Reactivity of lactones and GHB formation

Article abstract of DOI:10.1021/jo040271i

(Chemical Equation Presented) The behavior of lactones in their hydrolysis reactions is a good indicator of their reactivity as electrophilic molecules. The hydrolysis of four- to six-membered lactones was investigated in neutral (water) and slightly acid media and in water/dioxane media. The following conclusions were drawn: (i) The reactivity of β-propiolactone in neutral water is more than four times greater than that of β-butyrolactone, due to the flow of charge caused by the latter's methyl substituent. Reactivity is enthalpy-controlled. (ii) The reactivity of β-lactones diminishes in water/dioxane media when the percentage of dioxane increases. The increase in the dioxane percentage relaxing the intermolecular hydrogen bonds in the ordered structure of the water reduces ΔH^# and simultaneously increases the -ΔS^# value. (iii) An inverse solvent kinetic isotope effect in the acid-catalyzed hydrolysis of γ-butyrolactone and δ-valerolactone was observed, this being indicative of acyl cleavage. (iv) The ΔH^# and ΔS^# values permit discrimination between alkyl and acyl cleavage, (v) A correlation was found between the chemical reactivity of lactones and their carcinogenic activity. (vi) The results suggest that orally ingested γ-butyrolactone remains largely in its nonhydrolyzed form in the stomach before passing into the blood. (vii) The concentration equilibrium constant of GHB formation at human body temperature is K_eq (37°C) = 0.40. (viii) Study of GHB formation shows that, contrary to earlier results, this is an endothermic process, with Δ_rH= 3.6 kJ mol^-1.

Full text of DOI:10.1021/jo040271i

Reactivity of Lactones and GHB Formation
M. Teresa P e´ rez-Prior, Jos e´ A. Manso, M. del Pilar Garc ´ı a-Santos, Emilio Calle, and
Julio Casado*
Departamento de Qu ı´ mica F ı´ sica, Universidad de Salamanca, E-37008 Salamanca, Spain
jucali@usal.es
Received October 4, 2004
The behavior of lactones in their hydrolysis reactions is a good indicator of their reactivity as
electrophilic molecules. The hydrolysis of four- to six-membered lactones was investigated in neutral
(
water) and slightly acid media and in water/dioxane media. The following conclusions were
drawn: (i) The reactivity of â-propiolactone in neutral water is more than four times greater than
that of â-butyrolactone, due to the flow of charge caused by the latter’s methyl substituent. Reactivity
is enthalpy-controlled. (ii) The reactivity of â-lactones diminishes in water/dioxane media when
the percentage of dioxane increases. The increase in the dioxane percentage relaxing the
#
intermolecular hydrogen bonds in the ordered structure of the water reduces ∆H and simulta-
#
neously increases the -∆S value. (iii) An inverse solvent kinetic isotope effect in the acid-catalyzed
hydrolysis of γ-butyrolactone and δ-valerolactone was observed, this being indicative of acyl cleavage.
#
#
(
iv) The ∆H and ∆S values permit discrimination between alkyl and acyl cleavage. (v) A correlation
was found between the chemical reactivity of lactones and their carcinogenic activity. (vi) The results
suggest that orally ingested γ-butyrolactone remains largely in its nonhydrolyzed form in the
stomach before passing into the blood. (vii) The concentration equilibrium constant of GHB formation
at human body temperature is Keq (37 °C) ) 0.40. (viii) Study of GHB formation shows that, contrary
-
1
to earlier results, this is an endothermic process, with ∆ H ) 3.6 kJ mol .
r
Introduction
Since some lactones give alkylating reactions with any
of a number of nucleophilic sites in tissues, the present
4
The reactivity of amino acids with an -NH
2
group in
investigation was performed with two main objectives:
(i) To gain quantitative knowledge about the lactones’
reactivity and to relate this to their carcinogenic activity
and (ii) to gain insight into the formation of γ-hydroxy
butyric acid (GHB) in the hydrolysis of γ-butyrolactone.
γ-Butyrolactone is used widely as an industrial solvent
and is marketed as a dietary supplement. Since GHB is
a drug of abuse and is also used as a dietary supplement
claimed to improve physical performance and to reduce
nitrosation reactions and its relationship with the alkyl-
ating potential of their products have been investigated
in previous work.1 Among other conclusions, it was
observed that the species resulting from the nitrosation
of these amino acids are the corresponding lactones.
-3
(
1) Garc ´ı a Santos, M. P.; Calle, E.; Casado, J. J. Am. Chem. Soc.
