A mechanistic study of the ammonolysis of alkyl acetoacetates in water. Formation of 1,5-dimethyl-2,6,9-triaza-bicyclo[3.3.1]nonane-3,7-dione as the main product

Article abstract of DOI:10.1016/S0040-4020(01)01127-9

Ammonolysis of alkyl acetoacetates with 15% NH₃ in water at room temperature initially lead to formation of alkyl β-aminocrotonates which slowly converted into 1,5-dimethyl-2,6,9-triaza-bicyclo[3.3.1]nonane-3,7-dione as the main product.

Full text of DOI:10.1016/S0040-4020(01)01127-9

TETRAHEDRON
Pergamon
Tetrahedron 58 (2002) 55±60
A mechanistic study of the ammonolysis of alkyl acetoacetates
in water. Formation of 1,5-dimethyl-2,6,9-triaza-
bicyclo[3.3.1]nonane-3,7-dione as the main product
Rodrigo Paredes,a Rodrigo Abonia,a,p John Cadavid,a Rodolfo Moreno-Fuquen,a
Alonso Jaramillo,a Angelina Hormaza,b Alfonso Ramirezc and Allan Kennedyd
aDepartment of Chemistry, Universidad del Valle, A.A. 25360 Cali, Colombia
b
Â
Department of Chemistry, Universidad Nacional, A.A. 3840 Medellõn, Colombia
Â
c
Department of Chemistry, Universidad del Cauca, Calle 2 No 1AÐ45, Popayan, Colombia
dDepartment of Pure and Applied Chemistry, University of Strathclyde, 295 Cathedral Street, Glasgow G1 1XL, Scotland, UK
Received 25 September 2001; revised 8 November 2001; accepted 9 November 2001
AbstractÐAmmonolysis of alkyl acetoacetates with 15% NH3 in water at room temperature initially lead to formation of alkyl b-amino-
crotonates which slowly converted into 1,5-dimethyl-2,6,9-triaza-bicyclo[3.3.1]nonane-3,7-dione as the main product. q 2002 Published by
Elsevier Science Ltd.
1. Introduction
2. Results and discussion
We have presented evidence in favor of an enol mechanism
to explain the rapid alkaline hydrolysis of ethyl acetoacetate
in water and the ®rst-order kinetics observed when a dilute
equimolar (1.0£1024 M) solution of ester and HO2 is
allowed to react at room temperature.1 The mechanism
®rst involves the establishment of base catalyzed keto±
enol equilibrium, via the enolate anion, between the
ester and enols CH3C(OH)vCHCO2Et and CH3COCHv
C(OH)OEt. Of these enols, only the latter is reactive
toward nucleophylic addition of water, in the rate-
determining step, to form the dipolar ion CH3C(O2)v
CHC(OH)2O1(H)Et, whose breakdown losing EtOH leads
to the products.
2.1. Study of the products formed in the ammonolysis
reaction
In order to determine the nature of the organic product or
products formed in the ammonolysis reaction, we treated
CH3COCH2CO2Me with excess aqueous 15% NH3 at
room temperature. After about 20 min abundant white
crystals had formed whose analysis showed that they corre-
sponded to methyl b-aminocrotonate 6a (Scheme 1). After
initial formation of the white crystals, the mixture was left
standing at room temperature. After 4 weeks, the reaction
mixture had turned into a yellow aqueous solution. After
water and NH3 were taken off in vacuum, a brown syrupy
solid formed, which was washed with methanol. White
crystals were obtained after the washings. This white solid
was identi®ed as 1,5-dimethyl-2,6,9-triaza-bicyclo[3.3.1]-
In this work, we have investigated the ammonolysis in
water of ethyl and methyl acetoacetates and have
found that an enol mechanism, analogous to the one
proposed for the alkaline hydrolysis of ethyl acetoacetate,
explains the various experimental facts, characteristic of
the reaction.
1
nonane-3,7-dione 10 on the basis of its IR, H, 13C NMR
and mass spectra, X-ray crystallographic analysis and
con®rmed with the spectroscopic and analytical data
reported by Shim and co-workers.2 From the reaction
mixture, we isolated another solid. This solid corresponded
to a pure compound that appears to be another heterocycle
different from 10, 12 and 14 (Schemes 2and 3). We are at
present attempting to determine its structure.
