Method for estimating SN1 rate constants: Solvolytic reactivity of benzoates

Article abstract of DOI:10.1021/jo3013308

Nucleofugalities of pentafluorobenzoate (PFB) and 2,4,6-trifluorobenzoate (TFB) leaving groups have been derived from the solvolysis rate constants of X,Y-substituted benzhydryl PFBs and TFBs measured in a series of aqueous solvents, by applying the LFER equation: log k = s_f(E_f + N_f). The heterolysis rate constants of dianisylmethyl PFB and TFB, and those determined for 10 more dianisylmethyl benzoates in aqueous ethanol, constitute a set of reference benzoates whose experimental ΔG ^? have been correlated with the ΔH^? (calculated by PCM quantum-chemical method) of the model epoxy ring formation. Because of the excellent correlation (r = 0.997), the method for calculating the nucleofugalities of substituted benzoate LGs have been established, ultimately providing a method for determination of the S_N1 reactivity for any benzoate in a given solvent. Using the ΔG^? vs ΔH^? correlation, and taking s_f based on similarity, the nucleofugality parameters for about 70 benzoates have been determined in 90%, 80%, and 70% aqueous ethanol. The calculated intrinsic barriers for substituted benzoate leaving groups show that substrates producing more stabilized LGs proceed over lower intrinsic barriers. Substituents on the phenyl ring affect the solvolysis rate of benzhydryl benzoates by both field and inductive effects.

Full text of DOI:10.1021/jo3013308

Article
pubs.acs.org/joc
Method for Estimating SN1 Rate Constants: Solvolytic Reactivity of
Benzoates
Mirela Matic, Bernard Denegri,* and Olga Kronja*
́
̌ ́
University of Zagreb, Faculty of Pharmacy and Biochemistry, A. Kovacica 1, 10000 Zagreb, Croatia
S
* Supporting Information
ABSTRACT: Nucleofugalities of pentafluorobenzoate (PFB)
and 2,4,6-trifluorobenzoate (TFB) leaving groups have been
derived from the solvolysis rate constants of X,Y-substituted
benzhydryl PFBs and TFBs measured in a series of aqueous
solvents, by applying the LFER equation: log k = sf(Ef + Nf).
The heterolysis rate constants of dianisylmethyl PFB and TFB,
and those determined for 10 more dianisylmethyl benzoates in
aqueous ethanol, constitute a set of reference benzoates whose
experimental ΔG⧧ have been correlated with the ΔH⧧
(calculated by PCM quantum-chemical method) of the
model epoxy ring formation. Because of the excellent correlation (r = 0.997), the method for calculating the nucleofugalities
of substituted benzoate LGs have been established, ultimately providing a method for determination of the SN1 reactivity for any
benzoate in a given solvent. Using the ΔG⧧ vs ΔH⧧ correlation, and taking sf based on similarity, the nucleofugality parameters
for about 70 benzoates have been determined in 90%, 80%, and 70% aqueous ethanol. The calculated intrinsic barriers for
substituted benzoate leaving groups show that substrates producing more stabilized LGs proceed over lower intrinsic barriers.
Substituents on the phenyl ring affect the solvolysis rate of benzhydryl benzoates by both field and inductive effects.
= 0.00 for a dianisylcarbenium electrofuge (X = Y = 4-OCH3)
and sf = 1.00 for a chloride nucleofuge in pure ethanol.
According to eq 1, the nucleofugality (Nf) of a given leaving
group is defined as the negative intercept on the abscissa of the
log k vs Ef correlation line. This approach enables the
estimation of the absolute rates of heterolysis reactions (SN1)
semiquantitatively for various combinations of electrofuge
nucleofuge.4
In this work we have focused our attention toward SN1
solvolysis reactions of fluorinated benzoates, as well as toward
benzoates in general, in which the rate-determining heterolysis
step involves cleavage of the ROC(O)Ar bond, and
formation of the benzhydrylium carbocation intermediate and
the benzoate ion (Scheme 1). Nucleofugalities of the series of
substituted benzoates are determined both experimentally and
by using quantum-chemical calculation, and the influences of
INTRODUCTION
■
A universally valid nucleofugality scale does not exist because
the leaving group ability depends not only on the reaction
mechanism but also on various variables such as different
solvation effects in various solvents in ground and transition
states,1 back strain effects,2 anomeric effect,3 etc. However, a
comprehensive empirical scale in which nucleofugalities are
defined for combinations of leaving groups and solvents has
been developed, based on solvolysis rates of benzhydryl
derivatives.4,5 The nucleofugalities of numerous leaving groups
in given solvents have already been added to this scale.6
According to that approach, the absolute heterolysis rate
constant for any SN1 solvolysis reaction can be expressed with
the following three-parameter linear free energy relationship
(LFER) equation:
log k = sf(Ef + Nf)
(1)
Scheme 1
in which k is the first-order rate constant (s−1) at 25 °C, sf is the
nucleofuge-specific slope parameter, Nf is the nucleofugality
parameter, and Ef is the electrofugality parameter. Nucleofu-
gality and electrofugality are kinetic terms describing the
abilities of the leaving groups to depart with and without the
nonbonding electron pair, respectively. Ef is set up as a solvent-
independent variable that refers to the ability of the substrate
moiety to leave as a carbocation in a heterolysis reaction (SN1).
The nucleofuge-specific parameters (sf and Nf) describe the
leaving group ability in a given solvent. Such an approach
separates the contributions of an electrofuge and a nucleofuge
in the overall solvolytic reactivity. Predefined parameters are Ef
Received: July 19, 2012
Published: September 13, 2012
© 2012 American Chemical Society
8986
dx.doi.org/10.1021/jo3013308 | J. Org. Chem. 2012, 77, 8986−8998

The Journal of Organic Chemistry
Article
Table 1. Solvolysis Rate Constants of X,Y-Substituted Benzhydryl Pentafluorobenzoates and X,Y-Substituted Benzhydryl 2,4,6-
Trifluorobenzoates in Various Solvents at 25 °C
c
k/s−1
a
b
solvent
100E
substrate (X, Y)
Ef
pentafluorobenzoates (PFB)
trifluorobenzoates (TFB)
d
2 (PhO, H)
−3.52
−2.09
−1.32
−0.86
0
1.87 × 10−5
d
3 (MeO, H)
4 (MeO, Me)
5 (MeO, PhO)
6 (MeO, MeO)
1 (Me, H)
(3.16 0.03) × 10−4
(2.06 0.04) × 10−3
(7.69 0.04) × 10−3
2.08 × 10−5
(1.20 0.01) × 10−4
(4.60 0.07) × 10−4
(3.97 0.04) × 10−3
d
90E10W
80E20W
70E30W
−4.63
−3.52
−2.09
−1.32
−0.86
0
5.65 × 10−6
d
2 (PhO, H)
(7.74 0.08) × 10−5
(1.51 0.01) × 10−3
(9.21 0.20) × 10−3
(2.95 0.07) × 10−2
3.60 × 10−6
3 (MeO, H)
4 (MeO, Me)
5 (MeO, PhO)
6 (MeO, MeO)
1 (Me, H)
(7.63 0.07) × 10−5
(5.10 0.04) × 10−4
(1.77 0.01) × 10−3
(1.25 0.02) × 10−2
d
−4.63
−3.52
−2.09
−1.32
−0.86
0
1.60 × 10−5
d
2 (PhO, H)
(1.69 0.03) × 10−4
(3.08 0.05) × 10−3
(1.70 0.02) × 10−2
(3.93 0.09) × 10−2
7.92 × 10−6
3 (MeO, H)
4 (MeO, Me)
5 (MeO, PhO)
6 (MeO, MeO)
1 (Me, H)
(1.55 0.02) × 10−4
(9.49 0.10) × 10−4
(3.04 0.05) × 10−3
(2.06 0.02) × 10−2
−4.63
−3.52
−2.09
−1.32
−0.86
0
(2.86 0.01) × 10−5
(2.71 0.10) × 10−4
(4.71 0.10) × 10−3
(2.21 0.07) × 10−2
d
2 (PhO, H)
1.39 × 10−5
3 (MeO, H)
4 (MeO, Me)
5 (MeO, PhO)
6 (MeO, MeO)
3 (MeO, H)
4 (MeO, Me)
5 (MeO, PhO)
6 (MeO, MeO)
3 (MeO, H)
4 (MeO, Me)
5 (MeO, PhO)
6 (MeO, MeO)
2 (PhO, H)
(2.52 0.06) × 10−4
(1.48 0.01) × 10−3
(3.91 0.02) × 10−3
(2.86 0.03) × 10−2
90A10W
80A20W
70A30W
−2.09
−1.32
−0.86
0
(4.58 0.02) × 10−5
(2.58 0.02) × 10−4
(8.64 0.04) × 10−4
(7.23 0.06) × 10−3
(1.59 0.01) × 10−4
(9.45 0.05) × 10−4
(2.88 0.01) × 10−3
(2.32 0.03) × 10−2
d
6.89 × 10−6
(2.72 0.05) × 10−5
(2.29 0.04) × 10−4
−2.09
−1.32
−0.86
0
(3.70 0.04) × 10−5
(1.13 0.02) × 10−4
(9.76 0.07) × 10−4
d
−3.52
−2.09
−1.32
−0.86
0
1.96 × 10−5
d
3 (MeO, H)
4 (MeO, Me)
5 (MeO, PhO)
6 (MeO, MeO)
2 (PhO, H)
(4.47 0.08) × 10−4
(2.51 0.04) × 10−3
(6.64 0.10) × 10−3
(4.55 0.10) × 10−2
(4.91 0.02) × 10−5
(1.10 0.01) × 10−3
(5.70 0.05) × 10−3
(1.30 0.02) × 10−2
2.03 × 10−5
(1.14 0.03) × 10−4
(3.19 0.03) × 10−4
(2.55 0.04) × 10−3
60A40W
−3.52
−2.09
−1.32
−0.86
0
3 (MeO, H)
4 (MeO, Me)
5 (MeO, PhO)
6 (MeO, MeO)
(5.96 0.03) × 10−5
(3.32 0.04) × 10−4
(8.30 0.10) × 10−4
(6.33 0.06) × 10−3
a
b
Binary solvents are on a volume−volume basis at 25 °C. A = acetone, E = ethanol, W = water. Electrofugality parameters were taken from ref 4.
c
Average rate constants from at least three runs performed at 25 °C unless otherwise noted. Errors shown are biased standard deviations.
d
Extrapolated from data at higher temperatures by using the Eyring equation.
the aromatic ring substitution on overall barriers and intrinsic
barriers are examined. Furthermore, a quantum-chemical model
is proposed here, whose calculated activation enthalpies
correlate very well with experimental activation free energies
for solvolysis of the series of dianisylmethyl (X = Y = 4-MeO)
benzoates, ultimately enabling the prediction of solvolytic
reactivities for a wide spectrum of benzoates with various
substituents.
other leaving groups (LG). Benzhydryl derivatives have been
employed as substrates, at which the substituents on the
benzhydryl moiety have been adjusted to enable kinetic
measurements to be carried out by conventional methods.
Thus, a series of X,Y-substituted benzhydrylium pentafluor-
obenzoates (1-6-PFB) and 2,4,6- trifluorobenzoates (2-6-TFB)
have been subjected to kinetic measurements.
A series of benzhydryl pentafluorobenzoates (1-6-PFB) and
benzhydryl 2,4,6-trifluorobenzoates (2-6-TFB) have been
prepared from the corresponding benzhydrols according to
the methods presented in Experimental Section. The solvolysis
rates have been measured conductometrically in various
solvents at 25 °C. In a few cases the rates have been measured
RESULTS AND DISCUSSION
■
Kinetic Measurements. Equation 1 has been used to
assess the reactivity of fluorinated benzoate anions as leaving
groups in various solvents and to relate their nucleofugalities to
8987
dx.doi.org/10.1021/jo3013308 | J. Org. Chem. 2012, 77, 8986−8998

