Importance of equilibrium fluctuations between most stable conformers in the control of the reaction mechanism

Article abstract of DOI:10.1021/jo101366m

Hydrolysis of closely related compounds show how subtle structural differences markedly change reaction mechanisms. While in the hydrolysis of 3-acetoxy-2-naphthoic acid (3AC2NA) the reacting groups rotate freely, favoring intramolecular general base catalysis, the 1-acetoxy-2-naphthoic acid (1AC2NA) isomer is caged in an energy wall that freezes a conformation suitable for intramolecular nucleophilic attack, in contrast to the results expected for reactions governed largely by electronic effects. The results highlight the importance of the dynamics of equilibrium fluctuations between most stable conformers in the control of the reaction mechanism, (i) promoting the nucleophilic attack in 1AC2NA by allowing the most stable conformers to equilibrate only via rotation in a direction that intercepts the reaction coordinate and (ii) favoring a general base-catalyzed water attack in 3AC2NA by favoring equilibration via rotation that allows inclusion of a water molecule in a proper position for reaction.

Full text of DOI:10.1021/jo101366m

pubs.acs.org/joc
Importance of Equilibrium Fluctuations between Most Stable Conformers
in the Control of the Reaction Mechanism
Bruno S. Souza and Faruk Nome*
´
ꢀ
Departamento de Quımica, Universidade Federal de Santa Catarina, Florianopolis,
Santa Catarina 88040-900, Brazil
faruk@qmc.ufsc.br
Received July 12, 2010
Hydrolysis of closely related compounds show how subtle structural differences markedly change
reaction mechanisms. While in the hydrolysis of 3-acetoxy-2-naphthoic acid (3AC2NA) the reacting
groups rotate freely, favoring intramolecular general base catalysis, the 1-acetoxy-2-naphthoic acid
(1AC2NA) isomer is caged in an energy wall that freezes a conformation suitable for intramolecular
nucleophilic attack, in contrast to the results expected for reactions governed largely by electronic
effects. The results highlight the importance of the dynamics of equilibrium fluctuations between
most stable conformers in the control of the reaction mechanism, (i) promoting the nucleophilic
attack in 1AC2NA by allowing the most stable conformers to equilibrate only via rotation in a
direction that intercepts the reaction coordinate and (ii) favoring a general base-catalyzed water
attack in 3AC2NA by favoring equilibration via rotation that allows inclusion of a water molecule in
a proper position for reaction.
Introduction
SCHEME 1. Suggested Mechanisms for Decomposition of the
Acetyl Salicylates According to References 1 and 2
Hydrolyses of acetyl salicylic acid (aspirin) and its derivatives
are classical examples of intramolecular catalysis.1 In the pH-
independent region of rate constants, the carboxylate group acts
either as a general base or as a nucleophile, and this behavior is
highly dependent on electronic effects.2 In the parent compound
the carboxylate group is a general base catalyst, whereas in the
hydrolysis of the 3,5-dinitro derivative it is a nucleophile forming
a salicylic anhydride intermediate (Scheme 1). To explain this
behavior, Fersht and Kirby proposed that the tetrahedral
intermediates involved in the nucleophilic attack revert to
reagents when the leaving group ability is relatively low, as
with aspirin. In the highly activated acetyl 3,5-dinitro salicylic
acid the nucleophilic path is favored as a result of the similar
These observations are consistent with those of Gold et al.,3
who showed that intermolecular nucleophilic catalysis by
acetate ion is observed in the hydrolysis of acetate esters of
strongly acidic phenols, such as 2,4- and 2,6-dinitrophenol,
with formation of acetic anhydride. Conversely, the hydro-
lysis of phenyl and p-methylphenyl acetates are subject solely
to general base catalysis. It was proposed that nucleophilic
attack by the acetate group occurs only for esters whose
2b
0
pKa values of the leaving group and nucleophile.
(1) (a) Fersht, A. R.; Kirby, A. J. J. Am. Chem. Soc. 1967, 89, 5960–5961.
(b) Fersht, A. R.; Kirby, A. J. J. Am. Chem. Soc. 1967, 89, 5961–5962.
(2) (a) Fersht, A. R.; Kirby, A. J. J. Am. Chem. Soc. 1967, 89, 4857–4863.
(b) Fersht, A. R.; Kirby, A. J. J. Am. Chem. Soc. 1968, 90, 5818–5826.
(c) Fersht, A. R.; Kirby, A. J. J. Am. Chem. Soc. 1968, 90, 5826–5832.
(3) Gold, V.; Oakenful., Dg; Riley, T. J. Chem. Soc. B 1968, 515–519.