Mechanistically, we believe that, 6a formed as shown in
Scheme 1 and that conversion of 6a into heterocycle 10
took place via intermediates 5a, 7a, 8a and 9 (Scheme 1).
For the conversion of methyl acetoacetate into aceto-
Keywords: acetoacetates; ammonolysis; enol mechanism; b-amino-
crotonates; heterocycles.
Corresponding author. Fax: 157-2-3392440;
e-mail: abonia@quimica.univalle.edu.co
p
0040±4020/02/$ - see front matter q 2002 Published by Elsevier Science Ltd.
PII: S0040-4020(01)01127-9

56
R. Paredes et al. / Tetrahedron 58(2002) 55±60
Â
acetamide, Garcõa and co-workers treated the ester with 2%
ammonia in dioxane in the presence of Candida antarctica
lipase.3 These authors found that enamino esters such as 6
(Scheme 1) are not adequate substrates for the enzymatic
reaction, since, when they were treated with excess NH3 in
the presence of the enzyme, no reaction took place. In this
case, the lipase-catalyzed ammonolysis took a different path
than in our reaction. Presumably the ammonolysis took
place only at the ester group.
2.2. Kinetic study of the ammonolysis reaction
Ammonolysis reactions were carried out by mixing equal
volumes of aqueous NH3 and CH3COCH2CO2Et where a
large excess of ester was used. Each reaction was monitored
by measuring the decreasing NH3 and enolate concentra-
tions at time intervals. The NH3 concentrations were
measured by a potentiometric method and the enolate
concentrations were measured by UV spectrometry.
The following facts were observed during the course of the
reactions:
At the start a very fast and substantial drop in the NH3
concentration occurred. As seen in Table 1, for one of the
experiments, as soon as the NH3 and ester solutions were
mixed (time 0.0), the NH3 concentration dropped from
2.32£1024 to 2.5£1025 M (89% disappearance of NH3).
Initially, we thought that the disappearance of NH3 was
due to its very fast conversion into acetoacetamide.
However, work up of the reaction mixture at the start,
showed that no product had formed at all and that the
only organic compound present was unreacted ethyl aceto-
acetate 1b. Another possibility for a fast consumption of
NH3 is via a rapid acid±base reaction between the ester
and NH3 (Scheme 4).
For this reaction, its equilibrium constant can be evaluated
Scheme 1. Mechanistic sequence proposed for the ammonolysis in water of
alkyl acetoacetates 1a,b at room temperature, leading to the formation of
heterocycle 10.
Scheme 2. Stepwise formation of heterocycle 10 via proposed intermediate 12.

R. Paredes et al. / Tetrahedron 58(2002) 55±60
57
as follows:
hypothetical equilibrium constant is 0.087, a value 2.4
times larger than the expected 0.036. This indicated that
the sudden initial decrease in NH3 concentration possibly
occurred because the ester and NH3 reacted very fast to form
imine 4b (Scheme 1). This view is reinforced by the facts
described below.
À
Á
‰enolateŠ NH14
KaKb
KW
ˆ K ˆ
ˆ 0:036
‰esterŠ‰NH3Š
where Ka is the acidity constant for CH3COCH2CO2Et,
pKaˆ10.74, Kb the basicity constant for NH3ˆ1.8£1025
and Kw is the water constantˆ1.0£10214
.
A plot of ln[NH3] versus time for the experiment described
in Table 1 showed that after the initial fast drop in NH3
concentration, a sloping straight line was obtained from
time 10 to 30 as expected from the pseudo-®rst-order reac-
tion. Furthermore, when for each point at times 10, 15, 20,
25 and 30, the pKa for CH3COCH2CO2Et was calculated
from the experimental data, a value of 10.7 was obtained
(Table 2). This result indicated that from time 10 to 30, the
equilibrium between ester and enolate had been established.
Between time 0 and 10, the system moved toward the estab-
lishment of the ester±enolate equilibrium (Scheme 1), as
seen from the pKa values in Table 2.
Assuming that the above equilibrium was established at
once as soon as the reagents were mixed, the calculated
Three reactions between excess CH3COCH2CO2Et and NH3
in water were run. In all these reactions, we observed an
initial fast drop in NH3 concentration as seen from their
plots of ln[NH3] versus time. After about time 10, straight
lines were obtained whose slopes correspond to the kobs
values measured (Table 3). The results shown in Table 3
Table 3. Ammonolysis in water of CH3COCH2CO2Et at room temperature
[Ester]init.