001, 123, 7506.
2) Garc ´ı a Santos, M. P.; Gonz a´ lez Mancebo, S.; Hern a´ ndez Benito,
J.; Calle, E.; Casado, J. J. Am. Chem. Soc. 2002, 124, 2177.
3) Garc ´ı a Santos, M. P.; Calle, E.; Casado, J. Polyhedron 2003, 22,
059.
2
(
(4) Lawley, P. D. In Chemical Carcinogens, 2nd ed.; Searle, C. E.,
Ed.; ACS Monograph 182; American Chemical Society: Washington,
DC, 1984; Vol. 1, Chapter 7.
(
1
10.1021/jo040271i CCC: $30.25 © 2005 American Chemical Society
420
J. Org. Chem. 2005, 70, 420-426
Published on Web 12/28/2004

Reactivity of Lactones and GHB Formation
SCHEME 1. Hydrolysis of Lactones
FIGURE 1. Integrated form of the pseudo-first-order rate
equation (eq 1) for the hydrolysis of â-propiolactone and
5
stress, investigation of its formation is of biological/
o
â-butyrolactone. [L] ) 0.08 M, T ) 35 °C.
biochemical interest.
In addition, the reactivity of the cyclic esters is of
current interest because many biologically important
molecules contain cyclic phosphate and lactone struc-
TABLE 1. Rate Constantsa as a Function of
b
Temperature for the Neutral Hydrolysis of
â-Propiolactone and â-Butyrolactone
6
tures.
T (°C)
107kBPL
107kBBL
The behavior of lactones in their hydrolysis reactions
is a good indicator of their reactivity as electrophilic
molecules. Most earlier investigations were carried out
in alkaline media, and studies in neutral or slightly acid
M-1
-1 c
(
s )
1
1
2
25.0
27.5
3
3
3
5.0
7.5
2.5
3.54 ( 0.05
4.30 ( 0.04
7.6 ( 0.4
10.6 ( 0.1
14.9 ( 0.3
20.2 ( 0.7
27.4 ( 0.9
31.9 ( 0.9
0.94 ( 0.01
1.82 ( 0.01
2.22 ( 0.03
3.63 ( 0.02
4.38 ( 0.05
5.89 ( 0.08
7.69 ( 0.07
7-12
media have received scant attention,
even though they
are the most relevant in biochemical terms.4
,13
0.0
2.5
5.0
Previous research has shown that the hydrolysis of
â-butyrolactone by molecular water yields bimolecular
displacement reactions, with cleavage of the bond be-
a
As k in eq 1. b [BPL]o ) 0.08 M; [BBL]o ) 0.08 M. c Values of
rate constants are given within the 95% confidence interval.
1
0,14-17
tween the alcoholic carbon and oxygen.
al.
Olson et
1
6,17
demonstrated that the hydrolysis of â-butyrolac-
tone in basic and strongly acid solutions results in the
expected acyl-oxygen cleavage.
The present investigation was carried out in two
stages: (i) A kinetic study of the hydrolysis reactions of
â-propiolactone (BPL), â-butyrolactone (BBL), γ-butyro-
lactone (GBL), and δ-valerolactone (DVL) and (ii) a
kinetic and thermodynamic study of the formation of
GHB.
In the chemistry of synthesis, different attempts have
been made to find effective reagents for lactone ring-
opening; e.g., the preparation of ω-alkylthio carboxylic
acids through ω-carbon-oxygen bond cleavage.18 â-Pro-
piolactone has a curious dichotomy of reactivity; diethyl-
aluminumethanethiolate has been reported19 to cause
both O-alkyl and O-acyl cleavage at the same time.
More recently, Moore and Schwab,20 working with
strongly acidic conditions (6 N HCl), concluded that five-
membered lactones with bulky substituents at C-2 can
be hydrolyzed by a previously unreported alkyl cleavage
mechanism at the expense of the usually predominant
acyl cleavage.
Results and Discussion
Neutral Hydrolysis: Water Reactions. We inves-
tigated the hydrolysis of â-propiolactone and â-butyro-
lactone (Scheme 1a).
Experiments designed to determine the influence of the
lactone concentration revealed the reactions to be first-
order with respect to this reagent. Thus:
A review about the mechanisms of lactone hydrolysis
was published by Kaiser and K e´ zdy.6
rate ) k[H O][L] ) k [L]
(1)
2
1
(
(
(
(
(
(
(
(
(
5) Mason, P. E.; Kerns, W. P., II. Acad. Emerg. Med. 2002, 9, 730.