The Journal of Organic Chemistry
Article
at three different temperatures and extrapolated to 25 °C.
Details are given in Kinetic Methods (Experimental Section).
The first-order rate constants are presented in Table 1.
in aqueous acetones and the lines for trifluorobenzoates are
shown in Figures S1−S4 of the Supporting Information). The
nucleofuge-specific parameters calculated according to eq 1 are
presented in Table 2. The nucleofugalities show that the
reactivity of TFB is similar to that of DNB,4 while PFB reacts
about 1 order of magnitude faster. Systematically higher
reaction constants (sf) in all solvents for less reactive TFBs
than for PFBs confirm that TFB solvolyzes via a later TS in
which the charge separation is more advanced. The slope
parameters (sf) decrease for both nucleofuges as the polarity of
the solvent increases. Such behavior has been observed earlier
and rationalized as diminished solvation of the TS in more
polar solvent due to charge dispersion in both nucleofuge and
electrofuge moieties.7,8
At this point of research, experimental reactivities in a series
of solvents have been available for five different benzoates,
which are as follows: Bz (benzoate), PNB (p-nitrobenzoate),
DNB (3,5-dinitrobenzoate),4 and PFB and TFB presented
above. Because our intention was to correlate the experimental
results with a quantum-chemical model, it was advantageous to
expand the range of experimental reactivities that would be
correlated. Even more important was to gain experimental data
for substrates with structurally versatile leaving groups, those
with one or more electron-withdrawing and electron-donating
substituents being attached to different positions of the phenyl
ring of the benzoate moiety. Reactivities of dianisylmethyl (X =
Y = MeO) derivatives have been chosen for correlation, so the
rates of seven additional dianisylmethyl benzoates have been
determined. To find out the impacts of the enthalpy of
activation (ΔH⧧) and the entropy of activation (ΔS⧧) on the
overall reactivity, the solvolysis rates have further been
determined at three different temperatures in 90% aq ethanol.
The kinetic results are presented in Table S1 (Supporting
Information). This expanded set of 12 reference benzoates
covers various structures in which electron-accepting groups
(NO2, CN, CF3) and atoms (F, Cl), as well as electron-
donating groups (MeO, Me), are substituents on the phenyl
ring of the benzoate moiety. Also, the number of substituents
attached to the phenyl ring varies from 0 to 5.
The nucleofugality parameters (Nf) and the slope parameters
(sf) for PFB and TFB measured in various solvents have been
determined according to eq 1 by correlating the logarithms of
the first-order rate constants and the corresponding Ef values.
Plots of log k against Ef obtained for solvolysis of substituted
benzhydryl pentafluorobenzoates in ethanol−water binary
solvents are presented in Figure 1 (correlation lines obtained
Quantum-Chemical Model for Reference Leaving
Group Effects. Nucleofugality has often been associated
with the basicity of the leaving groups because a good linear
correlation between the reactivity of LGs (log k) and the pKa
values of the conjugated acids of the LGs has been obtained for
some substrates. The heterolysis barrier height in solvolysis is
defined by two factors: the intrinsic barrier (Λ) and the change
in free energy (ΔG°). It is not surprising that numerous
Figure 1. Plots of log k (25 °C) versus Ef for solvolysis of substituted
benzhydryl pentafluorobenzoates in ethanol and aqueous solvents of
ethanol. Solvent mixtures are given as v/v; E = ethanol and W = water.
Table 2. Nucleofugality Parameters Nf and sf for Pentafluorobenzoate (PFB) and 2,4,6-Trifluorobenzoate (TFB) in Various
Solvents
PFB
TFB
a
b
b
b
b
solvent
100E
Nf
sf
Nf
sf
−1.41 0.19
−0.76 0.06
−0.68 0.03
−0.58 0.01
−2.04 0.07
−1.58 0.02
−1.41 0.01
−1.15 0.07
0.97 0.05
0.97 0.02
0.90 0.01
0.87 0.01
1.06 0.02
1.04 0.01
0.96 0.01
0.92 0.02
−2.19 0.10
−1.90 0.08
−1.75 0.07
−1.68 0.06
−3.18 0.19
−2.80 0.01
−2.60 0.06
−2.30 0.07
1.10 0.03
1.01 0.02
0.98 0.02
0.94 0.02
1.17 0.06
1.08 0.01
1.00 0.02
0.97 0.02
90E10W
80E20W
70E30W
90A10W
80A20W
70A30W
60A40W
a
b
Binary solvents are v/v at 25 °C. A = acetone, E = ethanol, W = water. Errors shown are standard errors.
8988
dx.doi.org/10.1021/jo3013308 | J. Org. Chem. 2012, 77, 8986−8998

The Journal of Organic Chemistry
Article
Scheme 2. Epoxy Ring Formation Model Reaction
substrates show deviation from the log k vs pKa plot, because in
such correlations, the intrinsic factor is neglected. Clearly, the
solvolytic reactivity of a given substrate cannot be reliably
estimated from the basicity of the LG.9
calculated barriers at 25 °C for the epoxy ring formation and
the experimental barriers for solvolysis of corresponding
dianisylmethyl benzoates (see Figures S11 and S12 in
Supporting Information). In addition, due to the anionic
form of the substrates, small barriers (ca. 2 to 7 kcal/mol) have
been obtained. However, inclusion of the polarizable
continuum (water) in optimization increased the barriers and
improved the correlation significantly. By applying the PCM
model, which mimics the solvent, a different demand for
solvation of the ground and the transition state has been taken
into account. While the electrostatic solvation of the negative
charge located mainly on the oxygen in the ground-state
structures is important, the transition structures, in which the
negative charge is delocalized, might be less solvated with
increasing polarity of medium. The net effect is an increase of
the reaction barriers. It should be emphasized that the C−O
bond dissociates in all the cases; hence, the impact of hydrogen
bonds (electrophilic solvation) on relative reactivity is canceled
out and thus inclusion of explicit molecules of solvent in the
model reaction is not necessary.
To investigate the nucleofugality of numerous substituted
benzoate leaving groups, we have searched for a reliable
theoretical model, whose reaction barriers obtained by
quantum-chemical calculations correlate reasonably well with
the experimental barriers of the benzhydryl derivatives. The
total (electronic) energy in the heterolytic cleavage of the C
LG bond of neutral substrates continuously increases until
interaction between the charged LG and the nearest
neighboring hydrogen occurs, preventing location and opti-
mization of the transition state for the heterolytic step of the
monomolecular substitution. On the other hand, optimized
transition states for SN2 reactions10 as well as for displacement
reactions that proceed via participation of the neighboring
group (anchimeric assistance) are presented in the chemical
literature.11,12 Also, a quantum-chemical study has been used to
correlate the experimentally determined free energies of
activation for combination of benzhydrylium ions with various
π-electrophiles and free energies of activation calculated by
different theoretical methods.13 Although considerable system-
Geometry optimizations, using the Gaussian 09 package,14
were carried out without any constraints at the B3LYP/6-
311+G(2d,p) level. The integral equation formalism model for
PCM calculations (IEFPCM) in a dielectric continuum
representing water as a solvent has been employed.15 Details
of the calculations are given in Computational Methods below
and in Supporting Information (coordinates and energies). The
optimized substrate and transition-state structures for the
heterolytic epoxy ring formation with 2-oxyethyl 3,5-
dinitrobenzoate and 2-oxyethyl 2,4,6-trifluorobenzoate are
presented in Figure 2 (other structures are included in
Supporting Information).
Lengths of the partial (cleaving) bonds are basically the same
in all TS. This is not surprising because the substituents on the
benzene ring do not take part in the resonance delocalization of
the developing negative charge on the carboxyl group but only
electrostatically interact with it. The structures shown indicate
that the geometries of the ground and transition states are
influenced by the substituents on the phenyl ring. Generally,
substituents in the ortho-position cause a substantial twisting of
the benzene ring out of the carboxyl group plane, as is shown
for 2,4,6-trifluorobenzoate in Figure 2. This twisting effect is
more pronounced in the TS, so the dihedral angle between the
carboxyl plane and the benzene ring is larger in the TS than
that in the ground-state structure in all the cases. Further
increase of the dihedral angle in the transition-state structure
probably comes from the increase of the spatial electrostatic
repulsions between the ortho-substituents and the partially
negatively charged carboxyl group. In dynamic systems this
might induce the decrease of the rotational degrees of freedom
at the transition state, which leads to the decrease of the
activation entropy.
atic deviations of ΔG⧧ from the experimental values are
calc
obtained, the correlations are reasonably good. Furthermore,
Aggarwal et al. using the B3LYP method examined the PES for
reactions of ammonium, oxonium, phosphonium, and
sulfonium ylides with benzaldehyde and suggested that the
ability of the leaving group (nucleofugality) had a distinct
influence on the rate of the styrene oxide formation in the
intramolecular nucleophilic substitution step.11 They also
demonstrated a notable impact of the intrinsic barrier on the
ability of neutral leaving groups (OMe2, SMe2, NMe3, and
PMe3) in the heterolytic step.
We have chosen to compute the epoxy ring formation as a
model reaction in which the intramolecular backside n-electron
attack of the negatively charged oxygen is a driving force for the
C−OCOBz-Z bond cleavage (Scheme 2) and to correlate the
activation energies of the model reaction with the experimental
barriers of the heterolysis of dianisylmethyl benzoates. Our
basic assumption is that the differences in reaction energetics
for the series of model reactions are determined mostly from
the effects of LGs, while other variables contribute to overall
energies similarly (e.g., unfavorable entropy of activation caused
by n-electron assistance, solvation effects, etc.)
Quantum-chemical calculations for the model reaction have
been performed with various 2-oxyethyl benzoates, and the
corresponding transition structures have been optimized. In
preliminary calculations the transition-state structures for the
oxy-assisted heterolytic cleavage have been optimized in the gas
phase with the following leaving groups: PFB, TFB, DNB,
PNB, and Bz. Poor correlation has been obtained between the
8989
dx.doi.org/10.1021/jo3013308 | J. Org. Chem. 2012, 77, 8986−8998