7186 J. Org. Chem. 2010, 75, 7186–7193
Published on Web 10/08/2010
DOI: 10.1021/jo101366m
r
2010 American Chemical Society

Souza and Nome
JOCArticle
SCHEME 2. Proposed Mechanism for Hydrolysis of the Phosphodiester BMIPP, Involving Intramolecular General Acid Catalysis by
the Imidazolium Group and Nucleophilic Catalysis by the Imidazole Free Base
0
leaving group pKa is less than 3 units higher than that of
acetic acid.3
Recently, we reported a complete mechanistic study for
the hydrolysis of BMIPP (bis(2-(1-methyl-1H-imidazolyl)-
phenyl) phosphate), which like enzyme active sites deploying
two histidines has two catalytic groups in potentially pro-
ductive proximity of the unreactive phosphodiester center.4
We have found that BMIPP is hydrolyzed nearly 1011 times
faster than expected for diphenyl phosphate. The significant
combined effect of the two catalytic groups was explained in
terms of two factors: (i) an initial step that includes intra-
molecular nucleophilic catalysis (NUC) by the neighboring
imidazole free base, concerted with (ii) the expected efficient
intramolecular general acid catalysis (IGAC). Taken together,
factors i and ii result in the formation of the arylimidazole
Me-3 and a phosphorane intermediate (INT), which decom-
poses to the phosphate monoester Me-2 (Scheme 2).4
FIGURE 1. pH-rate profile for hydrolyses of 1AC2NA (black
squares) and 3AC2NA (open circles) at 45.0 °C and μ = 1 M (KCl).
The solid lines represent fits to the experimental data using eq 1.
Regions i, ii þ iii, iii, and iv were marked to indicate the main reac-
tions contributing to kobs, as defined in Scheme 3.
In the absence of a steric or stereoelectronic barrier, intra-
molecular nucleophilic catalysis is reliably more efficient than
intramolecular general base catalysis (IGBC), with effective
molarities of up to 109 in conformationally flexible systems,
compared with a limit of about 60 M for IGBC,5 and this
limit applies in particular to all known model systems invol-
ving water as the primary nucleophile.
Intramolecular reactions depend on the spatial relation-
ship of the reacting groups and time spent at the critical
reaction distance.6 To explore the importance of steric and
conformational effects in these well-known intramolecular
reactions, we examined hydrolysis in two structurally related
compounds, namely, 1-acetoxy-2-naphthoic acid (1AC2NA)
and 3-acetoxy-2-naphthoic acid (3AC2NA). Despite the appa-
rently similar stereochemistry of 1AC2NA and 3AC2NA, the
carboxylate groups of the two esters catalyze hydrolysis via
completely different mechanisms.
SCHEME 3. Proposed Reactions for Hydrolyses of 1AC2NA
and 3AC2NA, Involving Intramolecular Catalysis (Paths ii and iii)
and Specific Acid- and Base-Catalyzed Reactions (Paths i and iv)
form as that of aspirin2c hydrolysis. Kinetic data for both
compounds are consistent with Scheme 3, allowing deriva-
tion of eq 1.
þ
-
-
þ
kobs ¼ kH ½Hꢀ þ kOH ½OHꢀ
Results and Discussion
!
ꢀ
ꢁ
10- pH
Ka
10- pH þ Ka
Kinetics. pH-rate profiles for the hydrolysis of 1-acetoxy-
2-naphthoic acid (1AC2NA) and 3-acetoxy-2-naphthoic acid
(3AC2NA) are similar at 45.0 °C (Figure 1) and of the same
þ ka
þ kb
ð1Þ
10- pH þ Ka
(4) Orth, E. S.; Brandao, T. A. S.; Souza, B. S.; Pliego, J. R.; Vaz, B. G.;
Eberlin, M. N.; Kirby, A. J.; Nome, F. J. Am. Chem. Soc. 2010, 132, 8513–
8523.
(5) Kirby, A. J.; Hollfelder, F. From Enzyme Models to Model Enzymes;
RSC Publishing: Cambridge, 2009.
(6) (a) Menger, F. M. Pure Appl. Chem. 2005, 77, 1873–1886. (b) Yunes,
S. F.; Gesser, J. C.; Chaimovich, H.; Nome, F. J. Phys. Org. Chem. 1997, 10,
461–465. (c) Bruice, T. C.; Lightstone, F. C. Acc. Chem. Res. 1999, 32, 127–
136. (d) Hur, S.; Bruice, T. C. Proc. Natl. Acad. Sci. U.S.A. 2003, 100, 12015–
12020.
J. Org. Chem. Vol. 75, No. 21, 2010 7187

Souza and Nome
JOCArticle
FIGURE 2. ESI-MS data for hydrolysis products of 1AC2NA in (A) H2O and (B) 20% 18O-labeled water. Corresponding data for hydrolysis
of 3AC2NA in (C) H2O and (D) 20% 18O-labeled water. Arrows indicate m/z 187 and 189 peaks. The full m/z scale is shown in Figures S1 and
S2 in Supporting Information.
TABLE 1. Kinetic Data for Hydrolyses of 1AC2NA and 3AC2NA
Conversely, for the water reaction, the reactivity of 1AC2NA
is 4- to 6-fold greater than that of 3AC2NA, both in the anionic
(kb) and neutral forms (ka).