[NH3]init.
kobs (min21
)
0.100
0.050
0.020
3.5£1023
6.9£1024
2.3£1024
0.024
0.022
0.025
Scheme 3. Formation of heterocycle 10 according to Shim and co-workers.
Table 1. Ammonolysis in water of 0.0200 M CH3COCH2CO2Et with
2.32£1024 M NH3 at room temperature
indicate that for these reactions the rate law is ®rst order in
NH3 and zero order in ester. To explain this rate law, we
considered that the rate of formation of the dipolar ion 8b is:
Time (min)
[NH3] (M)
[Enolate] (M)
0.0
2.5
5.0
10.0
15.0
20.0
25.0
30.0
2.5£1025
1.9£1025
1.7£1025
1.4£1025
1.2£1025
1.1£1025
1.0£1025
8.7£1026
1.0£1024
1.0£1024
9.9£1025
9.4£1025
8.9£1025
8.5£1025
8.1£1025
7.7£1025
rate ˆ k5‰7bŠ‰NH3Š
Since 7b is a reactive, unstable species whose concentration
at all times after time 10 is very small and approximately
constant, we have:
rate ˆ k5‰7bŠ‰NH3Š ˆ kobs‰NH3Š
This last equation shows that the kinetics for the reaction is
®rst-order in NH3 and zero order in ester as determined
experimentally.
2.3. Mechanism for the formation of the heterocycle
Scheme 4.
Although Shim and co-workers2 prepared heterocycle 10 via
Table 2. Ammonolysis in water of 0.0200 M CH3COCH2CO2Et with 2.32£1024 M NH3 at room temperature (pKaˆ10.74, for each point the charge balance
[enolate]1[OH2]ˆ[NH41]1[H1] is maintained)
Time (min)
[Enolate] (M)
[NH3] (M)
[OH2] (M)
[NH41] (M)
[H1] (M)
Calculated pKa
0.0
5.0
1.0£1024
9.9£1025
9.4£1025
8.9£1025
8.5£1025
8.1£1025
7.7£1025
2.5£1025
1.7£1025
1.4£1025
1.2£1025
1.1£1025
1.0£1025
8.7£1026
4.3£1026
3.0£1026
2.6£1026
2.4£1026
2.3£1026
2.2£1026
2.0£1026
1.0£1024
1.0£1024
9.7£1025
9.1£1025
8.8£1025
8.3£1025
7.9£1025
2.5£1029
3.3£1029
3.8£1029
4.2£1029
4.4£1029
4.6£1029
5.0£1029
10.9
10.8
10.7
10.7
10.7
10.7
10.7
10.0
15.0
20.0
25.0
30.0

58
R. Paredes et al. / Tetrahedron 58(2002) 55±60
Table 4. Other crystal and re®nement data of compound 10
2.4. X-Ray crystallographic analysis for compound 10
Diffractometer
Crystal dimensions (mm3)
Scan technique
Data collected
Enraf-Nonius CAD4
0.2£0.2£0.15
Single crystal of C8H13N3O2, are at 123 K, monoclinic,
Ê
space group P21/n (No. 14), with aˆ8.3070(2) A,
v/2u
(0, 214, 213) to (10, 14, 12)
Ê
Ê
Ê
bˆ10.7160(3) A, cˆ10.0490(3) A, bˆ101.507(2) A, Vˆ
876.56(4) A , Zˆ4, Dcˆ1.388 g/cm3, m(Mo Ka)ˆ
Ê 3
Unique data
2198
1497
Observed re¯ections
Fo.4(s(Fo))
Decay
Rint (%)
R
wR20.12
Maximum shift/error
Computing structure solution
Computing structure re®nement
0.102mm 21. The structure was re®ned on F2. A summary
of crystal and re®nement data is given in Table 4. The cell
dimensions were re®ned by least-squares ®tting the u values
of 25 re¯ections with u range of 1.0±27.888. The ring H
atoms were added at calculated positions. The H atoms
treated with a riding model with SHELXL97 defaults,5
,1%
0.03
0.045
0.0
SHELXS-976
SHELXL-975
Ê
Ê
(C±H, 0.96±0.97 A and N±H, 0.86 A) were not re®ned
(Tables 5 and 6, Fig. 1).