The Journal of Organic Chemistry
Article
Figure 2. Optimized PCM-B3LYP/6-311+G(2d,p) structures of 2-
oxyethyl 3,5-dinitrobenzoate and 2-oxyethyl 2,4,6-trifluorobenzoate
and of the related heterolytic transition states. Selected distances are
given in angstroms. Dihedral angles (d.a.) between the carboxyl plane
and benzene ring are given in degrees.
Figure 3. Correlation of experimental activation free energies (kcal
mol−1) for solvolyses of dianisylmethyl benzoates in 90% ethanol at 25
°C against PCM-B3LYP/6-311+G(2d,p)-calculated activation free
energies (A), and calculated activation enthalpies (B) for heterolyses
of 2-oxyethyl benzoates.
Correlation of Experimental and Computed Leaving
Group Effects. To compare the leaving group effects on
reactivities of the model heterolytic epoxy ring formation
(Scheme 2) and of benzhydryl benzoates, we plotted the
experimental barriers for solvolysis of 12 reference dianisyl-
methyl benzoates (ΔG⧧) at 25 °C in a series of aqueous
ethanol (90%, 80% and 70%) mixtures against the calculated
free energies of activation (ΔG⧧) obtained by quantum-
chemical calculations for epoxidation using a polarizable
continuum that mimics water as a solvent. The correlation
plot in 90% aqueous ethanol is presented in Figure 3A (plots
obtained for the other two solvents are shown in Figures S9
and S10 in Supporting Information). The correlation is
reasonably good (r = 0.963), justifying the model used.
Moreover, slopes of the plots are close to unity, indicating that
the leaving groups indeed have effects on the barriers of the
heterolysis reactions similar to those of the modeled heterolytic
epoxy ring formations.
benzoates with different Z-substituents on the benzoate moiety
produce in solvolysis. Kevill demonstrated that in the
ethanolysis of adamantyl arenesulfonates, the rate-determining
variable is the entropy of activation (ΔS⧧), while the activation
enthalpies are essentially the same for all the substrates
investigated.16 However, this is not the case here for solvolysis
of benzhydryl benzoates. Most ΔS⧧ obtained for solvolysis of
dianisylmethyl benzoates (Table S1) are similar in magnitude
(between ca. −13 and −10 cal K−1 mol−1, so ΔΔS⧧ ≈ 1 cal
K−1 mol−1), indicating that for most cases, ΔH⧧ is the rate-
determining variable. Indeed, the corresponding experimental
ΔH⧧ values for solvolysis of dianisylmethyl benzoates in 90%
aq ethanol correlate reasonably well with the calculated ΔH⧧
values for the model reaction (the linear plot is shown in Figure
S19). However, for a few substrates, such as dianisylmethyl
DNB and PNB, the experimental ΔS⧧ parameters are less
negative than for others. For those very cases, experimental
ΔH⧧ values also deviate in the direction to cancel out the
deviations of ΔS⧧. The net effect is excellent correlation
between the calculated ΔH⧧ for the model reaction and the
experimental ΔG⧧. The root-mean-square error for correlated
experimental ΔG⧧ using the calculated ΔH⧧ in 90% aq ethanol
is 0.12 kcal/mol (vs 0.35 kcal/mol for the correlation of
experimental and calculated ΔG⧧ given in Figure 3A); statistical
data in other solvents are similar (footnotes in Tables 3 and
S7).
Our further aim was to estimate the barriers and the rate
constants for solvolysis of a vast number of different benzoates
in aqueous ethanol mixtures based on the above correlation and
to extract the corresponding nucleofugality parameters. Good
linear correlation is crucial to get accurate rate constants. We
found that the correlation was better between the experimental
ΔG⧧ and the calculated enthalpies of activation (ΔH⧧) at 25 °C
in all solvents used (r = 0.997 vs 0.963 for 90% ethanol)
whereas, importantly, the slopes of those lines were also about
unity (Figure 3). This might be because enthalpies obtained by
PCM calculation implicitly include contributions of solvation
entropies. On the other hand, without additional corrections
the entropy term included into the calculated Gibbs free energy
of activation tends to be quantitatively inaccurate, mostly due
to the impact of low frequencies.
The linear plot between the experimental ΔG⧧ obtained in
90% aq ethanol and the calculated ΔH⧧ for the model reaction
is presented in Figure 3B (correlations for 80% and 70% aq
The next reason for good ΔG⧧ vs ΔH⧧ correlation
exp
calc
comes from essentially invariant ΔS⧧ values that benzhydryl
8990
dx.doi.org/10.1021/jo3013308 | J. Org. Chem. 2012, 77, 8986−8998

