Amide and lactam hydrolysis of N-(2-hydroxyacetyl)-2-pyrrolidone: Effective catalysis

Article abstract of DOI:10.1002/poc.1067

When N-(2-hydroxyacetyl)-2-pyrrolidone (open form) is dissolved in water at pH > 8, irreversible cleavage of the exocyclic and endocyclic amide C - N bond occurs. The latter rupture corresponds to the lactam opening yielding N-(4-hydroxyacetyl)butanoic acid (NBA). NBA is produced from the ester hydrolysis of the esteramide macrocycle that is in equilibrium with the cyclol form of the open form. We have previously reported this latter equilibrium for N-(2-aminoacetyl)-2-lactams. 2-pyrrolidone (lactam) and glycolic acid are produced from direct hydrolysis of the open form by means of the amide exocyclic cleavage. The [NBA]/[lactam] ratio increases at higher pH since the NBA production is second order with respect to [OH^-] while the corresponding lactam formation is only first order. The obtained kobs is hence the sum of the rate constants that yield lactam and NBA, respectively. This kobs is uncatalyzed and specific base catalyzed with unusually high rate constants of 2.1 × 10^-6s^-1 and 0.025 M^-1 s ^-1, respectively. The stability of the corresponding tetrahedral intermediate formed and the intramolecular alkoxy nucleophilic attack on the lactam carbonyl group combined with an effective protonation of the lactam nitrogen that promotes the C - N cleavage, contribute to increase the reaction rates and lactam opening. Rate constants for the two parallel reactions are obtained from kobs and [NBA]/[lactam] versus pH plots. Copyright

Full text of DOI:10.1002/poc.1067

JOURNAL OF PHYSICAL ORGANIC CHEMISTRY
J. Phys. Org. Chem. 2006; 19: 737–743
Published online 6 November 2006 in Wiley InterScience (www.interscience.wiley.com). DOI: 10.1002/poc.1067
Amide and lactam hydrolysis of
N-(2-hydroxyacetyl)-2-pyrrolidone: effective catalysisy
*
Lisaedy Garc´ıa Borboa and Oswaldo Nu´ n˜ ez
´
´
´
´
´
Laboratorio de Fisicoquımica organica y Quımica ambiental, Departamento de Procesos y Sistemas, Universidad Simon Bolıvar,
Apartado postal 89000, Caracas, Venezuela
Received 7 November 2005; revised 13 February 2006; accepted 15 February 2006
ABSTRACT: When N-(2-hydroxyacetyl)-2-pyrrolidone (open form) is dissolved in water at pH > 8, irreversible
cleavage of the exocyclic and endocyclic amide C—N bond occurs. The latter rupture corresponds to the lactam
opening yielding N-(4-hydroxyacetyl)butanoic acid (NBA). NBA is produced from the ester hydrolysis of the ester-
amide macrocycle that is in equilibrium with the cyclol form of the open form. We have previously reported this latter
equilibrium for N-(2-aminoacetyl)-2-lactams. 2-pyrrolidone (lactam) and glycolic acid are produced from direct
hydrolysis of the open form by means of the amide exocyclic cleavage. The [NBA]/[lactam] ratio increases at higher
pH since the NBA production is second order with respect to [OHꢀ] while the corresponding lactam formation is only
first order. The obtained kobs is hence the sum of the rate constants that yield lactam and NBA, respectively. This
kobs is uncatalyzed and specific base catalyzed with unusually high rate constants of 2.1 ꢁ 10ꢀ6 sꢀ1 and
0.025 Mꢀ1 sꢀ1, respectively. The stability of the corresponding tetrahedral intermediate formed and the intramo-
lecular alkoxy nucleophilic attack on the lactam carbonyl group combined with an effective protonation of the
lactam nitrogen that promotes the C—N cleavage, contribute to increase the reaction rates and lactam opening.
Rate constants for the two parallel reactions are obtained from kobs and [NBA]/[lactam] versus pH plots.
Copyright # 2006 John Wiley & Sons, Ltd.
KEYWORDS: hydrolysis; catalysis; lactam; amide; stable tetrahedral intermediate; cyclol; macrocycle; open form
INTRODUCTION
N-(4-hydroxyacetyl)butanoic acid (NBA) (4). This pro-
duct corresponds to a lactam ring opening, a process of
wide interest in bioorganic chemistry. NBA is a derivative
of g-aminobutyric acid (GABA), an inhibitory neuro-
transmitter in the brain. GABA induces relaxation,
analgesia, and sleep. It is produced in the brain from
glutamate but is also available as a supplement pill.