6) Kaiser, E. T.; K e´ zdy, F. J. Prog. Bioorg. Chem. 1976, 4, 239.
7) Henry, P. Z. Physik. Chem. 1892, 10, 96.
where [L] represents the concentration of lactone and k
) k[H O] is the pseudo-first-order rate constant.
Figure 1 represents the integrated form of eq 1 in
terms of [L] , the concentration of lactone at time t, with
[L] being the initial concentration.
Table 1 gives the kBPL and kBBL rate constants (as k in
eq 1) for the hydrolysis reactions of BPL and BBL,
respectively, as a function of temperature (T).
The results given in Table 1 show that the rate
constants are more than four times shorter for â-butyro-
lactone than for â-propiolactone.
1
8) Johansson, H.; Sebelius, H. Ber. 1918, 51, 480.
2
9) Kailan, A. Z. Physik. Chem. 1920, 94, 111; 1922, 101, 63.
10) Bartlett, P. D.; Small, G., Jr., J. Am. Chem. Soc. 1950, 72, 4867.
11) Long, F. A.; Purchase, M. J. Am. Chem. Soc. 1950, 72, 3267.
12) Coffin, F. D.; Long, F. A. J. Am. Chem. Soc. 1952, 74, 5767.
13) Van Duuren, B. L.; Goldschmidt, B. M. J. Med. Chem. 1966, 9,
t
o
7
7.
(
(
(
(
(
14) Holmberg, B. Svensk. Kem. Tid. 1918, 30, 190, 215.
15) Long, F. A.; Olson, A. R. J. Phys. Chem. 1937, 41, 267.
16) Olson, A. R.; Miller, R. J. J. Am. Chem. Soc. 1938, 60, 2687.
17) Olson, A. R.; Hyde, J. L. J. Am. Chem. Soc. 1941, 63, 2459.
18) Node, M.; Nishide, K.; Ochiai, M.; Fuji, K.; Fujita, E. J. Org.
Chem. 1981, 46, 5163.
19) Hirabayashi, T.; Itoh, K.; Sakai, S.; Ishii, Y. J. Organomet.
Chem. 1970, 25, 33.
20) Moore, J. A.; Schwab, J. M. Tetrahedron Lett. 1991, 32, 2331.
(
Table 3 shows the values of the activation parameters
#
#
(
∆H and ∆S for both hydrolysis reactions.
J. Org. Chem, Vol. 70, No. 2, 2005 421

P e´ rez-Prior et al.
TABLE 2. bRate Constantsa as a Function of the Composition of the Media and the Temperature for the Neutral
Hydrolysis of â-Propiolactone and â-Butyrolactone
water/dioxane (volume ratio)
6/4
8/2
5/5
4/6
7
7
7
7
7
7
7
7
T (°C)
10 kBPL
10 kBBL
10 kBPL
10 kBBL
10 kBPL
10 kBBL
10 kBPL
10 kBBL
M-
1
s
-1 c
)
(
25.0
27.5
30.0
32.5
35.0
10.3 ( 0.3
1.75 ( 0.01
2.34 ( 0.01
3.10 ( 0.03
4.10 ( 0.04
5.41 ( 0.05
6.84 ( 0.07
8.46 ( 0.07
11.8 ( 0.2
14.7 ( 0.4
18.9 ( 0.5
0.87 ( 0.01
1.20 ( 0.01
1.55 ( 0.02
2.09 ( 0.03
2.63 ( 0.06
5.1 ( 0.1
7.02 ( 0.03
8.6 ( 0.3
11.0 ( 0.5
14.6 ( 0.3
0.61 ( 0.01
0.81 ( 0.02
1.09 ( 0.01
1.39 ( 0.01
1.85 ( 0.01
3.43 ( 0.04
4.50 ( 0.01
5.66 ( 0.08
7.31 ( 0.03
9.3 ( 0.1
0.387 ( 0.008
0.500 ( 0.007
0.67 ( 0.02
0.87 ( 0.01
1.11 ( 0.03
13.40 ( 0.06
17.5 ( 0.4
23.2 ( 0.6
30.4 ( 0.9
a
As k in eq 1. b [BPL]o ) 0.08 M; [BBL]o ) 0.08 M. c Values of rate constants are given within the 95% confidence interval.