The Journal of Organic Chemistry
Article
Table 3. Calculated Activation Free Energies, Solvolysis Rate Constants for Reference Dianisylmethyl Benzoates, and Related
Calculated Nucleofugalities at 25 °C
a
b
c
d
solvent
Z−benzoate
H (Bz)
ΔG⧧
kcalc/s−1
kcalc − kexp/s−1
krel
Nfcalc
−4.43
−2.14
−3.06
−3.00
calc
e
90E10W
23.20
20.22
21.41
21.34
19.80
21.69
21.14
21.39
19.99
18.59
22.93
22.71
22.75
19.90
21.04
20.97
19.50
21.31
20.78
21.02
19.68
18.34
22.50
22.28
22.48
19.69
20.81
20.73
21.07
20.55
20.79
19.47
18.16
22.23
22.02
6.14 × 10−5
9.32 × 10−3
1.25 × 10−3
1.42 × 10−3
1.90 × 10−2
7.83 × 10−4
1.97 × 10−3
1.30 × 10−3
1.39 × 10−2
1.46 × 10−1
9.61 × 10−5
1.41 × 10−4
1.30 × 10−4
1.60 × 10−2
2.34 × 10−3
2.65 × 10−3
3.18 × 10−2
1.49 × 10−3
3.61 × 10−3
2.43 × 10−3
2.35 × 10−2
2.24 × 10−1
2.00 × 10−4
2.89 × 10−4
2.06 × 10−4
2.29 × 10−2
3.48 × 10−3
3.93 × 10−3
2.24 × 10−3
5.33 × 10−3
3.61 × 10−3
3.33 × 10−2
3.02 × 10−1
3.13 × 10−4
4.50 × 10−4
−0.86 × 10−5 (−12%)
1.0
2-NO2
0.22 × 10−3 (+2%)
1.5 × 102
2.0 × 101
2.3 × 101
3.1 × 102
1.3 × 101
3.2 × 101
2.1 × 101
2.3 × 102
2.4 × 103
1.6
3-NO2
−0.28 × 10−3 (−18%)
−0.02 × 10−3 (−1%)
−0.49 × 10−2 (−21%)
−2.17 × 10−4 (−22%)
0.18 × 10−3 (+9%)
4-NO2 (PNB)
3,5-di-NO2 (DNB)
4-CN
−1.76 (−0.22)
−3.27
2-CF3
−2.85
−2.95
3,5-di-Cl
−0.58 × 10−3 (−31%)
0.14 × 10−2 (+10%)
−0.37 × 10−1 (−20%)
0.20 × 10−5 (+2%)
−0.54 × 10−4 (−28%)
0.04 × 10−4 (+3%)
0.11 × 10−2 (+7%)
2,4,6-tri-F (TFB)
penta-F (PFB)
2-MeO
−1.90 ( 0.00)
−0.93 (−0.17)
−4.23
2,4,6-tri-Me
H (Bz)
2.3
−3.93
−4.09
−1.89
−2.77
f
80E20W
1.0
2-NO2
1.2 × 102
1.8 × 101
2.0 × 101
2.4 × 102
1.1 × 101
2.8 × 101
1.9 × 101
1.8 × 102
1.7 × 103
1.5
3-NO2
−0.44 × 10−3 (−16%)
−0.00 × 10−3 ( 0%)
−0.75 × 10−2 (−19%)
−0.46 × 10−3 (−24%)
0.44 × 10−3 (+12%)
−0.68 × 10−3 (−22%)
0.29 × 10−2 (+12%)
−0.20 × 10−1 (−8%)
0.20 × 10−4 (+10%)
−0.94 × 10−4 (−25%)
0.25 × 10−4 (+12%)
0.11 × 10−2 (+5%)
−0.41 × 10−3 (−11%)
−0.45 × 10−3 (−10%)
−0.80 × 10−3 (−26%)
0.83 × 10−3 (+16%)
−0.90 × 10−3 (−20%)
0.47 × 10−2 (+14%)
−0.11 × 10−1 (−4%)
0.41 × 10−4 (+13%)
−0.89 × 10−4 (−17%)
4-NO2 (PNB)
3,5-di-NO2 (DNB)
4-CN
−2.71 (+0.07)
−1.53 (−0.10)
−2.97
2-CF3
−2.57
−2.67
3,5-di-Cl
2,4,6-tri-F (TFB)
penta-F (PFB)
2-MeO
−1.66 (+0.09)
−0.72 (−0.04)
−3.89
2,4,6-tri-Me
H (Bz)
2.2
−3.61
−3.88
−1.73
−2.59
−2.53
−2.79
−2.39
−2.49
g
70E30W
1.0
2-NO2
1.1 × 102
1.7 × 101
1.9 × 101
1.1 × 101
2.6 × 101
1.8 × 101
1.6 × 102
1.5 × 103
1.5
3-NO2
4-NO2 (PNB)
4-CN
2-CF3
3,5-di-Cl
2,4,6-tri-F (TFB)
penta-F (PFB)
2-MeO
−1.51 (+0.17)
−0.58 ( 0.00)
−3.69
2,4,6-tri-Me
2.2
−3.42
b
a
In kcal mol−1; calculated from ΔG⧧ (dianisylmethyl-LG)/ΔH⧧ (2-oxyethyl-LG) correlations given in Figures 3B, S6, and S7. Calculated from
c
d
ΔG⧧calc. Differences from experimental values are given in parentheses. krel = kcalc (X-Bz)/kcalc (Bz). Calculated from kcalc and related sf using eq 1
and the Ef value for dianisylmethyl electrofuge (Ef = 0.00). Applied sf values are given in Table S6. Deviations from experimental values are given in
e
parentheses. From the correlation in Figure 3B: slope = 0.97 0.02, intercept = 6.60 0.33 kcal mol−1 (r = 0.997; root-mean-square error = 0.12
f
kcal mol−1). From the correlation in Figure S6: slope = 0.93 0.02, intercept = 6.84 0.32 kcal mol−1 (r = 0.997; rms error = 0.10 kcal mol−1).
g
From the correlation in Figure S7: slope = 0.91 0.02, intercept = 6.91 0.36 kcal mol−1 (r = 0.997; rms error = 0.10 kcal mol−1).
ethanol are given in Figures S6 and S7, respectively, in
Supporting Information). From the correlation lines, we
calculated the rate constants for the 12 reference dianisylmethyl
benzoates in aqueous ethanol mixtures and the corresponding
nucleofugality parameter for a given benzoate using eq 1, as
well as the deviations from the experimental data (Table 3).
Data presented in Table 3 illustrate high accuracy of the model,
indicating that most deviations between the calculated and
experimental rate constants for solvolysis of dianisylmethyl
benzoates, as well as between corresponding calculated Nf
values and the experimental Nf values, are in the limits of the
experimental error. It should be emphasized that similar values
for rate constants have been obtained if the rate constants (and
the Nf values) have been calculated from the lines obtained by
correlation of the calculated ΔG⧧ for the model reaction and
the experimental ΔG⧧ (presented in Figure 3A, and for 80%
and 70% ethanol presented in Figures S9 and S10, respectively,
in Supporting Information), but the errors are larger (Table S7
in Supporting Information).
For comparison, some reference ground-state and corre-
sponding TS structures have been optimized in the gas phase,
and the PCM single-point calculations have been performed at
the same level of theory. The results (Table S5 in the
Supporting Information) show that barriers obtained by PCM
single-point calculations are between 1 and 2 kcal/mol lower
than those obtained by PCM optimization in all the cases.
However, both corelations (ΔG⧧ vs ΔG⧧ and ΔG⧧ vs
exp
calc
exp
ΔH⧧calc) are worse if PCM single-point energies have been
applied (r = 0.973, rms error = 0.34 kcal/mol and 0.990, rms
error = 0.19 kcal/mol, respectively, in 90% aq ethanol: Figure
S13), indicating that PCM optimization is advantageous for this
system. Our further attempt was to optimize the above-
described model epoxy ring fromation reaction using one of the
hybrid meta-GGA Minnesota functionals.17 Therefore, the
same 12 reference ground-state and related transition structures
8991
dx.doi.org/10.1021/jo3013308 | J. Org. Chem. 2012, 77, 8986−8998

The Journal of Organic Chemistry
Article
have been optimized using the PCM in combination with the
M06-2X functional. The barriers obtained by this functional are
about 6 kcal/mol higher than those obtained by B3LYP. The
M06-2X calculated barriers for the model reaction have been
correlated with the experimental barriers of the solvolysis
presented in Scheme 1 in 90% aqueous ethanol, as described
above (Figure S14 in the Supporting Information). The
correlations between the experimental barriers and those
obtained with M06-2X functional (r = 0.971, rms error =
0.32 kcal/mol for both ΔG⧧exp vs ΔG⧧calc and ΔG⧧exp vs ΔH⧧
calc
correlations) are considerably poorer than those obtained with
B3LYP functional.
Method for Estimating the Reactivity (Nucleofugality)
of Other Substituted Benzoates. The linear correlations
presented above between the experimental ΔG⧧ and the PCM-
B3LYP calculated ΔH⧧ for the model reaction provides a
method for estimating the solvolytic rate constants of any
dianisylmethyl benzoate in a given aq ethanol, by optimizing
ground-state and TS structures, and calculating the enthalpy of
activation of the model epoxy ring closure for a given benzoate.
The nucleofugalities (Nf parameters) can be calculated
reasonably correctly from eq 1 taking the assumption that
benzoates of similar structure have similar sf. Comparison of the
calculated Nf parameters (Table 3) with experimentally
obtained ones (Table 2 and reference 4) reveals that deviations
are mostly within the limits of experimental error. This
approach ultimately enables semiquantitative prediction of the
SN1 solvolysis rate of any combination of benzhydryl
electrofuge−benzoate simply by using eq 1.4
To define the reactivity of a vast number of benzoates, we
further optimized the structures of 67 different 2-oxyethyl-LGs
in which the substituents on the benzoate moiety have
systematically been varied. Also, the transition states for all
corresponding epoxy ring formation reactions via concerted
CLG bond cleavage have been located (Scheme 2). The
barriers and the corresponding rates for the solvolytic
displacement reactions in aqueous ethanol mixtures of 67
dianisylmethyl benzoates have been calculated from the ΔG⧧/
ΔH⧧ correlations presented in Figure 3B (in 90% aq ethanol)
and in Supporting Information (Figures S6 and S7 for 80% and
70% aq ethanol, respectively). The predicted absolute and
relative reaction rates of various dianisylmethyl benzoates in
90%, 80%, and 70% aqueous ethanol, as well as the derived
nucleofugality parameters, are given in Table S6. In summary,
the nucleofugalities (experimental and calculated) are now
available for the spectra of about 80 benzoates that cover
reactivities of 7 orders of magnitude. The most reactive LG
examined here is pentacyanobenzoate, while the least reactive is
3,4,5-triaminobenzoate (Table S6).
Figure 4. Hammett plot of calculated rate constants for solvolyses of
dianisylmethyl benzoates in 90% ethanol at 25 °C. The deviating
points represented with open circles are not included in the
correlation.
additive. Therefore, the relative reactivity among the variously
substituted benzoates is preserved with this model.
The Hammett ρ parameters obtained for the heterolysis
reactions from the calculated rate constants are as follows: 1.75
(90% aq ethanol), 1.68 (80% aq ethanol), and 1.65 (70% aq
ethanol). Those data are in accordance with the ρ values
obtained earlier for heterolysis reactions that solvolyze via
limited SN1 pathway. For example, in ethanolysis of 2-
adamantyl arenesulfonates and 1-adamantyl arenesulfonates
the ρ parameters are 1.86 and 1.76, respectively.16 Although for
the purpose of this work only heterolytic rate constants of
dianisylmethyl benzoates were used for correlations, we expect
that with different benzhydryl electrofuges the quality of
correlations would also be the same, which opens an even wider
perspective for application of the presented model.
Intrinsic Barriers in Heterolysis of Benzoates. It is
generally accepted that two major rate-determining variables
are the change in reaction thermodynamics (ΔG°) and the
intrinsic barrier (Λ).19 To obtain information about how the
variation of the substituents in benzoates affects the intrinsic
barriers, we computed the thermoneutral methyl exchange
reactions presented in Scheme 3 for 39 different benzoates,
Scheme 3. Methyl Exchange Identity Reaction
whose leaving group abilities have already been determined
computationally using the above epoxy ring-formation model
reaction. The calculations have also been performed at the
IEFPCM-B3LYP/6-311+G(2d,p) level, taking the polarizable
continuum representing water, as described in Computational
Methods.
Because the thermodynamic driving force in such an identity
reaction is zero, the barrier height represents the intrinsic
barrier for methyl exchange, which corresponds to both
intrinsic nucleofugality and intrinsic nucleophilicity of the
benzoate anion. Calculated intrinsic barriers are given in Table
S8.
The suitability of the above quantum-chemical model for
prediction of the reactivity of various dianisylmethyl benzoates
has additionally been verified by the Hammett correlation in
which the logarithms of the calculated solvolysis rate constants
for 6-LG (X = Y = MeO) in 90% aq ethanol are plotted against
the sum of the Hammett σ values (Figure 4).18 Again, a very
good correlation has been obtained in all solvents (r = 0.994,
root-mean-square error = 0.11 in 90% aq ethanol; correlations
in 80% and 70% aq ethanol are presented in Figures S16 and
S17, respectively). The fact that not only monosubstituted but
also di- and trisubstituted benzoates correlate well, for which
the sum of the σ values has been taken into account, indicates
that the calculated substituent effects on the aromatic ring are
It has repeatedly been shown for different types of reactions
that the intrinsic barriers of a structurally homologous series of
substrates change with variation of the structure.20−22 In line
8992
dx.doi.org/10.1021/jo3013308 | J. Org. Chem. 2012, 77, 8986−8998