However, as a supplement it is not always an effective way
to raise brain levels since it cannot easily cross the blood-
brain barrier. Therefore, the use of a derivative such as g-
aminobutyric lactam (1) or even NBA (4) becomes of
medical interest. On the other hand, lactam hydrolysis
reactivity is an important topic in the design of antibiotics
since it is well known that b-lactam antibiotics act by
acylating a serine hydroxyl group in the catalytic center of
bacterial protease. In fact, it has been found4, that 2-
pyrrolidone hydrolyzes with a specific base second-order
rate constant of 5.59 ꢁ 10ꢀ6 Mꢀ1 sꢀ1 similar to N-methyl
acetamide (3.32 ꢁ 10ꢀ6 Mꢀ1 sꢀ1), two orders of magni-
tude slower than b-propiolactam (2.37 ꢁ 10ꢀ4 Mꢀ1 sꢀ1).
This stability of 2-pyrrolidone towards ring opening
eliminated it as a potential antibiotic.
We have been interested in measuring intrinsic rates of
intramolecular transformations in stable tetrahedral
intermediates for some time. For example, we have
reported1,2 results on transannular intrinsic rates in
bislactam macrocycles formed from N-(2-aminoacetyl)-
2-lactams. We have used 2-pyrrolidone, 2-piperidone, and
2-caprolactam as the corresponding five-, six-, and seven-
membered lactams. We have also reported3 the rate of
formation of the corresponding stable tetrahedral inter-
mediates of N-heteroethylphtalimide (hetero: hydroxy,
amino, and thioxy) and their rates of rupture to
diacylimides. Subsequently, we synthesized N-(2-hydro-
xyacetyl)-2-pyrrolidone (3) expecting to observe similar
chemical behavior to that observed in the N-(2-aminoa-
cetyl)-2-lactams. However, somewhat unexpectedly, we
found at pD > 7.5 an irreversible cleavage of the exocyclic
and endocyclic C—N bond of 3 to yield, in the latter case,
´
´
*Correspondence to: O. Nun˜ez, Laboratorio de Fisicoquımica organica
y Quımica ambiental, Departamento de Procesos y Sistemas, Univer-
´
´
Peptide bonds hydrolyze with life-times of ca. 7 years
under mild conditions5 at 25 8C. Therefore, considerable
effort has been focused on the design of a catalyst that
hydrolyzes the amide bond under mild conditions using
metal-promoted6 artificial proteases. As non-metal
´
´
sidad Simon Bolıvar, Apartado postal 89000, Caracas, Venezuela.
E-mail: onunez@usb.ve
Contract/grant sponsor: Simon Bolivar University.
ySelected article presented at the Eighth Latin American Conference on
Physical Organic Chemistry (CLAFQO-8), 9–14 October 2005,
Florianopolis, Brazil.
Copyright # 2006 John Wiley & Sons, Ltd.
J. Phys. Org. Chem. 2006; 19: 737–743

´
˜
L. G. BORBOA AND O. NUNEZ
738
catalysts, 4-heterocyclohexanones have been used5 to
promote intramolecular alkoxy attack in the hemithioa-
cetal adduct formed by a-thiolacetamides nuclophilic
attack, a reaction that mimics the hydrolysis of peptides
by serine proteases. When tetrahydropyranose is used as
catalyst and p-trifluoromethyl a-thiolacetanilide as
substrate with [NaOH] ¼ 0.2–2 M, a third-order rate
constant of ca. 2 ꢁ 10ꢀ3 Mꢀ2 sꢀ1 is found. Therefore,
second order rate constants of ca. 10ꢀ3 Mꢀ1 sꢀ1 have been
reported as an important catalytic reaction acceleration.
HIV-1 protease7 hydrolyzes viral polyproteins into
functional protein products that are essential for viral
assembly and further activity, therefore, the effort to find
HIV-1 protease is an important medical target. In fact, it
has recently been pointed out8 that in the dimer aspartyl
protease active site there is formation of a tetrahedral-
transition state intermediate in the hydrolysis of the
peptides used as inhibitor. Therefore, the use of a
reversible macrocycle peptide, such as the one produced
in this work, may be of interest as an HIV-1 inhibitor.