TABLE 3. Activation Parameters as a Function of the
Composition of the Media for the Neutral Hydrolysis of
â-Propiolactone and â-Butyrolactone
a
-∆S#a (J K-1 mol-1
)
∆
H# (kJ mol-1
)
water/dioxane
volume ratio)
(
BPL
BBL
BPL
BBL
1
8
6
5
4
0/0
83 ( 2
80 ( 1
76 ( 3
75 ( 3
73 ( 1
88 ( 3
84 ( 1
82 ( 2
82 ( 1
79 ( 2
80 ( 8
90 ( 3
106 ( 10
112 ( 9
123 ( 3
78 ( 10
94 ( 1
105 ( 7
109 ( 4
122 ( 6
/2
/4
/5
/6
a
Values are given with their standard deviations.
With the aim of gaining deeper insight into the
transition state, the hydrolysis of BPL and BBL in water/
dioxane was investigated. Tables 2 and 3 show the results
obtained.
FIGURE 2. Isokinetic relationship in the hydrolysis of
â-propiolactone (2) and â-butyrolactone (b) in water/dioxane
media.
#
The very similar ∆S values found for the hydrolysis
of BPL and BBL (Table 3) are consistent with the
analogous geometry of these lactones. They also show
that the reactivity of these â-lactones is mainly enthalpy-
controlled.
should decrease when increasing those of the water/
dioxane ratio, as was also observed (Table 3).
#
The greater ∆H value obtained for the hydrolysis of
Since the strong effect of the dioxane on the kinetic
parameters is a consequence of a strong lactone-water
interaction, the enthalpy and entropy will both tend to
be linear functions of the volume fraction of that solvent,
â-butyrolactone must be caused by the methyl group as
a donor of charge on the â-carbon, with a decrease in its
electrophilic character (Scheme 1a). Since in general
terms substitution on carbon atoms 2 or 3 in lactones
23
and an isokinetic relationship should result. Figure 2
21
results in a decrease or loss in carcinogenic activity, the
steric hindrance of the BBL methyl group must also
shows that this effect was indeed observed.
Because the tumorigenicity of lactones would be cor-
#
contribute to the higher value of ∆H .
related with their alkylating capacity on nucleophilic
It can be observed that when the percentage of dioxane
increased: (i) the rate constant decreased appreciably,
4,21
substrates,
the greater reactivity of â-propiolactone
would imply a greater carcinogenic potential. The data
reported in Table 4 are consistent with such a conclusion.
Acid-Catalyzed Hydrolysis. Under the same condi-
tions as for BPL and BBL, no neutral hydrolysis reactions
of GBL and DVL were observed.
(
ii) the enthalpy of activation became smaller, and (iii)
the absolute value of the entropy of activation increased.
Mixed solvents behave regularly only insofar as the
components are similar; thus, they can be characterized
by the bulk relative permittivity, which is a monotonic
function of component permittivities. If one solvent is
protic and highly structured, its structure is broken by
a foreign solvent and the solvation power diminishes
rapidly, increasing the disorder of the protic solvent.
This latter being the current case: (i) the increase in the
dioxane percentage relaxing the intermolecular O‚‚‚H
hydrogen bonds in the ordered structure of the water
The loss of reactivity of GBL and DVL at neutral pH
4
is in agreement with their inactivity as carcinogens, as
well as with our previous results concerning the reactivity
1-3
of lactones resulting in the nitrosation of amino acids.
The alkylation time of 4-(p-nitrobenzyl)pyridine (a trap
2
2
for alkylating agents2
7,28
with similar nucleophilic char-
#
(23) Leffler, J. E.; Grunwald, E. Rates and Equilibria of Organic
Reactions As Treated by Statistical, Thermodynamic, and Extrather-
modynamic Methods; Dover: New York, 1963; p 402.
(24) Van Duuren, B. L.; Langseth, L.; Orris, L.; Teebor, G.; Nelson,
N.; Kuschner, M. J. Natl. Cancer Inst. 1966, 37, 825.
increases its hydrolytic activity by reducing ∆H (see
Table 3); (ii) as a consequence of the considerable disorder
#
induced by the dioxane in the reagent water, -∆S values
(
25) Van Duuren, B. L.; Langseth, L.; Orris, L.; Teebor, G.; Baden,
(
21) Van Duuren, B. L. Ann. N. Y. Acad. Sci. 1969, 163, 633.
22) Exner, O. Correlation Analysis of Chemical Data; Plenum
M.; Kuschner, M. J. Natl. Cancer Inst. 1967, 39, 1213.
