–775; (c) Brown, W. Idrugs 1999, 2, 1059–1068; (d) Albericio, F. Curr. Opin.
Chem. Biol. 2004, 8, 211–221.
3. (a) Jones, J. The Chemical Synthesis of Peptides; Oxford University Press: Oxford,
1991; (b) Bodanszky, M. Principles of Peptide Synthesis; Springer: Berlin, 1984.
4. (a) Martin, S. F.; Dwyer, M. P.; Lynch, C. L. Tetrahedron Lett. 1998, 39, 1517; (b)
Wee, A. G.; Liu, B.; McLeod, D. D. J. Org. Chem. 1998, 63, 4218; (c) Basha, A.;
Lipton, M.; Weinreb, S. M. Tetrahedron Lett. 1977, 18, 4171; (d) Marshall, J. A.;
Luke, G. P. J. Org. Chem. 1993, 58, 6229.
31+G(d,p) level. The Gibbs energy of reaction under standard conditions
DG°
was obtained as the difference [G(products) À G(reactants)]. All calculations
were carried out using Gaussian09 (Frisch, M. J. et al. Gaussian 09, Revision A.02
Wallingford CT, 2009).
11. Spectroscopic data (400.13 MHz, DMSO-d6).
Compound 1: 1H NMR (d). Uracil: (1H, s, NH), 7.43 (1H, d, J = 7.8 Hz, H6), 5.55
(1H, d, J = 7.8 Hz, H5); Acetamide: 4.66 (2H, s, –CH2); Hydroxyethyl chains: 4.94
(1H, t, J = 5.2 Hz, OH ), 4.70 (1H, t, J = 5.4 Hz, OHb), 3.58 (2H, dt, J = 5.3 Hz,
a
J = 5.2 Hz, CH2– OH ), 3.47 (2H, dt, J = 5.4 Hz, J = 5.6 Hz, CH2– OHb), 3.43 (2H, t,
a
J = 5.6 Hz, CH2–Nb), 3.35 (2H, t, J = 5.3 Hz, CH2–N ). Pure product as a white
a
solid: mp = 142 °C. Compound 2: 1H NMR (d). Major rotamer (70%): Uracil:
11.27 (1H, s, NH), 7.47 (1H, d, J = 7.9 Hz, H6), 5.55 (1H, d, J = 7.9 Hz, H5);
Acetamide: 4.70 (2H, s, –CH2); Hydroxyethyl: 4.98 (1H, t, J = 5.3 Hz, OH), 3.63
(2H, dt, J = 5.3 Hz, J = 5.3 Hz, –CH2–O), 3.49 (2H, t, J = 5.3 Hz, –CH2–N);
Propargyl: 4.19 (2H, d, J = 2.3 Hz, –CH2), 3.20 (1H, t, J = 2.3 Hz, H alkyne).
Minor rotamer (30%): Uracil: 11.27 (1H, s, NH), 7.49 (1H, d, J = 7.7 Hz, H6), 5.55
(1H, d, J = 7.7 Hz, H5); Acetamide: 4.68 (2H, s, –CH2); Hydroxyethyl: 4.74 (1H, t,
J = 5.2 Hz, OH), 3.49 (2H, br dt, –CH2–O), 3.40 (2H, m, –CH2–N); Propargyl: 4.29
(2H, d, J = 2.4 Hz, –CH2), 3.40 (1H, m, H alkyne). Pure product as a colorless oil.
Compound 3: 1H NMR (d). Major rotamer (70%): Uracil: 11.26 (1H, s, NH), 7.52
(1H, d, J = 7.8 Hz, H6), 5.57 (1H, d, J = 7.8 Hz, H5); Acetamide: 4.77 (2H, s, –CH2);
Hydroxyethyl: 4.96 (1H, t, J = 5.2 Hz, OH), 3.57 (2H, dt, J = 5.2 Hz, J = 5.2 Hz, –
CH2–O), 3.33 (2H, t, J = 5.2 Hz, –CH2–N); Benzyl: 7.41–7.22 (5H, m, HAr), 4.57
(2H, s, –CH2). Minor rotamer (30%): Uracil: 11.26 (1H, s, NH), 7.54 (1H, d,
J = 7.7 Hz, H6), 5.56 (1H, d, J = 7.7 Hz, H5); Acetamide: 4.67 (2H, s, –CH2);
Hydroxyethyl: 4.68 (1H, t, J = 6.0 Hz, OH), 3.45 (2H, dt, J = 6.0 Hz, J = 6.0 Hz, –
CH2–O), 3.28 (2H, t, J = 6.0 Hz, -CH2-N); Benzyl: 7.41–7.22 (5H, m, HAr), 4.64
(2H, s, –CH2). Pure product as a white solid: mp = 160 °C. Compound 5: 1H NMR
(d). Uracil: (1H, s, NH), 7.48 (1H, d, J = 7.8 Hz, H6), 5.57 (1H, d, J = 7.8 Hz, H5);
Acetamide: 4.62 (2H, s, –CH2); Morpholine: 3.62 (2H, t, J = 4.7 Hz, –CH2), 3.57
(2H, t, J = 4.7 Hz, –CH2), 3.46 (2H, t, J = 4.7 Hz, –CH2), 3.43 (2H, t, J = 4.7 Hz, –
CH2). Pure product as a white solid: mp = 178 °C. Compound 6: 1H NMR (d).