In this work, we report the efficient exocyclic carbonyl
hydrolysis of N-(2-hydroxyacetyl)-2-pyrrolidone (3) to
produce the two products 2-pyrrolidone (lactam) (1) and
glycolic acid (2). NBA (4) results from the endocyclic
hydrolysis of the carbonyl group of 3. (Scheme 1). The
rates and equilibrium constants involved in this scheme
have been measured and/or estimated. Furthermore, we
attempt to explain the importance of the processes
involved and of some of the products obtained.
a-Benzylacetic acid
Six grams (0.15 mol) of potassium was added to a solution
of 17 mL (0.16 mol) of benzylic alcohol in 200 mL of
toluene. The final solution was refluxed for 4 h. To this
solution, 0.8 g (0.07 mol) of a-chloroacetic acid in 50 mL
of benzene was added gradually while stirring. The final
mixture was refluxed for 16 h. After the reaction workup,
a yellow liquid was obtained. Yield: 6.1 g (51%). NMR
(1H, CDCl3): d: 4.10 (s,2H); 4.6 (s,2H); 7.30 (s,5H);
9.5 (s,1H).
a-Benzylacetic acid chloride
Four grams (0.024 mol) of a-benzylacetic acid was
dissolved in 25 mL of dry toluene. To this solution 3.5 mL
(0.048 mol) of thionyl chloride was added. After addition,
the solution was heated at 70–758C for 2 h. The thionyl
chloride was removed by distillation to finally obtain a
toluene solution of the target chloride. NMR (1H,CDCl3):
d: 4.49 (s,2H); 4.74 (s,2H); 7.30 (s,5H).
N-Benzylacetil-2-pyrrolidone
A solution of 4.1 g (0.048 mol) of 2-pyrrolidone and 3.8 g
(0.0048 mol) of pyridine in 10 mL of dry toluene was kept
at 08C in an ice bath. To this solution, was added 4.4 g
(0.024 mol) of a-benzylacetic acid chloride while
retaining the temperature at 08C. The reaction was kept
at this temperature for 2 h. After that time, the reaction
mixture was refluxed for another 2.5 h. A CaCl2 trap was
used during the whole reaction. After refluxing, the
reaction mixture was cooled to 08C and mixed with
150 mL of water-ice at 08C. The organic phase was
EXPERIMENTAL
N-(2-hydroxyacetyl)-2-pyrrolidone (3) was synthesized
from a-chloroacetic acid, benzylic alcohol, and
2-pyrrolidone (1) in four steps9.
Scheme 1. Competitive hydrolysis pathways for production of the pair 2-pyrolidone (1) and glycolic acid (2) from exocyclic
carbonyl hydrolysis of N-(2-hydroxyacetyl)-2-pyrrolidone (3) and NBA (4) from its endocyclic hydrolysis
Copyright # 2006 John Wiley & Sons, Ltd.
J. Phys. Org. Chem. 2006; 19: 737–743

AMIDE AND LACTAM HYDROLYSIS OF N-(2-HYDROXYACETYL)-2-PYRROLIDONE
739
separated and washed with a saturated solution of
NaHCO3. The organic phase was then dried with
anhydrous Na2SO4. The solution was filtered and excess
toluene evaporated under reduced pressure at 608C. A
solid was obtained. mp: 94–968C; yield: 3.4 g (61%).
NMR(1H,CDCl3): d: 2.10 (m,2H); 2.50 (t,2H); 3.80
(t,2H); 4.60 (s,4H); 7.30 (m,5H). Mass spectrum: m/z: 233
(Mþ), 91 (M-142); 127 (M-106).
2 < pD < 6, KH2PO4; 6 < pD < 10, K2HPO4; and
10 < pD < 13, K3PO4. Buffer concentrations of 0.3,
0.5, 0.7, and 1.0 M and a [KCl] ¼ 1 M were used. pD
values were obtained directly from pH measurements by
using the relation10: pD ¼ pH þ 0.40. The following
values of pD were used to evaluate the equilibrium and
hydrolysis rate: 0.4, 1.7, 4.9, 6.9, 7.4, 8.3, 10.5, 11.2, 11.8,
12.4, and 13.0. Equilibrium constants were evaluated in
triplicate; a standard deviation <10% was obtained.