(
(26) Van Duuren, B. L.; Goldschmidt, B. M.; Katz, C.; Langseth, L.;
Mercado, C.; Sivak, A. Arch. Environ. Health (Chicago) 1968, 16, 472.
Press: New York, 1988; p 186.
422 J. Org. Chem., Vol. 70, No. 2, 2005

Reactivity of Lactones and GHB Formation
TABLE 4. Dose and Tumorigenicity of Lactones by Subcutaneous Injection in Mice and Rats21,24-26
number of malignant tumors at
injection site/number of animals
dose in mg
subcutaneous injection in micea
â-propiolactone
â-butyrolactone
0.73
10
18/30
18/30
subcutaneous injection in ratsb
â-propiolactone
â-butyrolactone
4
13/20
9/20
100
a
Performed with 0.05 mL of tricaprylin as a vehicle. b Performed with 0.1 mL of tricaprylin as a vehicle.
FIGURE 3. Formation of GHB in the acid hydrolysis of γ-butyrolactone. (a) Variation in [GHB] with the reaction time. (b)
-
2
Integrated form of the rate equation (eq 1). [GBL]
o
) 0.1 M, [HCl] ) 1.18 × 10 M, T ) 20 °C.
acteristics to DNA) by â-lactones generated in the nit-
rosation of â-amino acids was Ak alkylation ) 2 to more than
7
0 h, with talkylation > 150 h for the γ-butyrolactone derived
1
from the nitrosation of γ-amino butyric acid.
Upon carrying out the reactions in weak hydrochloric
acid solutions, the following results were obtained.
Since the hydrolysis reaction of GBL is reversible, the
first-order rate constant for the forward reaction
+
(
Bk
1
) Bk GBL[H
3
O ], Bk GBL being the catalytic constant) is
given by
FIGURE 4. Variation in the GBL hydrolysis rate constant
with the acidity of the medium. [GBL] ) 0.1 M, T ) 35 °C.
[
GHB]eq
[GHB]eq
1
o
Bk )
ln
(2)
1
t [GBL]o
[GHB]eq - [GHB]
Keq ) Bk GBL/ Ak GHB ) [GHB] /[GBL] )
eq
eq
[
GHB] /([GBL] - [GHB] ) (3)
eq o eq
o
where [GBL] represents the initial lactone concentration,
and [GHB]eq and [GHB] are the concentrations of the
hydroxy acid at equilibrium and at time t, respectively.
is straightforward ( Bk GBL and Ak GHB represent the second-
order forward and backward rate constants, respectively).
Figure 3a shows a typical kinetic run, allowing one to
know [GHB]eq as the plateau value of [GHB]. Figure 3b
shows the good fit of the results to eq 2.
To check the hydrolysis of GBL occurring through an
acid-catalyzed mechanism, experiments at different pH
values were carried out. The slope value of R ) 1.00 in
Figure 4 is evidence of this.29
Regarding DVL, the experiments revealed that the
reactions were first-order with respect to this reagent:
+
rate ) kDVL[H
3
O ][DVL]. The results obtained with this
lactone are summarized in Tables 5 and 6 and show the
following: (i) The rate constant of the acid-catalyzed
hydrolysis of DVL (kDVL) is two-orders of magnitude
higher than that ( Bk GBL) of GBL hydrolysis. (ii) The Gibbs
The values of Bk GBL at different temperatures are
reported in Table 5. With the experimental value of
-1
energy of activation for DVL hydrolysis (81 kJ mol ; 25
[GHB]eq, knowledge of Keq, the concentration equilibrium
-1
°
C) is lower than that seen for GBL (94 kJ mol ; 25 °C).
constant of the hydrolysis of GBL,
The data reveal that the activity of DVL as an electro-
philic reagent is greater than that of GBL, so much so
that the hydrolysis of the latter is a clearly reversible
reaction.
(
27) Sawicki, E.; Bender, D. F.; Hauser, T. R.; Wilson, R. M., Jr.;
Meeker, J. E. Anal. Chem. 1963, 35, 1479.
28) Thomas, J. J.; Kim, J. H.; Mauro, D. M. Arch. Environ. Contam.
Toxicol. 1992, 22, 219.
29) Espenson, J. H. Chemical Kinetics and Reaction Mechanisms,
nd ed.; McGraw-Hill: New York, 1995; Chapter 10.
(
Another piece of information consistent with the notion
of acylic cleavage is the solvent kinetic isotope effect,
(
2
J. Org. Chem, Vol. 70, No. 2, 2005 423

P e´ rez-Prior et al.