While the rate constants in the specific acid and base cata-
lysis regions are not surprisingly different from those reported
for the salicylic acid reaction, the inversion of reactivity observed
in the water reaction is unexpected and indicates a change in
mechanism, especially because there is a minor change in the
spatial relations between the neighboring carboxylic and
acetoxy groups.
The values of kobs (= kb) in the pH-neutral plateau region
(around region iii in Figure 1) are considerably higher than
for the hydrolysis of phenyl acetate,7 and we focused our
studies on the mechanism of hydrolyses of the anionic forms
of the esters, which given the shapes of the pH-rate constant
profiles in the plateau region could involve either nucleophi-
lic or general base catalysis. Because the two mechanisms
are kinetically undistinguishable, we examined in the pH-
independent plateau region the hydrolyses of 1AC2NA and
3AC2NA in D2O and in 20% H218O.
Isotopic Labeling, Kinetic Isotope Effect and Thermody-
namic Parameters. Hydrolyses of 1AC2NA and 3AC2NA in
18O-labeled water were examined as described elsewhere,2a
and products were analyzed by ESI-MS in negative mode
with H2O/CH3CN 1:9 as mobile phase. The same procedure
was followed in H2O for comparison, and the data are given
in Figure 2.
at 45 °C and μ = 1 M (KCl)
104 ꢁ kHþ
kOH-
(M-1 s-1
104 ꢁ ka
104 ꢁ kb
ester
pKa
(M-1 s-1
)
)
(s-1
)
(s-1
)
1AC2NA
3AC2NA
2.79
3.43
0.29
1.55
0.17
0.77
0.04
0.01
0.80
0.15
For both compounds, kb is the first order rate constant
for hydrolysis in the pH-independent region (region iii in
Figure 1) and ka is the first order rate constant for the minor
contribution of the COOH form (region ii in Figure 1),
kHþ and kOH- are the usual second order rate constants for
specific acid and base catalysis whose major contributions
are in regions i and iv, respectively, and Ka is the acid dis-
sociation constant of the corresponding o-acetoxynaphthoic
acid. The initial products are acetic acid/acetate and
the corresponding 1-hydroxy-2-naphthoic (1OH2NA) and
3-hydroxy-2-naphthoic (3OH2NA) derivatives. In the acid
region, the initially formed 1OH2NA slowly decomposes to
form 1-naphthol, identified by comparison with an
authentic standard.
Table 1 shows the derived rate constants and pKa’s
obtained from the pH-rate profile by using eq 1. In both
compounds, both carboxyl and carboxylate groups catalyze
ester group hydrolysis, showing a behavior qualitatively
similar to that observed for aspirin.2c Hydrolysis of the anion
(kb) is 20 and 15 times faster than those of the neutral forms
(ka) for 1AC2NA and 3AC2NA, respectively.
Figure 2 shows the ratio of intensities of the M þ 2 (m/z
We note that rate constants kHþ and kOH- for the respective
specific acid and base catalysis for hydrolyses of 3AC2NA
are about 5.3- and 4.5-fold greater, respectively, than the
corresponding rate constants for hydrolysis of 1AC2NA.
189) and M (m/z 187) peaks, relative to those in samples with
(7) Kirby, A. J.; Lloyd, G. J. J. Chem. Soc., Perkin Trans 2 1976, 1753–
1761.
7188 J. Org. Chem. Vol. 75, No. 21, 2010

Souza and Nome
JOCArticle
SCHEME 4. Proposed Mechanisms Showing (i) Nucleophilic Intramolecular Attack in 1AC2NA with Incorporation of 18O in 1OH2NA
and (ii) General Base Catalysis for 3AC2NA, with Incorporation of 18O in the Acetic Product; *O Indicates 18
O
TABLE 2. Kinetic Isotope Effects and Activation Parameters for
Hydrolyses of 1AC2NA and 3AC2NA and Labeled Oxygen Incorporation
into Products 1OH2NA and 3OH2NA at pH 6.5 (Phosphate Buffer,
μ = 1.0 M)
in the acetate anion. Taken together, the results presented in
Figure 2 show that these apparently similar esters are hydro-
lyzed through very different mechanisms. The activation
parameters are those expected for the proposed mechanistic
routes, showing a small positive entropy of activation for the
intramolecular reaction of 1AC2NA (route i in Scheme 4)
and a larger and negative entropy of activation for the
general base mechanism proposed for the 3AC2NA ester
(route ii in Scheme 4).2a,b
1AC2NA
3AC2NA
%
18O incorporationa
20
1.16
26.48
27.19 ( 0.97
3.3 ( 2.4
0
2.42
27.06
19.27 ( 0.48
-26.1 ( 1.6
b
kH O/kD O
2
2
ΔGq at 25 °C (kcal mol-1 c
ΔHq (kcal mol-1 d
ΔSq (cal mol-1 K-1 e
)
)
)
aCalculated from the differences in the m/z 189 peak for hydrolyses in
The small kinetic isotope effect, kH O/kD O = 1.16,
2
2