N-(2-Hydroxyacetyl)-2-pyrrolidone (3)
Rate constant measurements
0.23 g (0.98 mmol) of N-benzylacetil-2-pyrrolidone was
dissolved in 30 mL of ethyl acetate (HPLC grade). To this
solution 0.43 g of Pd on Alumina (5%) catalyst in 35 mL
of ethyl acetate was added. The mixture was reduced
under a hydrogen atmosphere at 35 psi during 3 h. After
the reduction, the mixture was filtered and the resultant
liquid concentrated under reduced pressure. Awhite solid
was obtained. mp: 62–638C (lit10: 64–658C); yield: 0.12 g
(80%). NMR (1H,CDCl3): d:2.10 (m,2H); 2.60 (t,2H);
3.30 (s,1H); 3.85 (t,2H); 4.70 (s,2H). NMR(1H,D2O): d:
2.14 (m,2H); 2.65 (t,2H); 3.82 (t,2H); 4.70 (s,2H). Mass
spectrum: m/z: 143 (Mþ); 86 (M-57); 113 (M-30). IR
Kinetics were measured by following the disappearance
of the exocyclic methylene NMR (1H) signal of
compound 3 at ca. 4.6 ppm. Buffer phosphate (0.7 mL)
in D2O was used to dissolve compound 3 to yield a final
concentration of 0.25 M in the NMR tube. Two kinds of
experiments were conducted. One, keeping the buffer
concentration constant at 0.5 M at pD: 7.3, 8.7, 10.8, 11.3,
12.7, and 13 and the second keeping pD constant at 8.7
and 12.7 and changing the buffer concentration to 0.3, 0.7,
and 1.0 M. In all cases, a 400 MHz NMR, JEOL Eclipse
Plus instrument, was used. Rate constants were obtained
from the slope of a plot of ln (ItꢀIinf) versus t, where I is
the NMR integral of the signal. Rate constants were
measured in triplicate. The errors between measurements
did not exceed 5%.
(KBr): 3528 cmꢀ1 (OH), 1732 cmꢀ1 (C O lactam),
—
—
1699 cmꢀ1 (C O amide). UV (H O): 226 nm.
—
—
2
Methyl ester of NBA
0.22 g (0.94 mol) of N-benzylacetyl-2-pirrolidone was
dissolved in 30 mL of methanol (HPLC grade). To this
solution 0.46 g of Pd-Alumina (5%) catalyst dissolved
in 30 mL of methanol was added. The mixture was
reduced under a H2 atmosphere at 30 psi. After 3 h,
the reaction was stopped and the mixture filtered. The
resultant liquid was concentrated under reduced
pressure at 408C. The CGMS spectrum of the liquid
showed three peaks corresponding to toluene, N-(2-
hydroxyacetyl)-2-pyrrolidone (3) and the methyl ester
of N-(a-hydroxyacetyl)-g-aminobutyric acid. Mass
spectrum: 175 (Mþ); 146 (M-29); 104 (M-71); 91
(M-84); 65 (M-110). NMR(1H,CDCl3): d: 2.00 (m,2H);
2.45 (t,2H); 3.25 (t,2H); 3.80 (s,3H); 4.16 (s,2H). This
compound was used to identify one of the hydrolysis
products of 3, the NBA (4).
Identification of hydrolysis products
The lactam 2-pyrrolidone (1) and glycolic acid produced
during the reaction were identified by adding these
commercially available compounds to the reaction
mixture and identifying the signals whose intensity
increased. Signals of NBA (4) were identified by
comparison with the NMR signals of the methyl ester
of NBA that was synthesized in this work. Cyclol 5 and
macrocycle 6 NMR signals were identified by changing
the equilibrium toward the N-(2-hydroxyacetyl)-2-pyrro-
lidone (3) by decreasing the pH at ca. 2 of the reaction at
pH ca. 12 and observing the disappearance of these
signals. The cyclol and macrocycle signals were then
assigned by analogy with similar systems previously
identified1,2 in our laboratory. Magnetization transfer
experiments were also used to identify the corresponding
methylene signals. In Table 1, the NMR signals for the
reactants and products are shown. In Fig. 1, NMR spectra
at pD ¼ 3.0 and pD ¼ 12.7 are shown.
Sample preparation
1H NMR, was used to follow the hydrolysis of compound
3 at different pH values. The samples for each experiment
were prepared by dissolving 25 mg of compound 3 in
0.7 mL of D2O or in a solution of phosphate buffer in
D2O. NMR tubes of 5 mm were used to record the NMR
spectra. According to the required pH, the following
buffer solutions were used: 0 < pD < 2, H3PO4;
DISCUSSION
The product pair 2-pyrrolidone (lactam) (1) and glycolic
acid (2) is formed in the hydrolysis of N-(2-hydro-
xyacetyl)-2-pyrrolidone (3) via reaction of its exocyclic
Copyright # 2006 John Wiley & Sons, Ltd.