The Journal of Organic Chemistry
Article
with such observations are our results, presented in Figure 5A.
The relation between the experimental barriers for solvolysis of
study on heterolytic reactions of tritylium compounds, Mayr
demonstrated unexpectedly low reactivities of dimethylamino-
substituted trityl esters in comparison to methyl- and methoxy-
substituted trityl esters, indicating that the amino-group
stabilization of the carbocation by resonance develops notice-
ably later.24
The above premises can be used to rationalize the variable
intrinsic barriers obtained with the substituted benzoates.
Stabilization of the generated benzoate anion comes from the
resonance on the carboxylate moiety and from the overall polar
stabilization due to the electron-withdrawing substituents on
the phenyl ring (field effect and inductive effect). Each electron-
accepting substituent increases the polar stabilization effects of
the ring but decreases the extent of the resonance effect on the
carboxylate moiety in overall stabilization of TS; hence, the
intrinsic barrier is successively decreased. According to the
above consideration, both the thermodynamics and the intrinsic
barrier influence the overall barrier for heterolysis in the same
direction, the electron-withdrawing substituents stabilize the
LG and also enhance the reactions by reducing the intrinsic
barriers, and the electron-donating substituents on the phenyl
may increase the extent of resonance on the carboxylate group
and increase the intrinsic barrier. Therefore, it is not surprising
that the calculated enthalpies of activation (ΔH⧧) for the model
heterolytic epoxy ring formation presented in Scheme 2, and
the calculated enthalpies of activation for the thermoneutral
identity reaction presented in Scheme 3 correlate very well
(Figure 5B).
For estimating the extent of the polar substituent effects in
the TS, the normalized Hammett constant,20a which is here
based on changes in enthalpies, has been calculated for the
anchimerically assisted heterolytic dissociation of the oxyethyl-
LG models (Scheme 2) according to eq 2:
Figure 5. Correlation of calculated activation enthalpies for
thermoneutral LG-Me-LG exchange versus experimental activation
free energies for the solvolysis of 6-benzoates in 90% ethanol (A), and
calculated activation enthalpies for anchimerically assisted heterolytic
dissociation of 2-oxyethyl benzoates (B).
ρH
−2.5
activation
ρH
=
=
= 0.68
dianisylmethyl derivatives and the calculated intrinsic barriers
for methyl exchange (ortho-substituted benzoates are not
included in the correlation because of steric and other
intramolecular interactions of substituents) indicates that the
intrinsic barriers increase as the overall barriers of the
heterolysis increase, i.e., the intrinsic barrier is higher if a less
stable benzoate leaving group is generated. Consequently, for
the combination reaction (benzhydrylium ion + LG), the
intrinsic barriers increase as the stability of the benzoate
nucleophiles decrease.
Richard investigated numerous cases of the combination
reactions of benzylic carbocations with solvents, in which the
stability of carbocations varied by substituents at α-position,
and demonstrated that the intrinsic barriers for the
combination increased with destabilization of the carbocations
caused by α-electron-withdrawing substituents, due to increas-
ing participation of the vicinal aromatic ring electrons by
resonance.20,23 Thus, the overall barriers remain constant
regardless of the changes in thermodynamic driving forces of
combination reactions. The general conclusion was that the
changeable Λ is due to the nonsynchronized onset of the
resonance and the polar substituent effects in the TS, which has
been summarized as the Principle of nonperfect synchronization.21
In the case of combination reactions, the onset of the polar
effects lags behind the resonance effects and, consequently, a
more intense resonance effect in the TS increases the intrinsic
barrier.20a For heterolytic reactions, an inverse order of onset of
the stabilizing effects is expected. In the comprehensive kinetic
norm
ρH
−3.7
(2)
reaction
in which ρH
represents the Hammett reaction constant
activation
obtained by plotting the calculated enthalpies of activation
against the sum of corresponding Hammett σ values, whereas
represents the reaction constant obtained by plotting
the reaction enthalpies against the sum of σ parameters (the
ρH
reaction
correlation plots are shown in Figure S18).
The substituent effects on the phenyl ring of the benzoate
through polar influences are the specific rate-determining
variables. According to the quantum-chemical calculations, the
above fractional expression indicates that for the model reaction
(Scheme 2) about 70% of substituent polar interactions are
already effective in the transition states. On the other hand,
calculations showed that regardless of the LG, the epoxy ring-
formation model reaction (Scheme 2) is slightly exothermic
(Table S4), so it proceeds via a relatively early transition state.
One can speculate that if in an early TS approximately 70% of
polar interactions are already operative, then the further intense
stabilization after the TS must come from the resonance of the
carboxylate moiety, i.e., in the heterolytic cleavage of benzoate
the resonance effect lags behind the polar effects, as already
concluded from the variable intrinsic barriers.
Finally, it can be concluded that the variation of the
substituents on the benzoate moiety in the model reaction and
in the heterolysis reaction affects ΔG⧧ by the same amount,
whereas Λ values are changing proportionally to the reaction
8993
dx.doi.org/10.1021/jo3013308 | J. Org. Chem. 2012, 77, 8986−8998

The Journal of Organic Chemistry
Article
enthalpy. Combined influence of the two effects is responsible
for such good correlations presented in Figure 3.
Rate Effects of the Substituents. Abundance of
experimental and calculated data make it possible to examine
some aspects of the influence of the phenyl ring substituents on
the solvolytic reactivity of benzoates and also to find out how
the solvent polarity can alter those influences.
Kevill observed that 1-adamantyl and 2-adamantyl m-
nitrosulfonates solvolyze faster than the corresponding p-
nitrosulfonates in ethanol, which is not in accord with the
Hammett σ-parameters obtained in water (σp‑NO2 = 0.78;
σm‑NO2 = 0.71).16,18 Such an order of reactivity was attributed to
solvent effects.
Our experimental results indicate clearly that the relative
reactivities of m- and p-nitrobenzoates depend on the solvent
polarity. Furthermore, we were able to detect the inversion of
relative reactivities with changing solvent polarity. Thus, 6-PNB
(dianisylmethyl p-nitrobenzoate) solvolyzes faster in 70% aq
ethanol (see entries for PNB and m-NO2-Bz in Table S1), and
the rates of m- and p-nitrobenzoates are almost the same in
80% ethanol, whereas in 90% the solvolysis of m-nitrobenzoate
is faster. Similar behavior has been observed in solvolysis of
anisylphenylmethyl p- and m-nitrobenzoates (3-PNB and 3-m-
NO2-Bz; the last two entries in Table S1). Thus, m-
nitrobenzoate is more reactive in 90% aq ethanol, whereas in
70% the rates are the same in the limits of experimental error.
In summary, in more polar solvents p-nitrobenzoates solvolyze
faster, which is consistent with Hammett σ-parameters
determined in water, whereas in a less polar solvent m-
nitrobenzoates are more reactive.
Wiberg investigated the influences of substituents on
calculated gas-phase acidities of substituted benzoic acids and
showed that different field effects of differently substituted
phenyl rings on the carboxylate moiety are the major variables
that determine the acidities, including those of p- and m-nitro
groups.25 Using Wiberg’s formalism, inversion of the relative
reactivities of p- and m-nitrobenzoates can be rationalized as
follows: The nitro group (strong electron-withdrawing
substituent) on the phenyl ring separates the charges in the
phenyl ring by resonance. The partial positive charge, induced
by the electron-withdrawing nitro group in para-position,
comes conveniently to the vicinity of the negatively charged
carboxylate group, namely to the ipso-carbon, as shown by the
resonance structure A in Scheme 4. In m-nitrobenzoates the
causing m-nitrobenzoates to solvolyze faster. Therefore, in
more polar solvents the distinct field effect between the para-
substituted phenyl ring and the carboxylate moiety due to
charge separation by resonance is the rate-determining variable,
whereas in a less polar solvent it is the inductivity of the closer
substituent. Because the Hammett σ-parameters have originally
been determined in highly polar water, it is easy to understand
why those reflect the relative reactivities of nitrobenzoates in
more polar solvents. It should be mentioned that substituents
on the ortho-position enhance the reaction rate significantly
because of both inductive and field effects.
Using the ionizing power scale based on adamantyl tosylate,
the slope parameters (mOTs) of the log k vs YOTs line for
dianisylmethyl m- and p-nitrobenzoates in aqueous ethanol
showed that p-nitrobenzoates are somewhat more sensitive to
solvent polarity than m-nitrobenzoate (mOTs = 0.38 for 6-PNB;
mOTs = 0.33 for 6-3-NO2-Bz; derived from rate constants
presented in Table S1), indicating that in the transition
structures the charge separation is more favored in p-
nitrobenzoate.
To observe the substituent effects, it is useful to compare the
calculated relative rate constants. In both Tables 3 and S6 the
rate constant for unsubstituted dianisylmethyl benzoate is taken
as a reference (krel = 1). It seems that the reaction rates
obtained by calculations reflect the solvolytic behavior of the
substrates in more polar solvent. For substituents whose
influence on the benzoate ion stability comes from inductive
effect only, as for example halogens or phenyl, the order in
reactivity is ortho-substituted-Bz > meta-substituted-Bz > para-
substituted-Bz. For example, according to results in Table S6,
the relative reactivities of dianisyl o-, m-, and p-Cl-benzoates in
90% aq ethanol are: 8.2:1.8:1. However, if the charge separation
in the phenyl ring is enhanced in the TS by the strong electron-
withdrawing substituents, as shown above for nitrobenzoates,
the calculated rates show that para-substituted benzoates
solvolyze faster than meta-substituted benzoates, i.e., the
more favorable inductive effect of the closer meta-substituent
is canceled out with the favorable field effect of the phenyl ring
with para-substituent. Similar behavior is also predicted for
benzoates with strong electron-withdrawing CHO and CN
groups (Table S6). Such an order of reactivities is consistent
with experimental data obtained in 70% aq ethanol.
The methoxy group is the example of substituent which has a
stabilizing effect by inductivity but an unfavorable destabilizing
charge distribution effect by resonance (structures C and D in
Scheme 4). It seems that those two effects compensate for each
other in m-methoxybenzoate, so its calculated solvolysis rates in
aqueous ethanol solutions are essentially the same as the rates
of unsubstituted benzoate.
Scheme 4
In p-MeO-benzoates the inductive effect on a more remoted
position is less important, whereas the unfavorable repulsive
field effect is enhanced by the ability of the methoxy group to
distribute additional negative charge to the ipso-position, which
is in the vicinity of the negatively charged carboxylate moiety
(resonance structure C in Scheme 4). Therefore, p-methox-
ybenzoates solvolyze slower than the corresponding m-methoxy
benzoates. However, if the methoxy group is located in ortho-
position, the favorable inductive effect is strong enough to
overcome the destabilizing field effect (resonance structure D)
causing the enhancement of the rate in comparison to
unsubstituted benzoate.
positive charge is directed to ortho-positions (structure B),
which is of greater distance from the negatively charged
carboxylate group (Scheme 4). The larger field effect slightly
stabilizes the TS of p-nitrobenzoate over m-nitrobenzoate;
hence, the former solvolyzes faster. However, in less polar
solvents the internal charge separation in the benzoate moiety
is less favored, the favorable field effect is diminished, and the
greater inductivity of the meta-substituent compared with that
of the para-substituent becomes the rate-determining variable
8994
dx.doi.org/10.1021/jo3013308 | J. Org. Chem. 2012, 77, 8986−8998