coupled with the quantitative incorporation of 18O in the
reaction product from 1AC2NA, demonstrates the nucleo-
phiIic catalysis pathway, thus ruling out a general base
mechanism with participation of water as a reactant in the
rate-limiting step. The absence of a solvent hydrogen isotope
effect shows that attack of water in the second step of
reaction i in Scheme 4 must be fast relative to anhydride
formation. The 18O incorporation indicates that in the
reaction of 1AC2NA the attack of water on the anhydride
intermediate must occur almost exclusively at the carbonyl
group that is bonded to the aromatic ring given that the
content of 18O in the solution is 20%. Because in related
compounds, such as acetic benzoic anhydride, there is only
25% attack at the aromatic carbonyl carbon, the results
show that general base assistance by the neighboring phe-
nolate generated in the first step plays an important role,
favoring attack on the aromatic carbonyl center. Since
protonation of 2AN anion should proceed readily under
the experimental conditions, general acid catalysis could also
contribute to favor the regioseletive water attack. The ob-
served kinetic isotope effect in the hydrolysis of 3AC2NA,
kH O/kD O = 2.42, is typical of reactions proceeding via
natural vs 20% 18O-labeled water. bValues are (3%, at 45 °C. Esti-
c
mated using ΔGq = RT ln[(kBT)/(hkb)], where kB and h are the
Boltzmann and Planck constants, respectively, kb is taken from the
Eyring plot at 25 °C, and T is the temperature. dObtained from the
e
Eyring plot given in Figure S3 in Supporting Information. Estimated
using ΔGq = ΔHq - TΔSq, at 25 °C.
natural abundance, thereby providing information on the
extent of incorporation of H218O into the products of the
hydrolyses of 1AC2NA (Figure 2A and B) and 3AC2NA
(Figure 2C and D). Clearly, the two esters exhibit dramati-
cally different behavior, as summarized in Table 2. Compar-
ing panels A and B of Figure 2, it can be clearly seen that in
the hydrolysis of 1AC2NA a large increase in the M þ 2 peak
is observed in the reaction conducted in 18O-enriched water;
this large increase allowed us to calculate the incorporation
of 18O in 1OH2NA given in Table 2. Conversely, panels C
and D of Figure 2 show basically identical M and M þ 2
peaks for 3OH2NA, showing that 18O is not incorporated in
the hydrolysis of 3AC2NA. Also included in Table 2 are
solvent hydrogen kinetic isotope effects (kH O/kD O) and
2
2
activation parameters for both hydrolyses in the plateau
region (pH 6.5).
2
2
In the case of 3AC2NA hydrolysis, the solvent 18O isotopic
labeling results are fully consistent with a mechanism involving
general base catalysis where no labeled oxygen is incorporated
into the product, 3OH2NA. As expected from the proposed
general base catalysis, 18O isotopic labeling for the hydrolysis
of 3AC2NA showed incorporation of 18O into the acetate
anion product. By contrast, the isotopic labeling results for
the hydrolysis of 1AC2NA are fully consistent with a me-
chanism involving nucleophilic attack. In fact, the incor-
poration of labeled oxygen into the 1OH2NA product is
compelling evidence for the formation of the anhydride
intermediate 2AN (Scheme 4) in the hydrolysis of 1AC2NA
in the plateau region. That is, the 1AC2NA reaction pro-
ceeds via nucleophilic catalysis without incorporation of 18O
general base catalysis; this kinetic isotope effect is also
consistent with the incorporation of 18O into the acetate
product (reaction ii in Scheme 4).
This mechanistic divergence cannot be explained in terms
of simple electronic effects because the second pKa’s of
1OH2NA and 3OH2NA (about 13.0 and 12.5, respectively,
obtained by titration; see Figures S4 and S5 in Supporting
Information) are not significantly different. Presumably
significant ground state differences in 1AC2NA and 3AC2NA
control the different mechanisms.
Computational Calculations. On this basis, we constructed
the relaxed potential energy surface (RPES) for anionic
1AC2NA and 3AC2NA at the B3LYP/6-31þG(d) level,
locating all conformations associated with rotations of the
J. Org. Chem. Vol. 75, No. 21, 2010 7189

Souza and Nome
JOCArticle
FIGURE 3. B3LYP/6-31þG(d) relaxed energy potential surface for the anionic (A) 1AC2NA and (B) 3AC2NA obtained by twisting the ester
and carboxylate groups relative to the aromatic mean plane. Numbering is according to Scheme 5.