Table 5. Fractional atomic coordinates and equivalent isotropic displace-
Ê 2
ment parameters (A )
x
y
z
Ueq
O12
O13
N9
0.07890(14)
0.00643(14)
0.05608(16)
0.0674(2)
0.20475(12)
0.28989(11)
0.52440(14)
0.29677(16)
20.00249(12)
20.53806(11)
20.21399(13)
20.07905(15)
20.12164(13)
20.21679(15)
20.36059(16)
20.42835(15)
20.36964(13)
20.24386(15)
20.12363(16)
20.26840(17)
20.16980(17)
0.0248(3)
0.0204(3)
0.0167(3)
0.0177(4)
0.0183(3)
0.0166(4)
0.0185(4)
0.0171(4)
0.0171(3)
0.0159(4)
0.0174(4)
0.0208(4)
0.0220(4)
C3
N20.19874(16) 81(01.344)2
C1
C8
C7
N6
C5
C4
C10
C11
0.19969(19)
0.1933(2)
0.0257(2)
20.10078(17)
20.09361(19)
20.0934(2)
20.2427(2)
0.3551(2)
0.44696(16)
0.39495(16)
0.34841(16)
0.37524(13)
0.45112(15)
0.36374(16)
0.53608(16)
0.52299(17)
intermediate 14 by thermolysis of b-aminocrotonamide as
shown in Scheme 3, we believe that under our mild reaction
conditions heterocycle 10 is formed via intermediate 12,
from a double Michael addition of NH3 on 12, as shown
in Scheme 2. Under our reaction conditions, we did not
detect intermediate 14, while Shim separated it as a stable
compound from the reaction mixture.
Figure 1. Ortep-3 view7 of compound 10 with the atom labeling. Thermal
ellipsoids are shown at 50% probability, except for H atoms which are
represented by spheres of arbitrary size.
Table 6. Geometric parameters of compound 10
O12±C3
N9±C1
N2±C1
C8±C7
C5±C10
1.243(2)
1.458(2)
1.471(2)
1.508(2)
1.517(2)
O13±C7
C3±N2
C1±C11
C7±N6
C5±C4
1.2506(19)
1.343(2)
1.520(2)
1.335(2)
1.528(2)
N9±C5
C3±C4
C1±C8
N6±C5
1.451(2)
1.504(2)
1.540(2)
1.494(2)
C5±N9±C1
N2±C3±C4
N9±C1±C11
N2±C1±C8
O13±C7±N6
C7±N6±C5
N6±C5±C10
C10±C5±C4
111.03(13)
117.37(14)
109.67(14)
109.38(14)
121.62(15)
126.10(14)
107.19(12)
111.60(13)
O12±C3±N2
C3±N2±C1
N2±C1±C11
C11±C1±C8
O13±C7±C8
N9±C5±N6
N9±C5±C4
C3±C4±C5
121.44(16)
126.58(14)
108.61(13)
110.71(13)
120.54(14)
110.95(12)
107.60(12)
112.88(13)
O12±C3±C4
N9±C1±N2
N9±C1±C8
C7±C8±C1
N6±C7±C8
N9±C5±C10
N6±C5±C4
121.16(15)
107.99(12)
110.42(12)
113.55(13)
117.81(14)
110.32(13)
109.20(13)
O12±C3±N2±C1
C5±N9±C1±N2
C5±N9±C1±C8
C3±N2±C1±C11
N9±C1±C8±C7
C11±C1±C8±C7
C1±C8±C7±N6
C8±C7±N6±C5
C1±N9±C5±C10
176.96(15)
255.51(15)
64.03(16)
142.22(16)
243.04(19)
2164.69(14)
10.9(2)
C4±C3±N2±C1
C5±N9±C1±C11
C3±N2±C1±N9
C3±N2±C1±C8
N2±C1±C8±C7
C1±C8±C7±O13
O13±C7±N6±C5
C1±N9±C5±N6
C1±N9±C5±C4
25.2(2)
2173.70(12)
23.4(2)
296.84(18)
75.66(17)
2171.19(15)
2175.93(15)
250.06(16)
69.35(15)
N6±C5±C4±C3
C10±C5±C4±C3
N2±C3±C4±C5
N9±C5±C4±C3
C7±N6±C5±C4
O12±C3±C4±C5
C7±N6±C5±N9
C7±N6±C5±C10
73.14(16)
2168.53(13)
16.9(2)
247.38(17)
2101.09(18)
2165.22(14)
17.3(2)
2.0(2)
2168.69(12)
137.85(16)

R. Paredes et al. / Tetrahedron 58(2002) 55±60
59
3. Conclusions
4.4. Determination of enolate concentrations in the
ammonolysis of CH3COCH2CO2Et in water at room
temperature
Results from our work show that ammonolysis in water of
alkyl acetoacetates at room temperature does not lead to the
formation of acetoacetamide as could be expected. Instead
alkyl b-aminocrotonates form as kinetically controlled
products which then convert slowly into a mixture of at
least two heterocycles of which the main product is
1,5-dimethyl-2,6,9-triaza-bicyclo[3.3.1]nonane-3,7-dione.