TABLE 5. Rate Constants as a Function of
a
Temperature for the Acid-Catalyzed Hydrolysis of
γ-Butyrolactone and δ-Valerolactone
4
2
T (°C)
10 Bk GBL
10 kDVL
M-
s )
1
-1 b
(
15.0
17.5
20.0
22.5
25.0
27.5
30.0
32.5
35.0
0.99 ( 0.01
1.27 ( 0.01
1.55 ( 0.03
1.98 ( 0.03
2.40 ( 0.02
3.11 ( 0.05
3.82 ( 0.04
4.95 ( 0.07
6.22 ( 0.06
1.97 ( 0.03
2.35 ( 0.06
2.76 ( 0.05
3.24 ( 0.05
4.06 ( 0.07
4.7 ( 0.2
5.4 ( 0.1
6.7 ( 0.2
7.3 ( 0.2
a
-2
[
GBL]o ) 0.1 M; [DVL]o ) 0.1 M; [HCl]GBL ) 1.18 × 10 M;
-3 b
[
HCl]DVL ) 4.72 × 10 M. Values of rate constants are given
within the 95% confidence interval.
FIGURE 5. (a) Orientation of the dipoles at oxygen for
unprotonated and protonated δ-valerolactone. In the proto-
nated molecule, the dipoles’ orientation is more favorable. (b)
Contraction of the lone pair orbitals leading to decreased lone
pair-lone pair repulsion.
TABLE 6. Activation Parameters for the Acid-
Catalyzed Hydrolysis of γ-Butyrolactone and
δ-Valerolactone
HB * (∆ HA )
a
* b
-∆ SB *a (-∆ SA )
* b
∆
-
1
-1
-1
lactone
(kJ mol
)
(J K mol )
TABLE 8. Variation in the Equilibrium Constant with
Temperature in the GHB Formationa
γ-butyrolactone
δ-valerolactone
65 ( 1 (61 ( 1)
47 ( 1
96 ( 3 (100 ( 4)
113 ( 5
10Keqb
a
Values are given with their standard deviations. b Magnitudes
T (°C)
within parentheses correspond to the back reaction.
15.0
20.0
25.0
30.0
35.0
3.66 ( 0.04
3.73 ( 0.04
3.79 ( 0.04
3.93 ( 0.04
4.03 ( 0.04
TABLE 7. Solvent Kinetic Isotope Effect in the
Acid-Catalyzed Hydrolysis of Lactones
lactone
kH O/kD O
2
2
a
GBL]o ) 0.1 M; [HCl] ) 1.18 × 10-2 M. b Mean values of three
[
γ-butyrolactone
δ-valerolactone
0.67
0.41
experiments.
At the molecular level, there are arguments that
rationalize the kinetic and thermodynamic parameters
found in the acid-catalyzed hydrolysis of GBL and DVL.
Some of the major effects of protonation are a change in
the direction of the dipole component at the carbonyl
oxygen (Figure 5a) and a contraction of the lone pair
orbitals33 (Figure 5b). Accordingly, the hydrolytic encoun-
ter between the protonated lactones and water molecules
will be favored by diminishing intermolecular lone pair-
lone pair repulsion. This should cause a lower enthalpy
of activation and a more negative entropy of activation
in the reactions occurring with protonation of the car-
boxyl carbon, as was observed (Table 6).
SKIE.30 Table 7 shows this effect to be inverse. This is
indicative of preequilibrium protonation of the lactone
molecule prior to the reaction, as would be anticipated
for an acid-catalyzed mechanism.31 Brown et al. re-
ported a number of examples of transition states for acid
acyl cleavage mechanisms with inverse solvent effects (it
should be noted that SKIE was not observed in the
noncatalyzed hydrolysis of BPL and BBL).
32
Comparison of the results from the study of neutral
and acid-catalyzed reactions provided the following con-
clusions:
#
(
i) Lower ∆H values for the acid-catalyzed reactions.
The acyl-oxygen fission route being the normal path, the
Study of the reactions of triethyloxonium ion with GBL
+
presence of H ions enhancing the electrophilic character
3
3
and DVL to form the ethylated lactone has shown that
#
of the carbonyl carbon should imply lower ∆H values,
-
1
the basicity of DVL is 13 kJ mol greater than that of
as was observed.
GBL. Ab initio calculations33 also reveal that the proton
#
(
ii) More negative ∆S values for the acid-catalyzed
-
1
affinity of DVL is 8 kcal mol greater than that of GBL.