The Journal of Organic Chemistry
Article
Ar2CH), 7.30−7.42 (m, 7H; ArH); 13C NMR (75 MHz, CDCl3, 20
°C): δ = 55.4 (ArOCH3), 79.4 (Ar2CH), 114.2, 121.7, 127.0, 128.4,
128.8, 129.0, 139.1, 159.8 (Ar); 19F NMR (282 MHz, CDCl3, 20 °C):
δ = −137.9, −148.5, −160.5 (F5Ar) ppm; elemental analysis calcd (%)
for C21O3F5H13 (408.30): C 61.77, H 3.21; found C 61.55, H 3.26.
4-Methoxy-4′-methylbenzhydryl pentafluorobenzoate: (4-PFB)
from 4-methoxy-4′-methylbenzhydrol (1.00 g; 4.4 mmol), pyridine
(0.76 g; 9.6 mmol), and pentafluorobenzoyl chloride (1.11 g; 4.8
mmol); yield 1.30 g, 71.9%; mp 114.6−116.9 °C; 1H NMR (300 MHz,
CDCl3, 20 °C): δ = 2.34 (s, 3H; ArCH3), 3.79 (s, 3H; ArOCH3), 6.88
(d, J = 8.0 Hz, 2H; ArH), 7.06 (s, 1H; Ar2CH), 7.16−7.36 (m, 6H;
ArH); 13C NMR (75 MHz, CDCl3, 20 °C): δ = 21.3 (ArCH3), 55.4
(ArOCH3), 79.7 (Ar2CH), 114.1, 127.1, 128.9, 129.5, 131.4, 136.4,
138.3, 159.6 (Ar); 19F NMR (282 MHz, CDCl3, 20 °C): δ = −137.9,
−148.6, −160.5 (F5Ar) ppm. MALDI-TOF/TOF MS calcd for
C22H15F5O3 [M + Na+]: 445.0833; found: 445.0855.
CONCLUSION
■
The solvolysis of X,Y-substituted benzhydryl benzoates
proceeds via an SN1 path in which the substituents on the
phenyl ring of the benzoate moiety represent a major rate-
determining factor by polar effects, i.e., by field and inductive
effects. The intrinsic barriers of such reactions are variable and
depend on the substituents on the phenyl ring in the benzoate
moiety in a way that the more stabilized LG proceeds over a
lower intrinsic barrier. The experimental barriers for the
solvolysis of benzhydryl benzoates and those for the heterolytic
epoxy ring formation, calculated by PCM quantum-chemical
calculations, correlate linearly, providing the model for
extrapolation of the reactivity of any benzhydryl benzoate and
also for calculation of the nucleofugality of a given benzoate in a
given solvent. Using the method presented above, the
theoretical nucleofugality parameters for about 70 benzoates
have been determined in 90%, 80%, and 70% aqueous ethanol.
Furthermore, using eq 1 and taking those nucleofugalities
determined here, the heterolysis reaction for any combination
of the electrofuge−benzoate nucleofuge can be determined
semiquantitatively in a series of aqueous ethanol solutions.
4-Methoxy-4′-phenoxybenzhydryl pentafluorobenzoate: (5-
PFB) from 4-methoxy-4′-phenoxybenzhydrol (0.80 g; 2.6 mmol),
pyridine (0.45 g; 5.7 mmol), and pentafluorobenzoyl chloride (0.66 g;
1
2.9 mmol); yield 1.14 g, 87.0%; mp 73.6−75.8 °C; H NMR (300
MHz, CDCl3, 20 °C): δ = 3.80 (s, 3H; ArOCH3), 6.88−7.37 (m, 14H;
Ar2CH + ArH); 13C NMR (75 MHz, CDCl3, 20 °C): δ = 55.5
(ArOCH3), 79.3 (Ar2CH), 114.2, 118.7, 119.4, 123.8, 128.8, 130.0,
131.4, 134.0, 156.8, 157.7, 159.8 (Ar); 19F NMR (282 MHz, CDCl3,
20 °C): δ = −137.8, −148.4, −160.4 (F5Ar) ppm; MALDI-TOF/TOF
MS calcd for C27H17F5O4 [M + Na+]: 523.0939; found: 523.0945.
4,4′-Dimethoxybenzhydryl pentafluorobenzoate: (6-PFB) from
4,4′-dimethoxybenzhydrol (1.00 g; 4.1 mmol), pyridine (0.71 g; 9.0
mmol), and pentafluorobenzoyl chloride (1.04 g; 4.5 mmol); yield
EXPERIMENTAL SECTION
■
Substrate Preparation. 4-Methylbenzhydrol, 4-methoxybenzhydrol,
and 4,4′-dimethoxybenzhydrol were prepared by the reduction of the
commercially available substituted benzophenones with sodium
borohydride in methanol.
1
1.17 g, 65.0%; mp 106.7−108.5 °C; H NMR (300 MHz, CDCl3, 20
°C): δ = 3.79 (s, 6H; ArOCH3), 6.89 (d, J = 8.8 Hz, 4H; ArH), 7.06
(s, 1H; Ar2CH), 7.32 (d, J = 8.7 Hz, 4H; ArH); 13C NMR (75 MHz,
CDCl3, 20 °C): δ = 55.4 (ArOCH3), 79.5 (Ar2CH), 114.1, 128.7,
131.4, 159.8 (Ar); 19F NMR (282 MHz, CDCl3, 20 °C): δ = −138.1,
−148.6, −160.6 (F5Ar) ppm; elemental analysis calcd (%) for
C22O4F5H15 (438.34): C 60.28, H 3.45; found C 60.43, H 3.31
General Procedure for the Synthesis of Benzhydryl 2,4,6-
Trifluorobenzoates. The procedure is the same as previously
described for pentafluorobenzoates except that 2,4,6-trifluorobenzoyl
chloride was used instead the pentafluoro analogue. 4-Phenoxy- and 4-
methoxy-4′-phenoxybenzhydryl 2,4,6-trifluorobenzoate were obtained
as pale yellow oils, while the others were obtained as white crystals
(yield 71.2−81.1%).
4-Phenoxybenzhydrol, 4-methoxy-4′-methylbenzhydrol, and 4-methoxy-4′-
phenoxybenzhydrol were prepared according to the procedure given in
the ref 6a.
General Procedure for the Synthesis of Benzhydryl
Pentafluorobenzoates. A solution of pentafluorobenzoyl chloride
(3.3 mmol) in anhydrous benzene (10 mL) was added dropwise to the
previously prepared vigorously stirring solution of appropriate
benzhydrol (3.0 mmol) and pyridine (6.7 mmol) in anhydrous
benzene (30 mL). The reaction mixture was stirred for 12 h under an
atmosphere of argon at ambient temperature. Precipitated pyridinium
chloride was removed by filtration, and the excess of pyridine was
removed by 10% hydrochloric acid. The benzene layer was separated
and washed with water. After drying over anhydrous sodium sulfate,
benzene was evaporated in vacuo to give pale-yellow oil.
Recrystallization from light petroleum ether afforded white crystals
(yield 59−87%).
4-Phenoxybenzhydryl 2,4,6-trifluorobenzoate: (2-TFB) from 4-
phenoxybenzhydrol (0.70 g; 2.5 mmol), pyridine (0.44 g; 5.6 mmol),
and trifluorobenzoyl chloride (0.54 g; 2.8 mmol); yield 0.77 g, 70.0%);
1H NMR (300 MHz, CDCl3, 20 °C): δ = 6.70−7.45 (Ar2CH + ArH +
4-Methylbenzhydryl pentafluorobenzoate: (1-PFB) from 4-
methylbenzhydrol (0.60 g; 3.0 mmol), pyridine (0.53 g; 6.7 mmol),
and pentafluorobenzoyl chloride (0.77 g; 3.3 mmol); yield 0.96 g,
F2ArH); 13C NMR (75 MHz, CDCl3, 20 °C): δ = 78.3 (Ar2CH),
101.4 (t, J = 26.2, 2C; F3ArH), 118.7, 119.4, 123.8, 127.2, 128.3, 128.8,
129.0, 130.0, 134.4, 139.6, 156.9, 157.5 (Ar); 19F NMR (282 MHz,
CDCl3, 20 °C): δ = −101.6, −105.5 (F3Ar) ppm. MALDI-TOF/TOF
MS calcd for C26H17F3O3 [M + K+]: 473.0761; found: 473.0774.
4-Methoxybenzhydryl 2,4,6-trifluorobenzoate: (3-TFB) from 4-
methoxybenzhydrol (0.60 g; 2.8 mmol), pyridine (0.49 g; 6.2 mmol),
and trifluorobenzoyl chloride (0.60 g; 3.1 mmol); yield 0.74 g, 71.2%;
1
81.0%; mp 91.1−91.7 °C; H NMR (300 MHz, CDCl3, 20 °C): δ =
2.33 (s, 3H; ArCH3), 7.09−7.42 (m, 10H; Ar2CH); 13C NMR (75
MHz, CDCl3, 20 °C): δ = 21.3 (ArCH3), 79.9 (Ar2CH), 127.2, 127.4,
128.4, 128.8, 129.5, 136.3, 138.4, 139.3 (Ar); 19F NMR (282 MHz,
CDCl3, 20 °C): δ = −137.8, −148.5, −160.5 (F5Ar) ppm. MALDI-
TOF/TOF MS calcd for C21H13F5O2 [M + K+]: 431.0467; found:
431.0486.
1
mp 85.7−87.9 °C; H NMR (300 MHz, CDCl3, 20 °C): δ = 3.73 (s,
3H; ArOCH3), 6.67 (t, J = 8.4 Hz, 2H; F3ArH), 6.84 (d, J = 8.7 Hz,
2H; ArH), 7.04 (s, 1H; Ar2CH), 7.24−7.39 (m, 7H; ArH); 13C NMR
(75 MHz, CDCl3, 20 °C): δ = 55.4 (ArOCH3), 78.6 (Ar2CH), 101.4
(t, J = 26.2, 2C; F3ArH), 114.1, 121.8, 127.1, 128.2, 128.7, 129.0,
131.9, 139.9, 159.6 (Ar); 19F NMR (282 MHz, CDCl3, 20 °C): δ =
−101.9, −105.7 (F3Ar) ppm . MALDI-TOF/TOF MS calcd for
C21H15F3O3 [M + K+]: 411.0605; found: 411.0623.
4-Phenoxybenzhydryl pentafluorobenzoate: (2-PFB) from 4-
phenoxybenzhydrol (0.75 g; 2.7 mmol), pyridine (0.47 g; 6.0 mmol),
and pentafluorobenzoyl chloride (0.60 g; 3.0 mmol); yield 0.80 g,
1
59.0%; mp 76.7−77.6 °C; H NMR (300 MHz, CDCl3, 20 °C): δ =
6.96−7.44 (m, Ar2CH + ArH); 13C NMR (75 MHz, CDCl3, 20 °C): δ
= 79.5 (Ar2CH), 118.7, 119.5, 123.7, 127.1, 128.5, 128.9, 129.1, 130.0,
133.9, 139.2, 156.7, 157.8 (Ar); 19F NMR (282 MHz, CDCl3, 20 °C):
δ = −137.7, −148.2, −160.3 (F5Ar) ppm; elemental analysis calcd (%)
for C26O3F5H15 (470.38): C 66.38, H 3.21; found C 66.38, H 3.00.
4-Methoxybenzhydryl pentafluorobenzoate: (3-PFB) from 4-
methoxybenzhydrol (0.93 g; 4.3 mmol), pyridine (0.76 g; 9.6 mmol),
and pentafluorobenzoyl chloride (1.10 g; 4.8 mmol); yield 1.26 g,
4-Methoxy-4′-methylbenzhydryl 2,4,6-trifluorobenzoate: (4-
TFB) from 4-methoxy-4′-methylbenzhydrol (0.60 g; 2.6 mmol),
pyridine (0.46 g; 5.8 mmol), and trifluorobenzoyl chloride (0.56 g; 2.9
1
mmol); yield 0.73 g, 71.6%; mp 76.8−78.8 °C; H NMR (300 MHz,
CDCl3, 20 °C): δ = 2.28 (s, 3H; ArCH3), 3.73 (s, 3H; ArOCH3), 6.66
(t, J = 8.4 Hz, 2H; F3ArH), 6.81 (d, J = 8.8 Hz, 2H; ArH), 7.01 (s, 1H;
Ar2CH), 7.10−7.31 (m, 6H; ArH); 13C NMR (75 MHz, CDCl3, 20
1
71.0%; mp 85.1−86.4 °C; H NMR (300 MHz, CDCl3, 20 °C): δ =
3.79 (s, 3H; ArOCH3), 6.89 (d, J = 8.8 Hz, 2H; ArH), 7.09 (s, 1H;
8995
dx.doi.org/10.1021/jo3013308 | J. Org. Chem. 2012, 77, 8986−8998