SCHEME 5. Geometrical Features and Numbering Used
To Construct the RPESs Shown in Figure 3; Arrow Indicates
the 8-Hydrogen in 1AC2NA
the ester group is pushed toward the carboxylate group, and
the reactive groups are forced to reside longer in a conformation
suitable for intramolecular nucleophilic attack. It is interest-
ing to note that the global minima structures of 1AC2NA and
3AC2NA show suitable conformations for intramolecular
nucleophilic attack (Figure 4 and geometrical parameters
given in Table 3) and the most stable conformers have ade-
quate conformations to be the more reactive ones.6a-c In
˚
these structures, the O7-C4 distance is 2.67 and 2.66 A and
the O7-C4-O6 angle is 103.96° and 104.03° for 1AC2NA
and 3AC2NA, respectively. Table 3 also presents the geome-
trical details of 1AC2NA and 3AC2NA optimized by using
the polarizable continuum model (PCM) to mimic the water
environment. As expected, on going from the gas phase to
the polarizable continuum the carboxylate negative charge is
stabilized, and there is a considerable increase in the O7-C4
distance. In addition the O7-C4-O6 angle deviates from the
ester and carboxylate groups relative to the mean aromatic
ring plane (Scheme 5). We refer to the RPES because every
point in the 3D grid represents an optimized structure in which
two dihedral angles are frozen, while all other bond dis-
tances, angles, and dihedrals were set free in order to obtain
local minima. The results are in Figure 3 and show the energy
costs for rotation of the reactive groups.
€
Burgi-Dunitz angle, reflecting a poorer interaction between
The RPES shows that in 3AC2NA the maximum energy
conformer has a relative energy of about 12 kcal mol-1 and is
located in the plane where the C2-C3-O5-C4 dihedral angle
is about 0° as a result of the repulsive interaction between the
carboxylate and ester groups, with the latter coplanar to the
aromatic ring. Besides this unfavorable interaction, all other
conformers are within a few kilocalories, and 75% of the energy
map contains structures within 5 kcal mol-1 of the most
stable conformer, so that the ester group rotates almost freely.
Conversely, in 1AC2NA, as well as a similar maximum
energy conformer in the plane of the C2-C3-O5-C4
dihedral angle of 0°, there is a high barrier associated with
angles higher than 150° for the ester/ring dihedral. Thus, the
ester group does not rotate freely and only 48% of the map is
populated by structures within 5 kcal mol-1 energy. This
high energy wall is structurally associated with peri-interac-
tions with the 8-hydrogen (see arrow in Scheme 5), which acts
as a bumper, pushing back the ester group. Thus, in 1AC2NA,
the carboxylate and the ester carbonyl groups. Nevertheless,
when compared, both in gas and in the PCM phase, there is
no major difference between 1AC2NA and 3AC2NA that
would give key information regarding differences in the
hydrolysis mechanisms of both esters. The crystal structure
of the neutral acid 1AC2NA8 is qualitatively consistent with
that calculated, especially in relation to the stereochemistry
and distances between ester and carboxylic acid groups, as
found for aspirin,9 which is structurally similar to 3AC2NA.
The most stable conformers of 1AC2NA and 3AC2NA
show conformations that are strongly favorable for intra-
molecular nucleophilic attack. This observation is consistent
with the experimental results obtained for the hydrolysis of
1AC2NA; on the other hand, this observation undermines
(8) Souza, B. S.; Bortoluzzi, A. J.; Nome, F. Acta Crystallogr., Sect. E
2007, 63, O4523–U3325.
(9) Wilson, C. C. New J. Chem. 2002, 26, 1733–1739.
7190 J. Org. Chem. Vol. 75, No. 21, 2010

Souza and Nome
JOCArticle
A detailed examination of the ∼800 analyzed structures used
to build the energy potential surface showed that in both
compounds there was a probability of about 90% of the
˚
O7-C4 distance being between 2.4 and 3.0 A. Statistical
analysis of the O7-C4 distance showed that the distribution
was slightly narrower for 1AC2NA with full width at half-
˚
maximum (fwhm) of 0.22 A, compared with a fwhm of
˚
0.38 A for the 3AC2NA ester (see Supporting Information,
Figures S6 and S7), a result consistent with the less restrained
structure of 3AC2NA.
The only significant difference between both compounds
is the 13 kcal/mol-1 barrier associated with forcing the ester
group to achieve angles higher than 150° for the ester/ring
dihedral angle in the 1AC2NA ester. This barrier should be
compared with the small 3 kcal/mol barrier in 3AC2NA,
which allows the ester group to rotate almost freely from one
side of the ring to the other. The dynamics of the process can
be examined by considering rates of interconversion of
acyclic conformers,10 and considering ΔS ≈ 0 for the iso-
merization process, eq 2 can be used to estimate the inter-
conversion constant k, where ΔH ≈ ΔEEl and is in calories
per mole.