Ammonolysis of methyl acetoacetate catalyzed by C.
antarctica lipase lead to the formation of acetoacetamide.
After mixing the aqueous NH3 and ester solutions,
absorbance readings were taken at time intervals during
30 or 60 min. Enolate concentrations were calculated as
shown:
A
‰enolateŠ ˆ
1max
4.5. Calibration curve of ln[NH3] versus potential
4. Experimental
4.1. Materials
Several aqueous NH3 standard solutions of appropriate
concentrations were prepared. The NH3 concentration of
each solution was determined using the Orion ion analyzer.
For each solution, its potential (mV) was measured. A plot
of ln[NH3] versus potential gave a straight line with
intercept 215.54 and slope 20.0498.
All chemicals were of the highest possible purity available.
Solutions for the kinetic studies were prepared daily in
distilled water.
4.6. Determination of NH3 concentrations in the
ammonolysis of CH3COCH2CO2Et in water at room
temperature
4.2. Instruments
A UV±visible Shimadzu UV-160A spectrophotometer was
used for the spectrometric measurements. An Orion AE940
ion analyzer equipped with a NH3 electrode and a reference
electrode was used to measure NH3 concentrations in the
After mixing the aqueous NH3 and ester solutions, potential
readings were taken at time intervals during 30 or 60 min.
NH3 concentrations were determined from the calibration
curve.
È
standard solutions. A digital Schott±Gerate CG 820 pH
4.7. Determination of NH41and OH2 concentrations in
the ammonolysis of CH3COCH2CO2Et in water at room
temperature
È
meter equipped with a Schott±Gerate H61 combined elec-
trode was used for the potentiometric measurements. The 1H
and 13C NMR spectra were recorded on a Varian Gemini
200 instrument, chemical shifts (d) are reported in ppm
relative to internal tetramethylsilane, DMSO-d6 and D2O
as solvent. Silica gel plates (Merck F254) were used for
analytical TLC. The IR spectra were taken in a FT-IR
ATI MATTSON spectrophotometer. The mass spectrum
was taken on a Varian Model MAT MS-311 spectrometer
at 70 eV.
NH41originates from the reaction of CH3COCH2CO2Et
with NH3 and from the reaction of NH3 with water. [OH2]
and [NH41] for each point were calculated from the follow-
ing equation:
Â
Ã
NH14 ‰OH2Š
{‰enolateŠ 1 ‰OH2Š}‰OH2Š
‰NH3Š
ˆ
ˆ Kb
‰NH3Š
ˆ 1:8 £ 1025
4.3. Determination of lmax and 1max for the enolate of
CH3COCH2CO2Et in water at room temperature
4.7.1. Determination of the products formed in the
ammonolysis of CH3COCH2CO2Et in water at room
temperature. 8.56 g (0.0738 mol) of CH3COCH2CO2Me
was added to 50 mL 15% aqueous NH3. After 20 min, an
abundant white solid had formed. The solid was ®ltered off,
washed with water and dried at room temperature. It
amounted to 7.94 g of a product identi®ed as methyl
b-aminocrotonate 6a (94% yield, mp 81.2±838C) on
comparison with an authentic sample of the compound
obtained from Aldrich.9
The enolate ion of 1,3-dicarbonyl compounds absorb
strongly in the UV in the 270±300 nm region.8 An aqueous
0.250 M KOH and 2.28£1025 M CH3COCH2CO2Et
solution displays maximum absorption at 273 nm due to
the complete conversion of the ester into enolate ion, as
calculated from the pKaˆ10.7 for the ester.4
The formation of the enolate was accomplished by mixing
50.0 mL of aqueous 0.500 M KOH with 50.0 mL of aqueous
4.55£1025 M CH3COCH2CO2Et. Aliquots were poured into
1 cm UV cells and readings of absorbance A at 273 nm were
taken every 5 min during 100 min. A plot of ln A versus
time gave a straight line with intercept 20.693 and slope
20.00334.