This allows one to assume a lower Gibbs energy of
activation for DVL-catalyzed hydrolysis reaction than
that for GBL, as was in fact observed.
reactions. In the acyl fission mechanism, cleavage of the
C-O bond is not required in the transition state. In
addition, nucleophilic attack by water on the lactone
molecule occurs on the previously protonated carbonyl
It may be concluded that the kinetic and thermody-
namic results can be used as a criterion to discern
between mechanisms occurring through alkyl-oxygen or
acyl-oxygen cleavage.
+
C-atom, i.e., a C -atom. This involves a compact tetra-
hedral intermediate in the case of the acid-catalyzed
#
hydrolysis, i.e., a more negative ∆S value (Table 6).
Finally, it must be pointed out that when we studied
the hydrolysis of BPL and BBL, the possibility of acid-
catalyzed hydrolysis was also investigated. Results showed
(
30) Laughton, P. M.; Robertson, R. E. In Solute-Solvent Interac-
tions; Coetzee, J. F., Ritchie, C. D., Eds.; Marcel Dekker: New York,
1
1
1
969; pp 399-538.
31) Deraniyagala, S. A.; Adediran, S. A.; Pratt, R. F. J. Org. Chem.
995, 60, 1619.
32) Brown, R. S.; Bennet, A. J.; Slebocka-Tilk, H. Acc. Chem. Res.
992, 25, 481.
(
(
(33) Wiberg, K. B.; Waldron, R. F. J. Am. Chem. Soc. 1991, 113,
7705.
424 J. Org. Chem., Vol. 70, No. 2, 2005

Reactivity of Lactones and GHB Formation
SCHEME 2. Reaction Path of the GHB Formation
in the Acid-Catalyzed Hydrolysis of
γ-Butyrolactone
more liposoluble than GHB, which helps it to pass
through the lipid layers of tissues before it can be
hydrolyzed. Such tissues may act as a GBL reservoir and
38-40
extend the duration of its action as compared to GHB.
These results would also help to explain the fact that
the oral bioavailability of GBL (the degree to which it
becomes available to target tissues) in rats is 85%, higher
than that of GHB40 and also explain why GBL has a
longer duration of action than GHB, possibly due to
differences in drug distribution.4
1-43
Conclusions
(i) â-Propiolactone and â-butyrolactone show hydrolysis
reactions in neutral (water) media, while γ-butyrolactone
and δ-valerolactone need acid media to allow their
hydrolysis to occur. (ii) The reactivity of the â-lactones
is enthalpy-controlled; that of â-propiolactone is more
than four times greater than that of â-butyrolactone, due
to the flow of charge caused by the latter’s methyl
substituent. (iii) The reactivity of â-lactones diminishes
in water-dioxane media when the percentage of dioxane
increases. The increase in the dioxane percentage relax-
ing the intermolecular hydrogen bonds in the ordered
that, in the working conditions, catalyzed hydrolysis can
be discarded.
Formation of GHB: Heat of Reaction. Since the
fraction of lactone and hydroxy acid existing at equilib-
rium is dependent on the relative thermodynamic stabil-
ity of the acid and its precursor lactone, the concentration
equilibrium constant Keq was measured.
With eq 3, Keq values were determined at different
temperatures. Table 8 shows the results obtained by
using the van’t Hoff equation34
#
structure of the water reduces ∆H and simultaneously
#
increases the -∆S value. (iv) The inverse solvent kinetic
isotope effect in the acid-catalyzed hydrolysis of γ-butyro-
lactone and δ-valerolactone points to an acyl-cleavage
#
#
mechanism. (v) The ∆H and ∆S values permit discrimi-
nation between alkyl and acyl cleavage. (vi) A correlation
was found between the chemical reactivity of the lactones
and their carcinogenic activity. (vii) The results suggest
that orally ingested γ-butyrolactone remains largely in
its nonhydrolyzed form in the stomach before passing into
the blood, where it is rapidly hydrolyzed to GHB by the
action of a γ-lactonase. (viii) The concentration equilib-
rium constant of GHB formation at human body tem-
perature is Keq (37 °C) ) 0.40. (ix) Contrary to earlier
d ln Keq
∆ H
r
)
-
(4)
d(1/T)
R
From the equilibrium constants of Table 8, ∆
hydrolysis reaction is calculated to be ∆
mol . This result shows that the acid-catalyzed hydroly-
sis of GBL to form GHB is an endothermic reaction
r
H for the
H ) 3.6 (0.3 kJ
r
-1
(
Scheme 2). Such a conclusion is contrary to that of Coffin
12
and Long, who reported a value of ∆ H ) -710 cal (=-3
r
-
1
r
results, this is an endothermic process, with ∆ H ) 3.6
kJ mol ), implying that the reaction would be exother-
mic.