FIGURE 4. B3LYP/6-311þþG(2df,p) global minima of 1AC2NA
anion and 3AC2NA anion calculated in (A) the gas phase and
(B) using the PCM to mimic the water environment. Main geo-
metrical parameters are given in Table 3.
k ¼ 2:084 ꢁ 1010Te- ΔH=1:986T
ð2Þ
Equilibration of conformers via rotation of the acetoxi
group on the direction from -80° f 0° f 80° favors the
intramolecular reaction, since intercepts the actual reaction
coordinate for nucleophilic attack, and the energy barrier is
about 5.5 kcal mol-1 for both compounds, which corre-
sponds to a rate of intercorversion of ca. 5.7 ꢁ 108 s-1 at
25 °C. Conversely, equilibration of conformers via rotation
of the acetoxy group on the direction from -80° f -180° f
80°, for the 3AC2NA isomer, needs to surpass a small barrier
of 3.8 kcal mol-1 (see Figure 3B), which corresponds to a rate
of interconversion (between conformers where the C2-C3-
O5-C4 dihedral angle is about 80° to conformers with the
same dihedral at -80°) of about 1.0 ꢁ 1010 s-1 at 25 °C. The
equivalent rate for 1AC2NA, which passes over a high energy
wall (about 13 kcal mol-1) due to the peri-interactions with
TABLE 3. Selected Distances and Angles for Optimized Structures
of 1AC2NA and 3AC2NA Calculated at the B3LYP/6-311þþG(2df,p)
Level in the Gas Phase and by Using PCM To Mimic the Water
Environmenta
1AC2NA
3AC2NA
gas phase
PCMa
gas phase
PCMa
O7-C4
2.69
1.25
2.75
2.99
1.25
2.88
2.68
1.25
2.74
3.10
1.25
2.91
O7-C1
O7-O5
O7-C4-O6
O7-C1-C2-C3
C2-C3-O5-C4
104.16
27.34
79.39
95.35
47.65
96.60
104.18
24.80
80.04
94.90
50.57
100.59
aAtomic radii obtained from the Simple United Atom Topological
Model.
the 8-hydrogen (vide supra), decreases to 7.7 ꢁ 103 s-1
.
the IGBC mechanism observed for 3AC2NA. A closer
examination of the RPES shows that, for both esters, the
most stable conformers show contact distances consistent
with Menger’s spatiotemporal hypothesis. This hypothesis
states that reactions are fast when the reactive groups are
held rigidly at a contact distance (less than the diameter of
Clearly, the dynamics of the equilibrium between confor-
mers provides key insights to understand the difference in
mechanisms. In the hydrolysis reaction of 1AC2NA, the
most stable conformer is in a conformation where there is
not enough space to position a water molecule for a general
base-catalyzed reaction to proceed. The dynamics of the
nucleophilic reaction involves rotation of the acetoxy group
in the direction -80° f 0° f 80°, thus intercepting the
reaction coordinate for nucleophilic attack. Rotation in this
direction proceeds with an interconversion rate of 5.7 ꢁ 108,
which is 7.4 ꢁ 104-fold faster than the interconversion rate
needed for a plausible water reaction (where rotation of the
acetoxy group in the direction of the peri-hydrogen is needed
to include a water molecule between the reacting groups),
which requires overcoming a 13 kcal mol-1 wall, correspond-
ing to a rate of 7.7 ꢁ 103 s-1. Obviously, the nucleophilic
reaction is highly favored by the energy wall. A similar
analysis for the 3AC2NA isomer shows that in the 3AC2NA
case, the reverse is true. It is less expensive in terms of energy
˚
water, ca. 3 A); thus, imposing time and distance constraints
on two reactants necessarily brings them into proximity and
speeds the reaction.6a In a sense, although the spatiotemporal
concept does not account quantitatively for the experimental
observations, it allows an initial qualitative analysis that is
useful for understanding the differences in the mechanism
for the two reactions being studied.
Dynamics of Conformer Equilibrium on the RPES. The
RPES reveals that in both esters, 1AC2NA and 3AC2NA,
high energy conformers with relative energies close to 12 kcal
mol-1 arise from repulsive interactions between the carbox-
ylate and ester groups, with the latter lying in the aromatic
plane. Moreover, when the ester/ring dihedral is between
75° and 85° (C2-C3-O5-C4), the carboxylate group
rotates freely, and the equilibria between the different struc-
tures proceed on a relatively flat potential energy surface.
(10) Eliel, E. L.; Wilen, S. H.; Mander, L. N. Stereochemistry of Organic
Compounds; Wiley: New York, 1994.
J. Org. Chem. Vol. 75, No. 21, 2010 7191

Souza and Nome
JOCArticle
to move the acetoxy group, for the equilibration of confor-
mers, via rotation in the direction -80° f -180° f 80°, with
a small barrier of 3.8 kcal mol-1 and a rate of interconversion
of 1.0 ꢁ 1010 s-1, which is ca. 20-fold faster than the rate of
interconversion in the direction -80° f 0° f 80°, which as
described above favors the nucleophilic attack. Rotation in
the direction -80° f -180° f 80° increases the distance
between the reacting groups and helps to position a water
molecule in the correct conformation for the general base
catalysis reaction.
negative mode, mobile phase H2O/CH3CN 1:9): 270 (17.80%,
M - H þ CH3CN), 229 (25%, M - H), 187 (100%, M -
CH3CO). The complete FTIR, ESI-MS, and 1H NMR (400
MHz, acetone-d6) spectral data for both 1AC2NA and 3AC2NA
are given in Supporting Information.