J. Phys. Org. Chem. 2006; 19: 737–743

´
˜
L. G. BORBOA AND O. NUNEZ
740
Table 1. 1H NMR chemical shift of reactants and products
1HNMR d (ppm) in D2O, pD ¼ 12
Compounds
N–CH2–O
–CH2N
–CH2CO–
–CH2–CH2–CH2–
(1)
(2)
(3)
(4)
(5)
(6)
—
3.45 (t)
—
3.82 (t)
3.26 (t)
3.44 (t)
3.33 (t)
2.36 (t)
—
2.65 (t)
2.21 (t)
2.85 (t)
2.53 (t)
2.12 (m)
—
2.14 (m)
1.78 (m)
2.13 (m)
1.90 (m)
3.96 (s)
4.70 (s)
4.08 (s)
4.55 (s)
4.35 (s)
Figure 1. 1NMR (400 MHz) spectra at pD ¼ 3.0 (bottom) and pD ¼ 12.7 (top). Bottom signals: 4.70 ppm –CO–CH2–O
compound 3, 3.82 ppm, –CH2N– compound 3, 2.65,–CH2–CO compound 3, 2.14 ppm, –CH2–CH2–CH2– compound 3.
Top signals: 4.70 ppm –CO–CH2–O compound 3, 4.55 ppm, –CO–CH2–O Cyclol 4.35 –CO–CH2–O, Macrocycle, 4.08 –CO–
CH2–O, NBA, 3.96 ppm –CH2– glycolic acid (conjugate base), 3.82 ppm, –CH2N– compound 3, 3.45 –CH2N– 2-pyrrolidone,
2.65, –CH2–CO compound 3, 2.36, ppm –CH2–CO 2-pyrrolidone and 2.14 ppm, –CH2–CH2–CH2– compound 3 and
2-pyrrolidone
carbonyl group. At the same time, NBA (4) is also
produced in the reaction but via hydrolysis of the
macrocycle (6) which is in equilibrium with the cyclol (5)
(Scheme 1). Production of compound 4 is promoted at
higher pD as depicted in Fig. 2 where a plot of log ([4]/
[1]) versus pD is shown. The reaction rate is also
catalyzed by OHꢀ as can be seen in a plot of log kobs
versus pD (Fig. 3). The kobs at any pD is given by the sum
of the rates of 1 and 2 formation plus the rate of formation
of 4. According to Scheme 1, the rate of formation of the
pair 1 and 2, is rL ¼ k3 [ODꢀ] [1]. Assuming a steady state
of [1] yields rL ¼ (k1k3[ODꢀ]/(k3[ODꢀ] þ kꢀ2)) [5]. The
0
5
7
9
11
13
15
-0,2
-0,4
-0,6
-0,8
-1
-1,2
-1,4
pD
Figure 2. Plot of log ([4]/[1]) versus pD. Points: experimental measurements. Solid line: Best fit according to equation in text
Copyright # 2006 John Wiley & Sons, Ltd. J. Phys. Org. Chem. 2006; 19: 737–743

AMIDE AND LACTAM HYDROLYSIS OF N-(2-HYDROXYACETYL)-2-PYRROLIDONE
741
0
5
7
9
11
13
15
-0,5
-1
-1,5
-2
-2,5
-3
-3,5
-4
-4,5
pD
Figure 3. Plot of log kobs versus pD. Points: experimental measurements. Experimental pKa ¼ 11.9, from log kmaxꢀ0.30
rate of formation of 6 is then given by rNBA ¼ k1[ODꢀ]
This means that the rate of lactam formation is first
[6], which can be transformed into rNBA ¼ (k[OHꢀ]k
/
ꢀ1
order in [ODꢀ] and that the rate of NBA formation is
second order in [ODꢀ]. This conclusion is in agreement
with the experimental result of Fig. 2 in which an increase
in the ratio [4]/[1] ¼ kNBA/kL is obtained when increasing
the pD. Figure 3 also shows the essentially correctness of
equations deduced above since kobs ¼ kL þ kNBA in
which kL dominates when pD < pKa.
(k[ODꢀ] þ kꢀ1)) [5ꢀ], on assuming steady state for 6.