-
1
kJ mol .
r
Since (i) the Coffin and Long ∆ H value was calculated
by working only at three temperatures and (ii) the Keq
values used by those authors were mean values from
Experimental Section
â-Propiolactone, â-butyrolactone, and γ-butyrolactone were
obtained from Sigma. δ-Valerolactone was a Fluka product.
Since the commercial δ-valerolactone may be polymerized up
to 25%, its purity was checked periodically, and when ap-
propriate, it was reconstituted by distillation under reduced
pressure.
The heavy water used to study the solvent isotope effect was
an Aldrich product (99.9% D), and the dioxane used in the
study of the influence of the reaction medium was from
Panreac.
To monitor the reaction kinetics, hydroxy acids resulting
in hydrolysis of the lactones were used as control species. The
concentration of the hydroxy acid was determined by titration
with NaOH, this latter being titrated with potassium hydrogen
phthalate. In the titration procedure, 1 mL aliquots of the
experiments carried out under nonhomogeneous condi-
+
tions (in a 9-fold [H ] range), the Coffin and Long ∆
r
H
value would be affected by an appreciable error.
Because the kinetic experiments were carried out
under pH conditions (pH ∼ 2) comparable to those
existing in the stomach, from the data in Tables 5 and 6
it can be deduced that at human body temperature the
amount of GBL converted into GHB is less than 3% after
2
h (as shown in Figure 3a, 1 week is necessary for the
equilibrium situation to be reached). This means that
when ingested (see Introduction), GBL flows into the
blood in the nonhydrolyzed, nonopened form.
Although GBL is rapidly hydrolyzed in the blood by a
γ-lactonase (with a half-life of less than 1 min3
5-37
), it is
(
38) Kohrs, F. P.; Porter, H. Ann. Emerg. Med. 1999, 33, 475.
(
34) McQuarrie, D. A.; Simon, J. D. Physical Chemistry. A Molecular
(39) Shannon, M.; Quang, L. S. Pediatr. Emerg. Care 2000, 16, 435.
(40) Dupont, P.; Thornton, J. Hum. Exp. Toxicol. 2001, 20, 19.
(41) Lettieri, J.; Fung, H. Res. Commun. Chem. Path. Pharmacol.
1978, 22, 107.
Approach; University Science Books: Sausalito, CA, 1997; pp 1063-
1
064.
(
35) Giarman, N. J.; Roth, R. H. Science 1964, 145, 583.
36) Roth, R. H.; Giarman, N. J. Biochem. Pharmacol. 1966, 15,
(
(42) Cash, C. D. Neurosci. Biobehav. Rev. 1994, 18, 291.
(43) Roth, R. H.; Delgado, J. M. R.; Giarman, N. J. Int. J. Neuro-
pharmacol. 1966, 5, 421.
1
333.
(37) Roth, R. H.; Giarman, N. J. Biochem. Pharmacol. 1965, 14, 177.
J. Org. Chem, Vol. 70, No. 2, 2005 425

P e´ rez-Prior et al.
reacting mixture (lactone and unbuffered CO
2
-free water in
Acknowledgment. The authors thank the Spanish
the case of neutral hydrolysis; lactone and diluted HCl
solutions for acid-catalyzed reactions) were removed from time
to time, added to 9 mL of ice water, and titrated immediately
Ministerios de Ciencia y Tecnolog ´ı a (Project BQU2001-
1
934) and Educaci o´ n y Ciencia (Project CTQ2004-05048/
BQU), as well as the Spanish Junta de Castilla y Le o´ n
(Grant SA003/02) for supporting the research reported
in this article. M.T.P.P. and J.A.M. also thank, respec-
tively, the Ministerio de Ciencia y Tecnolog ´ı a and the
Junta de Castilla y Le o´ n for Ph.D. grants.
a
to a bromthymol blue endpoint (the hydroxy acid pK values
4
4
are in the range 4.5-4.7 ). Kinetic runs were performed in
triplicate. Temperatures were kept constant to (0.04 °C using
a Lauda-Ecoline RE 120 thermostat.
(
44) CRC Handbook of Chemistry and Physics, 83rd ed.; CRC
Press: Boca Rat o´ n, FL, 2003.
JO040271I
426 J. Org. Chem., Vol. 70, No. 2, 2005