Kinetics. Reactions followed spectrophotometrically were
started by adding 20 μL of a stock solution of the substrate
(0.01 M in acetonitrile) to a buffered aqueous solution (3 mL) at
the appropriate pH. Kinetics were followed, at constant tem-
perature and ionic strength 1.0 M (KCl), for at least five half-
lives, by monitoring the appearance of 1OH2NA (344 nm for pH
values smaller than 3.40 and 314 nm for pH values higher than
3.40) or 3OH2NA (350 nm) on a spectrophotometer equipped
with a thermostatted cell holder. Activation parameters were
calculated by using the Eyring equation from rate constants at
pH 6.65 (phosphate buffer, 0.01 M, μ = 1.0M). The pH values
of the reaction mixtures were measured at the end of each run.
Observed first order rate constants (kobs) were calculated by
nonlinear least-squares fitting of the absorbance versus time;
correlation coefficients were 0.998 or better. Buffers were HCl/
KCl (pH e 2), chloroacetate (pH 2.0-3.5), formate (pH 3.0-
4.0), acetate (pH 4.0-5.5), phosphate (pH 5.5-7.0), bis-TRIS
(pH 7.0-9.0), carbonate (pH 9.5-11.0), and NaOH (pH g 12).
Data from kinetic measurements are summarized in Tables
S1-S2 in Supporting Information. The solvent isotope effect
was calculated from the ratio of kobs in H2O and D2O in
phosphate buffer at pH 7.0 or pD = 7.3,12 μ = 1.0 M (KCl),
and 45 °C.
Labeling Experiments. Hydrolysis of 1AC2NA and 3AC2NA
in the presence of 18O-isotope labeled water was performed as
described elsewhere.2a A 1-mL glass ampule containing 5 mg of
1AC2NA or 3AC2NA, 82 mg of sodium acetate, 0.2 mL of THF,
0.1 mL of HCl 1 M, and 0.6 mL of 20 atom % enriched H218O
was sealed and incubated at 45.00 °C for 5t1/2 (at pH 5.00). The
products were analyzed by ESI-MS in negative mode using
H2O/CH3CN 1:9 as mobile phase, and the reaction was also
carried out in unlabeled water following the same procedure
with no significant m/z 189 peak.
Conclusions
These closely related compounds show how subtle struc-
tural differences may lead to major changes in reaction
mechanism that, although kinetically equivalent, could be
distinguished by means of isotopic labeling, kinetic isotope
effects, and activation parameters. Such differences in
mechanisms are certainly present in many enzymatic reactions,
where reaction mechanism may well be controlled by con-
formational effects, which are optimized to have an appro-
priate time window for the reaction to proceed. In the
reaction of the conformationally unrestrained 3-acetoxy
compound, there is intramolecular general base catalysis.
Conversely, for the 1-acetoxy derivative, a conformational
constraint due to peri-interactions forces the reacting car-
boxylate and ester pair to remain longer in conformations
that favor a reaction route proceeding via intramolecular
nucleophilic attack. The results highlight the importance
of the role of the dynamics of equilibrium between most
stable conformers in the control of the reaction mechanism,
(i) promoting the nucleophilic attack in 1AC2NA by allowing
the most stable conformers to equilibrate only via rotation in
a direction that intercepts the reaction coordinate and (ii)
favoring a general base-catalyzed water attack in 3AC2NA
by favoring equilibration via rotation that allows inclusion
of a water molecule in a proper position for reaction.
Mass Spectrometry. Mass spectroscopic identifications were
performed in a low-resolution instrument with the direct injec-
tion mode of 10-5 M sample solutions as carried out in other
studies.13 The interface, CDL, and block temperatures were set
at 250, 250, and 200 °C, respectively. The detector was main-
tained at 1.50 kV, the flow of N2 was 1.5 L min-1, and the mobile
phase was 1:9 H2O/CH3CN
Experimental Section
Materials. 1-Acetoxy-2-naphthoic acid (1AC2NA) was pre-
pared by a procedure similar to that reported by Bergeron
et al.11 Concentrated sulfuric acid (10 drops) was added to a
refluxing mixture of 1-hydroxy-2-naphthoic acid (3.50 g, 18.6
mmol) in acetic anhydride (8 mL, 89.7 mmol). The mixture was
kept under reflux for 10 additional minutes, and after cooling to
room temperature, the pale solid was filtered off and recrystal-
lized in aqueous ethanol. The pale crystals melt at 138-139 °C.