However, the two rates can be expressed in terms of the
total amount of 5, that is, [5]T ¼ [5] þ [5ꢀ], and, since 5ꢀ
is the conjugate base of 5, [5] ¼ [Dþ]/(Ka þ [Dþ]) and
[5ꢀ] ¼ Ka/(Ka þ [Dþ]) where Ka is the acidity equi-
librium constant of 5. Therefore, the rate constants for
lactam (kL) and NBA (kNBA) production are then given,
respectively by:
When pD > pKa, Eqns (1) and (2) may be transformed
into the following expressions:
kL ¼ ðk2k3 ½ODꢀꢂ=ðk3½ODꢀꢂ þ kꢀ2ÞÞ
kL ¼ ðk2k3½ODꢀꢂ=kꢀ2Þ ½Dþꢂ=ðKaÞ
(6)
(7)
ꢁ ½Dþꢂ=ðKa þ ½DþꢂÞ
(1)
kNBA ¼ k½ODꢀꢂk1=ðkꢀ1
Þ
kNBA ¼ ðk½ODꢀꢂk1=ðk½ODꢀꢂ þ kꢀ1ÞÞKa=ðKa
þ ½DþꢂÞ
According to the last two equations, the rate of lactam
formation becomes pD independent but the rate of NBA
formation is now first order in [ODꢀ]. This means that
even at high pD, the ratio [4]/[1] ¼ kNBA/kL still depends
on [ODꢀ] but kobs should reach a constant value since the
kL term still dominates in the kobs expression. These
conclusions are indeed in agreement with the exper-
imental results of Figs 2 and 3.
(2)
The Ka value of 5 can be readily estimated from linear
free energy relationships. For instance, using the pKa
value of Ethanol (15.9), a ri value11 for substitution at the
a carbon and the following si values12:
It is important to remark that the last equations have
been deduced in terms of the total concentration of
compound 5. In order to transform these equations as a
function of the total concentration of reactants, the
following considerations must be taken into account:
[1]Total ¼ [1] þ [5]Total þ [6], [5]Total/[1] ¼ KC and [6]/
[5]Total ¼ KM, the total concentration of 5 in terms of
the total amount of reactants becomes:
pKað5Þ ¼ 15:9 ꢀ 8:4 ðsi ðOHÞ þ si ðNHCOEtÞÞ
¼ 15:9 ꢀ 8:4 ð0:25 þ 0:25Þ ¼ 11:7
Two scenarios may then be used to simplify Eqns (1)
(3)
and (2) at pD < 11.7 and at pD > 11.7. In the first
case, kL ¼ (k1k3[ODꢀ]/(k3[ODꢀ] þ kꢀ2)) and kNBA
¼
(k[ODꢀ]k1/(k[ODꢀ] þ kꢀ1)) Ka/[Dþ]. Assuming now
that the equilibrium between the species 3, 5, and 6 is
achieved rapidly as previously reported1,2 for analogous
compounds, kꢀ2 and kꢀ1 > k3[ODꢀ] and k[ODꢀ], respect-
ively. Therefore Eqns (1) and (2) can be transformed to:
½5ꢂTotal ¼ ½1ꢂTotal=ð1 þ KM þ 1=KCÞ
(8)
Therefore, if KM and KC < 1, then [5]Total ¼ [1]Total. In
any case, the correction of Eqns (1) and (2) and the
equations derived from them, involves dividing or
multiplying these equations by a constant that does not
change the conclusions pointed out above.
kL ¼ k2k3½ODꢀꢂ=ðkꢀ2
Þ
(4)
(5)
kNBA ¼ ðk½ODꢀꢂ k1=kꢀ1Þ ðKa=½DþꢂÞ
Copyright # 2006 John Wiley & Sons, Ltd.
J. Phys. Org. Chem. 2006; 19: 737–743

´
˜
L. G. BORBOA AND O. NUNEZ
742
Table 2. Rate and equilibrium constants experimentally measured and estimated
kobs (sꢀ1)1 (k3 þ k) ꢃ 5%
2.1 ꢁ 10ꢀ6
KB (Mꢀ1 sꢀ1
)
pKa (cyclol)
KM ꢃ 5%
KC ꢃ 10%
(k3 þ k)
2.5 ꢁ 10ꢀ2 (B ¼ OHꢀ)1
5 ꢁ 10ꢀ7 (B ¼ HPO24ꢀ
1
)
6 ꢁ 10ꢀ4 (B ¼ PO34ꢀ
1
)
k3
11.93 (11.7)4
15
0.1 (pD ¼ 12)5
4 (B ¼ OHꢀ)2
0.6 (B ¼ OHꢀ)2
k
1 From kobs versus [B] plots.
2 Estimated (see text).
3 Experimental value from Fig. 3.
4 Estimated value from Eqn (3).
5 From 1HNMR signal integration.
Estimates of the KM and KC values can be deduced
from the NMR spectra at different pH. For instance, at
high pH where the cyclol (5) and macrocycle (6) forms
are detected (Fig. 1) with a low and almost equal signal
intensity. Therefore, KM is ca. 1. However, the KC value is
lower than 1 (ca. 0.1) as can be deduced from the relative
intensity of the open form 3 and the cyclol 5 signals at pH
ca. 12.