10.53 g (0.0908 mol) of CH3COCH2CO2Me was added to
50 mL 15% aqueous NH3. After formation of the white
solid, the mixture was left standing at room temperature.
TLC showed that methyl b-aminocrotonate was converting
into products. After 4 weeks, the conversion was complete
and the reaction mixture was a yellow solution. After water
and NH3 were taken off in vacuum, a brown syrupy solid
A0
e20:693
2:28 £ 1025
1max
ˆ
ˆ
ˆ 2:19 £ 104
‰enolateŠ
0

60
R. Paredes et al. / Tetrahedron 58(2002) 55±60
formed which was washed with methanol. 1.93 g of white
crystals were obtained after the washings and corresponded
to 1,5-dimethyl-2,6,9-triaza-bicyclo[3.3.1]nonane-3,7-dione
10. 23% isolated yield, mp, 198.0±198.28C [197±198]; IR
(KBr, cm21): 3282 [3280], 3198 [3179], 1662 [1660], 1508
[1510], 1396 [1400]; 1H NMR (DMSO-d6): d 1.28 (s, 6H, 1-
and 5-CH3) [1.28 (s, 6H)], 2.19 (s, 4H, 4- and 8-CH2) [2.15
(s, 4H)], 3.55 (s, 1H, 9-NH) [3.45 (s, 1H)], 8.01 (s, 2H, 2-
and 6-NH) [7.89 (s, 2H)]; 1H NMR (D2O): d 1.39 (s, 6H, 1-
and 5-CH3) [1.52(s, 6H)], 2.42(s, 4H, 4- and 8-CH 2) [2.53
(s, 4H)]; 13C NMR (DMDO-d6): d 28.1 (1- and 5-CH3), 44.5
(4- and 8-C), 65.7 (1- and 5-C), 168.0 (3-and 7-CO); 13C
NMR (D2O): d 26.2, 42.4, 65.7, 170.1; MS m/z (%): 184
(96, M11), 183 (100, M1) [183], 166 (8, M2NH3). Signals
in square brackets correspond to those reported by Shim and
co-workers.2 The overall yield of 10 is higher than 23%
since we detected more of this product mixed with other
products in the methanolic washings.
Organic Chemistry, University of Mainz (Germany) and
DAAD for the ®nancial support of her Doctoral program.
References
1. Paredes, R.; Rios, L.; Ocampo, R. Bol. Soc. Chil. Quim. 1991,
36, 195±201.
2. Shim, S. C.; Kim, D. W.; Moon, S. S.; Chae, Y. B. J. Org.
Chem. 1984, 49, 1449±1450.
Â
3. Garcõa, M. J.; Rebolledo, F.; Gotor, V. Tetrahedron 1994, 50,
6935±6940.
4. Cram, D. J. Fundamentals of Carbanion Chemistry; Academic:
New York, 1965; pp 8±20.
5. Sheldrick, G. M. SHELXL97: Program for the Re®nement of
È
Crystal Structures; University of Gottingen: Germany, 1997.
6. Sheldrick, G. M. Acta Crystallogr. 1990, A46, 467±473.
7. Farrugia, L. S. J. Appl. Crystallogr. 1997, 30, 565.
8. Silverstein, R. M.; Bassler, G. C.; Morrill, T. C. Spectrometric
Identi®cation of Organic Compounds; Wiley: New York, 1991
p 301.
9. Catalog Handbook of Fine Chemicals; Aldrich Chemical
Company, Inc., 2000±2001; p 1086.
Acknowledgements
We are grateful to Universidad del Valle for the support
given to this work. A. Hormaza thanks the Institute of