The Journal of Organic Chemistry
Article
°C): δ = 21.3 (ArCH3), 55.4 (ArOCH3), 78.7 (Ar2CH), 101.3 (t, J =
26.1, 2C; F3ArH), 114.1, 127.1, 128.9, 129.4, 132.1, 137.0, 137.9, 159.6
(Ar); 19F NMR (282 MHz, CDCl3, 20 °C): δ = −102.0, −105.7
(F3Ar) ppm. MALDI-TOF/TOF MS calcd. for C22H17F3O3 [M + K+]:
425.0761; Found: 425.0768.
TOF/TOF MS Calcd for C22H19NO6 [M + (e−)]: 393.1207; found:
393.1204.
4,4′-Dimethoxybenzhydryl 4-cyanobenzoate: from 4,4′-dime-
thoxybenzhydrol (0.50 g; 2.1 mmol), pyridine (0.47 g; 5.9 mmol),
and 4-cyanobenzoyl chloride (0.49 g; 2.9 mmol); yield 0.43 g, 55.1%;
mp 128.1−130.0 °C; 1H NMR (300 MHz, CDCl3, 20 °C): δ = 3.73 (s,
6H; ArOCH3), 6.84 (d, J = 8.2 Hz, 4H; ArH), 7.01 (s, 1H; Ar2CH),
7.28 (d, J = 8.7 Hz, 4H; ArH), 7.68 (d, J = 8.8 Hz, 2H; NC-ArH), 8.15
(d, J = 8.8 Hz, 2H; NC-ArH); 13C NMR (75 MHz, CDCl3, 20 °C): δ
= 55.5 (ArOCH3), 78.1 (Ar2CH), 114.2, 116.6, 118.2, 127.9, 128.7,
130.4, 132.4, 134.4, 159.6 (Ar), 164.3 (CO). MALDI-TOF/TOF
MS calcd for C23H19NO4 [M + (e−)]: 373.1309; found: 373.1299.
4,4′-Dimethoxybenzhydryl 2-(trifluoromethyl)benzoate: from
4,4′-dimethoxybenzhydrol (0.50 g; 2.1 mmol), pyridine (0.47 g; 5.9
mmol), and 2-(trifluoromethyl)benzoyl chloride (0.61 g; 2.9 mmol);
yield 0.69 g, 79.3%); 1H NMR (300 MHz, CDCl3, 20 °C): δ = 3.75 (s,
6H; ArOCH3), 6.84 (d, J = 8.8 Hz, 4H; ArH), 7.03 (s, 1H; Ar2CH),
7.28 (d, J = 8.8 Hz, 4H; ArH), 7.51−7.55 (m, 2H; F3C−Ar), 7.67−
7.75 (m, 2H; F3C−Ar); 13C NMR (75 MHz, CDCl3, 20 °C): δ = 55.4
(ArOCH3), 78.4 (Ar2CH), 114.0, 121.7, 125.3, 126.8 (q, J = 5.4 Hz,
1C; F3C-Ar), 127.9, 128.8, 130.4, 131.3, 131.9, 132.1, 159.5 (Ar),
165.9 (CO); 19F NMR (282 MHz, CDCl3, 20 °C): δ = −59.8
(F3C−Ar) ppm. MALDI-TOF/TOF MS Calcd for C23H19F3O4 [M +
H+]: 417.1308; found: 417.1304.
4-Methoxy-4′-phenoxybenzhydryl 2,4,6-trifluorobenzoate: (5-
TFB) from 4-methoxy-4′-phenoxybenzhydrol (0.70 g; 2.3 mmol),
pyridine (0.40 g; 5.1 mmol), and trifluorobenzoyl chloride (0.49 g; 2.5
mmol); yield 0.86 g, 81.1%); 1H NMR (300 MHz, CDCl3, 20 °C): δ =
3.79 (s, 3H; ArOCH3), 6.69 − 7.38 (m, F3ArH + Ar2CH + ArH); 13
C
NMR (75 MHz, CDCl3, 20 °C): δ = 55.4 (ArOCH3), 78.4 (Ar2CH),
101.4 (t, J = 26.1, 2C; F3ArH), 114.1, 118.7, 119.4, 123.7, 128.8, 130.0,
132.0, 134.8, 157.0, 157.3, 159.5 (Ar); 19F NMR (282 MHz, CDCl3,
20 °C): δ = −101.7, −105.7 (F3Ar) ppm. MALDI-TOF/TOF MS
Calcd for C27H19F3O4 [M + H+]: 465.1308; found: 465.1298.
4,4′-Dimethoxybenzhydryl 2,4,6-trifluorobenzoate: (6-TFB)
from 4,4′-dimethoxybenzhydrol (0.60 g; 2.5 mmol), pyridine (0.43
g; 5.4 mmol), and trifluorobenzoyl chloride (0.53 g; 2.7 mmol); yield
1
0.75 g, 75.8%; mp 98.5−99.9 °C; H NMR (300 MHz, CDCl3, 20
°C): δ = 3.78 (s, 6H; ArOCH3), 6.71 (t, J = 8.4 Hz, 2H; F3ArH), 6.89
(d, J = 8.8 Hz, 4H; ArH), 7.05 (s, 1H; Ar2CH), 7.34 (d, J = 8.8 Hz,
4H; ArH); 13C NMR (75 MHz, CDCl3, 20 °C): δ = 55.4 (ArOCH3),
78.5 (Ar2CH), 101.3 (t, J = 26.6, 2C; F3ArH), 114.1, 128.7, 132.1,
159.5 (Ar); 19F NMR (282 MHz, CDCl3, 20 °C): δ = −102.1, −105.8
(F3Ar) ppm; elemental analysis calcd (%) for C22O4F3H17 (402.36): C
65.67, H 4.26; found C 65.59, H 4.42.
4,4′-Dimethoxybenzhydryl 3,5-dichlorobenzoate: from 4,4′-dime-
thoxybenzhydrol (0.50 g; 2.1 mmol), pyridine (0.47 g; 5.9 mmol), and
3,5-dichlorobenzoyl chloride (0.61 g; 2.9 mmol); yield 0.71 g, 80.7%;
4,4′-Dimethoxy 4-nitrobenzoate, and 4-methoxybenzhydryl 4-nitroben-
zoate were prepared according to the procedure given in the ref 5a.
4,4′-Dimethoxybenzhydryl 2-Nitrobenzoate. A solution of 2-
nitrobenzoyl chloride (0.53 g, 2.9 mmol) in anhydrous benzene (10
mL) was added dropwise to the previously prepared vigorously stirring
solution of 4,4′-dimethoxybenzhydrol (0.50 g, 2.1 mmol) and pyridine
(0.45 g, 5.7 mmol) in anhydrous benzene (30 mL). The reaction
mixture was stirred overnight under an atmosphere of argon at
ambient temperature. Precipitated pyridinium chloride was removed
by filtration, and the excess of pyridine was removed by 10%
hydrochloric acid. The benzene layer was separated and washed with
water. After being dried over anhydrous sodium sulfate, the benzene
was evaporated in vacuo. The crude product was dissolved in diethyl
ether (30 mL), and then about 30 mL of concd aq NaOH was added.
The mixture was stirred for 1 h, and then the organic layer was
separated and washed with water. After being dried over anhydrous
sodium sulfate, the solvent was removed in vacuo to a give pale-yellow
1
mp 91.8−93.8 °C; H NMR (300 MHz, CDCl3, 20 °C): δ = 3.74 (s,