From Fig. 3, where a leveling in the plot is observed in
accord with Eqn (6), the pKa of the cyclol 5 can be
experimentally evaluated. According to the plot in Fig. 3
this pKa value is 11.7. This value is in good agreement with
the one estimated, from linear free energy relationships, of
11.9. From Fig. 3 and according to Eqn (6) the leveling is
equal to k2k310ꢀ14/kꢀ2Ka, that is, k2k310ꢀ2/kꢀ2. But k2/
in NBA catalysis is a consequence of the one order of
magnitude difference between the second-order rate
constants k3 and k. Therefore, the net effect (according to
Eqn (9)) is the independence of the product ratio on
[ODꢀ]. In Table 2, the experimentally obtained equi-
librium and rate constants, estimated and derived in this
work are shown.
Using the three first kB values in the second column of
Table 2 and the corresponding pKa values for KH2PO4
(7.21), K2HPO4 (12.32), and H2O (15.74), a b ¼ 0.55
(R2 ¼ 0.995) is obtained. Since k3 is the main path
contributing to the base catalysis, the b ¼ 0.55 value
means that the proton transferred from water to the
conjugate base of the catalyst has been transferred ca.
50% at the transition state of the k3 step. Bi-functional
catalysis with participation of the base-conjugated acid is
also a possibility due to the need for protonation of the
amino amide-leaving group. In fact, protonation of the
leaving lactam (k3 step) or amide group (k1 step) is rate
determining for the reaction product formation. At high
pH water or the buffer-conjugated acid form may be
acting as general acid catalyst. The planar five-membered
ring of 2-pyrrolidone may also facilitate this protonation.
In fact, protonation of the electron pair on nitrogen that is
needed to promote the departure of the lactam occurs
while maintaining the electron pair coming from the
cleavage synperiplanar with the p orbital of the lactam
carbonyl group. This induced stabilization at the
transition state probably makes the difference in the
observed chemistry of N-(2-hydroxyacetyl)-2-piperidone
in which we have not detected lactam formation as
compared with the one of compound 3.
k
¼ 1/KC ¼ 0.1 and the rate constant value at the leveling,
ꢀ2
according to Fig. 3, is 0.24 minꢀ1 (4 ꢁ 10ꢀ3 sꢀ1). There-
fore, 0.24 minꢀ1 ¼ k3 ꢁ 0.1 ꢁ 10ꢀ2 M and k3 ¼ 240 Mꢀ1
minꢀ1 (4 Mꢀ1 sꢀ1).
Dividing Eqn (5) by (4), the [NBA]/[lactam] ¼ [4]/[1]
ratio is obtained: [4]/[1] ¼ (k ꢁ 10 Mꢀ1/k3)[ODꢀ]. From
a plot of [4]/[1] versus [ODꢀ] (not shown), a slope
(taking the last four points) of 1.52 (R2 ¼ 0.98) is
obtained. Therefore, slope ¼ 1.52 Mꢀ1 ¼ k ꢁ 103 Mꢀ1
(0.6 Mꢀ1 sꢀ1).
/
k3 Mꢀ1; from where a k ¼ 1.52 ꢁ k3/10 ¼ 36 Mꢀ1 minꢀ1
Under conditions in which lactam and NBA production
is not catalyzed by base (Eqns (4) and (5), with the terms
k3[ODꢀ] and k[ODꢀ], respectively, equal to kw) still the
NBA production is inversely dependent on [Dþ] as shown
in Eqn (9) that represents Eqn (5) under non-catalyzed
conditions.
kNBA ¼ ðkwkꢀ1=kꢀ1Þ ðKa=½DþꢂÞ
(9)
CONCLUSIONS
This inverse dependency on [Dþ] is manifested in the
ratio [NBA]/[lactam] at low pH as shown in Fig. 2. After
this dependency a leveling is observed in Fig. 2. This
means that in that range the NBA production is still non-
catalyzed by base while lactam production is. This delay
The second-order rate constant for hydrolysis of
compound 3 to yield 2-pyrrolidone (1) and glycolic acid
(2) is quite fast (4 Mꢀ1 sꢀ1) and comparable13 to the
hydrolysis rate of reactive p-nitrophenyl acetate
Copyright # 2006 John Wiley & Sons, Ltd.