Uracil: (1H, s, NH), 7.53 (1H, d, J = 7.8 Hz, H6), 5.54 (1H, d, J = 7.8 Hz, H5);
Acetamide: 8.17 (1H, t, J = 5.3 Hz, NH), 4.31 (2H, s, –CH2); Hydroxyethyl: 4.69
(1H, br t, OH), 3.40 (2H, br dt, –CH2–O), 3.14 (2H, dt, J = 5.8 Hz, J = 5.8 Hz, –CH2–
N). Pure product as a white solid: mp = 238 °C. Compound 8: 1H NMR (d). Uracil:
(1H, s, NH), 7.58 (1H, d, J = 7.8 Hz, H6), 5.56 (1H, d, J = 7.8 Hz, H5); Acetamide:
8.66 (1H, t, J = 5.8 Hz, NH), 4.38 (2H, s, –CH2); Benzyl: 7.34–7.22 (5H, m, HAr),
4.30 (2H, d, J = 5.8 Hz, –CH2). Pure product as a white solid: mp = 236 °C.
12. Compound 9 is significantly less stable than 4 and 7 (Table 2), while the yield is
5. Schlecker, W.; Huth, A.; Ottow, E.; Mulzer, J. Synthesis 1995, 1225.
6. (a) Shimizu, T.; Osako, K.; Nakata, T. Tetrahedron Lett. 1997, 38, 2685; (b)
Akakura, M.; Yamanmoto, H. Synlett 1997, 277; (c) Patterson, J. W. J. Org. Chem.
1995, 60, 4542; (d) Beerli, R.; Rebek, J., Jr. Tetrahedron Lett. 1995, 36, 1813; (e)
Sidler, D. R.; Lovelace, T. C.; McNamara, J. M.; Reider, P. J. J. Org. Chem. 1994, 59,
1231; (f) Kim, M. Y.; Starreet, J. E.; Weinreb, S. M. J. Org. Chem. 1981, 46, 5383.
7. Hadj-Bouazza, A.; Teste, K.; Colombeau, L.; Chaleix, V.; Zerrouki, R.; Kraemer,
M.; Sainte Catherine, O. Nucleosides Nucleotides and Nucleic Acids 2008, 27, 439–
448.
8. General procedure (Table 2): 1-Ethoxycarbonyluracil (100 mg, 0.5 mmol) in
ethanol (3 mL), at 70 °C. Amine (1.02 equiv) was then added after dissolution of
the alkylated uracil. Products have been obtained by a simple recrystallization
from ethanol.
9. Compound 3, actually exhibits three conformers with very similar energies
having (i) one H-bond between the hydroxyethyl group and the carbonyl group
of the amid bond, (ii)
p-stacking stabilization as for compound 8 and (iii)
combination of both. Disregarding the comparison obtained from calculations
which must be very accurate to discriminate relative stabilities, here we
believe that the third is the most probable conformer in solution.
10. Calculations were performed at the DFT (density functional theory) level using
functionals introducing dispersive corrections. Both wB97XD and B3LYP-D
were used and provided the same trend. The classical hybrid functionals do not
allow providing accurate results due to the importance of H-bond and non-
bonding interaction in this study. Only the results obtained with wB97XD are
shown in Table 2. Results obtained with the other functional are given as
Supplementary data. The 6–31+G(d,p) basis set was used. The conformational
analysis was carefully performed at the DFT level, scanning flexible torsion
angles. The lowest energy conformers (ground-state geometries) were
confirmed by vibrational frequency analysis that indicated the absence of
imaginary frequencies. The solvent effects were implicitly taken into account
with the PCM (polarizable continuum model) method, in which the solute is
embedded in a shape-adapted cavity surrounding by a dielectric continuum
characterized by its dielectric constant. Ethanol is a relatively good HBA (H-
bond acceptor) solvent, which probably allows specific non-bonding
0
(no yield actually) for the three compounds and therefore cannot be
distinguished. Compound 9 is probably much more difficult to be formed with
respect to 4 and 7 but this cannot be observed.