1H NMR (400 MHz, acetone-d6, TMS): δ 8.16 (1H, d, J = 8.26);
8.07 (1H, d, J = 8.67); 8.03 (1H, d, J = 7.97); 7.91 (1H, d, J =
8.67); 7.70 (1H, t, J = 7.43); 7.65 (1H, t, J = 7.65), 2.46 (3H, s).
νmax (KBr) 3080-2600, 1766, 1683, 1296, and 1207 cm-1. m/z
(ESI-MS, negative mode, mobile phase H2O/CH3CN 1:9): 270
(4,60%, M - H þ CH3CN), 229 (32,5%, M - H), 187 (100%,
M - CH3CO). The isomer 3-acetoxy-2-naphthoic acid (3AC2NA)
was prepared following the same procedure used to prepare
1AC2NA using 3-hydroxy-2-naphthoic (3OH2NA) acid instead
of 1-hydroxy-2-naphthoic acid (1OH2NA). Slightly yellow
crystals, mp 183-184 °C. 1H NMR (400 MHz, acetone-d6,
TMS): δ 8.69 (1H, s); 8.11 (1H, d, J = 8.12); 7.97 (1H, d, J =
8.17); 7.68 (2H, m); 7.61 (1H, t, J = 7.37), 2.30 (3H, s). νmax
(KBr) 3084-2800, 1764, 1700, 1294, 1207 cm-1. m/z (ESI-MS,
1
NMR Spectroscopy. All H spectra were monitored at 400
MHz, 25 °C, in acetone-d6. The 1H chemical shifts are referred to
internal TMS.
Computational Methods. All computational calculations were
performed using Gaussian 0314 at the DFT level employing the
hybrid B3LYP15 functional. Global minima structures were
optimized using the 6-311þþG(2df,p) basis set and properly
characterized in force constant calculations by the absence of a
single negative eigenvalue. The construction of the RPESs was
performed using the 6-31þG(d) basis set using the “Scan”keyword,
which requests that a potential energy surface (PES) scan be
done over a rectangular grid involving selected internal coordi-
nates. The selected variables were the mean ring plane/ester
(12) Fife, T. H.; Bruice, T. C. J. Phys. Chem. 1961, 65, 1079–1080.
(13) Kirby, A. J.; Medeiros, M.; Oliveira, P. S. M.; Brandao, T. A. S.;
Nome, F. Chem.;Eur. J. 2009, 15, 8475–8479.
(14) Frisch, M. J., et al. Gaussian 03, revision D.01; Gaussian, Inc.:
Pittsburgh, PA, 2004.
(15) (a) Becke, A. D. J. Chem. Phys. 1993, 98, 5648–5652. (b) Stephens,
P. J.; Devlin, F. J.; Chabalowski, C. F.; Frisch, M. J. J. Phys. Chem. 1994, 98,
11623–11627. (c) Vosko, S. H.; Wilk, L.; Nusair, M. Can. J. Phys. 1980, 58,
1200–1211. (d) Lee, C. T.; Yang, W. T.; Parr, R. G. Phys. Rev. B 1988, 37,
785–789.
(11) Bergeron, R. J.; Wiegand, J.; Wollenweber, M.; McManis, J. S.;
Algee, S. E.; Ratliff-Thompson, K. J. Med. Chem. 1996, 39, 1575–1581.
7192 J. Org. Chem. Vol. 75, No. 21, 2010

Souza and Nome
JOCArticle
dihedral angle and the mean ring plane/carboxylate dihedral
angle as defined above. The calculations were performed using
the “Loose” keyword in order to achieve local minima struc-
tures, with an step size of 8°. Relative energies were calculated
from the electronic energies differences (ΔEEl).
Spectrofluorimetric Titration. The spectrofluorimetric titra-
tion of 1-hydroxy-2-naphthoic acid (1OH2NA) and 3-hydroxy-
2-naphthoic acid (3OH2NA) were performed at room tempera-
ture, by measuring the emission intensity at 318 and 312 nm,
respectively, of a 66.6 μM solution of the hydroxy acids. The pH
was changed by addition of small aliquots of HCl or KOH 1 M
from pH 2.00 to 12.00, and the intensity was corrected by the
dilution factor. Above and under pH 2.00-12.00, standard
KOH or HCl solutions were used, and the ionic strength was
kept constant (1.0 M) by addition of KCl. Results are given in
Supporting Information, Figures S4 and S5.
ꢀ
Acknowledgment. We are grateful to INCT-Catalise,
PRONEX, FAPESC, CNPq and CAPES for support of this
work and to Professor Ricardo L. Longo for its help in con-
structing the RPES.
Supporting Information Available: Characterization of
1AC2NA and 3AC2NA, kinetic details, titration curves of
1OH2NA and 3OH2NA, 2D plots of the RPESs, Cartesian
coordinates of 1AC2NA and 3AC2NA, and Gaussian full ref-
erence. This material is available free of charge via the Internet
at http://pubs.acs.org.
J. Org. Chem. Vol. 75, No. 21, 2010 7193