J. Phys. Org. Chem. 2006; 19: 737–743

AMIDE AND LACTAM HYDROLYSIS OF N-(2-HYDROXYACETYL)-2-PYRROLIDONE
743
the N-(2-hydroxyacetyl)-2-piperidone, the cyclol deriva-
tive is the more stable form at any pH. However, in the
case of compound 3, the cyclol form becomes relatively
important only at high pD.
These results are then quite important to promote the
hydrolysis of the unreactive five-membered 2-pyrrolidone
lactam and may also be of relevance in the in situ
production, at brain level, of a derivative of GABA that
may act as inhibitory neurotransmitter. The reversible
equilibrium established between cyclol 5 and macrocycle
6 present an interesting system that may attach reversibly
to the active site of HIV-1 protease and act as its inhibitor.
Scheme 2. Resonance form of compound 3 which favor
nucleophilic hydroxyl attack on the exocyclic carbonyl carbon
and retards the endocyclic carbonyl attack
(15 Mꢀ1 sꢀ1) at 25 8C and two orders of magnitude slower
than the hydrolysis of N-acetylimidazole (316 Mꢀ1 sꢀ1).
This unusual reactivity is due to the restricted resonance
between the nitrogen electron pair and the exocyclic
carbonyl group that makes this carbonyl group reactive.
The restriction is due to an efficient delocalization of the
electron pair with the five-membered lactam carbonyl
group (Scheme 2). In fact, this kind of delocalization has
also been indicated13 as that responsible for the high
reactivity of N-acetyl-imidazole. In both cases, the
delocalization of the electron pair on nitrogen is promoted
by the planar structure of the five-membered ring that
favors the maximum n-p orbital overlapping.
Acknowledgements
We acknowledge the Environment Management Unit
(UGA) at Simon Bolivar University for financial support.
REFERENCES
The formation of the GABA derivative NBA (4) is also
fast (0.6 Mꢀ1 sꢀ1) and its formation, relative to formation
of 1, is enhanced at higher pH. This tendency has been
attributed to the formation of the derivative not directly
from OHꢀ attack on the lactam carbonyl group but rather
from the macrocycle 6 in equilibrium with the cyclol 5
and the N-(2-hydroxyacetyl)-2-pyrrolidone (3). The
kinetic behavior shown in this work and the equilibrium
established between the three forms mentioned above are
strong arguments to support the proposal. The alternative,
direct hydroxyl attack on the endocyclic carbonyl group is
ruled out due to the kinetic argument and to the decrease
in reactivity of this carbonyl group induced by the
resonance form shown in Scheme 2. In fact, this low
reactivity also influences the intramolecular alkoxyl
attack. For instance, we have found14 that in the case of
´
1. Tineo E, Pekerar SV, Nun˜ez O. J. Chem. Soc. Perkin Trans. 2 2002;
244–246.
2. Guedez T, Nun˜ez A, Tineo E, Nun˜ez O. J. Chem. Soc. Perkin Trans.
´
2 2002; 2078–2082.
´
´
´
3. Nun˜ez O, Rodrıguez J, Angulo L. J. Phys. Org. Chem. 1993; 7: 80–
89.
4. Imming P, Klar B, Dix D. J. Med. Chem. 2000; 43: 4328–4331.
5. Ghosh M, Conroy JL, Seto CT. Angew. Chem. Int. Ed. 1999; 38(4):
514–516.
6. Groves JT, Baron LA. J. Am. Chem. Soc. 1989; 111: 5442–5448.
7. Carrieri P, Spire B, Duran S, Katlama C, Peyramond D, Francois C,
Chene G, Lang JM, Moatti JP, Leport C. J. Acquir. Immune Defic.
Syndr. 2003; 32: 38–47.
8. Dunn B. Chem. Rev. 2002; 102: 4431–4468.
9. Griot R, Frey A. Tetrahedron 1963; 19: 1661–1673.
10. Glosoe P, Long F. J. Phys. Chem. 1960; 64: 188–190.
11. Fox JP, Jencks WP. J. Am. Chem. Soc. 1974; 96: 1436–1449.
12. Ritchie CD, Sager WF. Progr. Phys. Org. Chem. 1964; 2: 323–456.
13. Fife TH. Acc. Chem. Res. 1993; 26: 325–331.
´ ´
14. Nun˜ez O. Acta Cientıfica Venezolana 1989; 40: 301–302.
Copyright # 2006 John Wiley & Sons, Ltd.
J. Phys. Org. Chem. 2006; 19: 737–743