6H; ArOCH3), 6.85 (d, J = 8.8 Hz, 4H; ArH), 6.99 (s, 1H; Ar2CH),
7.27 (d, J = 8.8 Hz, 4H; ArH), 7.49 (t, J = 2.0 Hz, 1H; 3,5-di-Cl-ArH),
7.91 (d, J = 2.0 Hz, 2H; 3,5-di-Cl-ArH); 13C NMR (75 MHz, CDCl3,
20 °C): δ = 55.5 (ArOCH3), 78.1 (Ar2CH), 114.1, 128.3, 128.7, 132.1,
133.0, 133.4, 135.5, 159.6 (Ar), 163.6 (CO). MALDI-TOF/TOF
MS Calcd for C22H18Cl2O4 [M + H+]: 417.0655; found: 417.0642.
4,4′-Dimethoxybenzhydryl 2-methoxybenzoate: from 4,4′-dime-
thoxybenzhydrol (0.50 g; 2.1 mmol), pyridine (0.47 g; 5.9 mmol), and
2-methoxybenzoyl chloride (0.49 g; 2.9 mmol); yield 0.60 g, 75.9%);
1H NMR (300 MHz, CDCl3, 20 °C): δ = 3.73 (s, 6H; ArOCH3), 3.84
(s, 3H; ArOCH3), 6.83 (d, J = 8.8 Hz, 4H; ArH), 6.92 (q, J = 6.3 Hz,
2H; ArOCH3), 7.00 (s, 1H; Ar2CH), 7.32 (d, J = 8.8 Hz, 4H; ArH),
7.41 (t, J = 9.1 Hz, 1H; ArOCH3), 7.84 (d, J = 8.0 Hz, 1H; ArOCH3);
13C NMR (75 MHz, CDCl3, 20 °C): δ = 55.4 (ArOCH3), 56.1
(ArOCH3), 76.8 (Ar2CH), 112.3, 114.0, 120.3, 128.0, 128.7, 132.0,
133.2, 133.8, 159.3, 159.7(Ar), 165.3 (CO) ppm. MALDI-TOF/
TOF MS Calcd for C23H22O5 [M + K+]: 417.1099; found: 417.1098.
4,4′-Dimethoxybenzhydryl benzoate: from 4,4′-dimethoxybenzhy-
drol (0.50 g; 2.1 mmol), pyridine (0.47 g; 5.9 mmol), and benzoyl
1
oil (0.60 g; yield 74.5%). H NMR (300 MHz, CDCl3, 20 °C): δ =
3.73 (s, 6H, ArOCH3), 6.84 (d, J = 8.8 Hz, 4H; ArH), 7.00 (s, 1H,
Ar2CH), 7.25 (d, J = 8.6 Hz, 4H; ArH), 7.51−7.61 (m, 2H, O2N-
ArH), 7.68 (d, J = 6.9 Hz, 1H; O2N-ArH), 7.81 (d, J = 6.7 Hz, 1H;
O2N-ArH); 13C NMR (75 MHz, CDCl3, 20 °C): δ = 55.4 (ArOCH3),
79.0 (Ar2CH), 114.1, 124.0, 127.6, 128.8, 130.2, 131.7, 132.0, 132.9,
159.6 (Ar), 164.5 (CO) ppm. MALDI-TOF/TOF MS Calcd for
C22H19NO6 [M + (e−)]: 393.1207; found: 393.1210.
1
chloride (0.41 g; 2.9 mmol); yield 0.49 g, 67.1%); H NMR (300
MHz, CDCl3, 20 °C): δ = 3.77 (s, 6H; ArOCH3), 6.88 (d, J = 8.7 Hz,
4H; ArH), 7.06 (s, 1H; Ar2CH), 7.35 (d, J = 8.3 Hz, 4H; ArH), 7.44
(t, J = 7.8 Hz, 2H; ArH), 7.55 (t, J = 7.5 Hz, 1H; ArH), 8.12 (d, J = 6.9
Hz, 2H; ArH); 13C NMR (75 MHz, CDCl3, 20 °C): δ = 55.5
(ArOCH3), 77.0 (Ar2CH), 114.1, 127.9, 128.7, 129.9, 130.6, 132.9,
133.2, 159.4 (Ar), 165.8 (CO) ppm. MALDI-TOF/TOF MS calcd
for C22H20O4 [M + K+]: 387.0993; found: 387.0975.
Synthesis of Other 4,4′-Dimethoxybenzhydryl Benzoates
and 4-Methoxybenzhydryl 3-Nitrobenzoate. The procedure is
the same as previously described for 4,4′-dimethoxybenzhydryl 2-
nitrobenzoate, except that the appropriate benzoyl chloride was used.
For the synthesis of 2,4,6-trimethylbenzoate, dichloromethane instead
of benzene was used as a solvent. Most of products were obtained as
pale-yellow oils, except for 4,4′-dimethoxybenzhydryl 4-cyanoben-
zoate, 4,4′-dimethoxybenzhydryl 3,5-dichlorobenzoate, and 4,4′-
dimethoxybenzhydryl 2,4,6-trimethylbenzoate, which were obtained
as white crystals.
4,4′-Dimethoxybenzhydryl 2,4,6-trimethylbenzoate: from 4,4′-
dimethoxybenzhydrol (0.50 g; 2.1 mmol), pyridine (0.47 g; 5.9
mmol), and 2,4,6-trimethylbenzoyl chloride (0.54 g; 2.9 mmol); yield
1
0.68 g, 82.9%; mp 71.6−73.2 °C; H NMR (300 MHz, CDCl3): δ =
2.12 (s, 6H; (CH3)3ArH), 2.22 (s, 3H; (CH3)3ArH), 3.74 (s, 6H;
ArOCH3), 6.78−6.84 (s, 2H; (CH3)3ArH + d, J = 8.8 Hz, 4H;
ArOCH3), 7.06 (s, 1H; Ar2CH), 7.29 (d, J = 8.6 Hz, 4H; ArH). 13C
NMR (75 MHz, CDCl3): δ = 19.8, 21.3 ((CH3)3ArH), 55.4
(ArOCH3), 77.1 (Ar2CH), 114.0, 128.5, 128.9, 131.2, 132.7, 135.2,
139.4, 159.4 (Ar), 170.0 (CO) ppm. MALDI-TOF/TOF MS calcd
for C25H26O4 [M + K+]: 429.1462; found: 429.1474.
4,4′-Dimethoxybenzhydryl 3-nitrobenzoate: from 4,4′-dimethox-
ybenzhydrol (0.50 g; 2.1 mmol), pyridine (0.47 g; 5.9 mmol), and 3-
1
nitrobenzoyl chloride (0.54 g; 2.9 mmol); yield 0.55 g, 70.5%); H
NMR (300 MHz, CDCl3, 20 °C): δ = 3.79 (s, 6H; ArOCH3), 6.90 (d,
J = 8.8 Hz, 4H; ArH), 7.10 (s, 1H; Ar2CH), 7.35 (d, J = 8.8 Hz, 4H;
ArH), 7.64 (t, J = 8.1 Hz, 1H; O2N-ArH), 8.38−8.64 (m, 2H; O2N-
ArH), 8.91 (s, 1H; O2N-ArH); 13C NMR (75 MHz, CDCl3, 20 °C): δ
= 55.5 (ArOCH3), 78.2 (Ar2CH), 114,2, 124.8, 127.7, 128.7, 129.9,
132.1, 132.3, 135.6, 148.5, 159.6 (Ar), 163.8 (CO) ppm. MALDI-
4-Methoxybenzhydryl 3-nitrobenzoate: from 4-methoxybenzhy-
drol (0.50 g; 2.3 mmol), pyridine (0.51 g; 6.4 mmol), and 3-
1
nitrobenzoyl chloride (0.60 g; 3.2 mmol); yield 0.61 g, 72.8%); H
NMR (300 MHz, CDCl3, 20 °C): δ = 3.79 (s, 6H; ArOCH3), 6.90 (d,
8996
dx.doi.org/10.1021/jo3013308 | J. Org. Chem. 2012, 77, 8986−8998

The Journal of Organic Chemistry
Article
J = 8.8 Hz, 4H; ArH), 7.13 (s, 1H; Ar2CH), 7.31−7.45 (m, 7H; ArH),
7.66 (t, J = 8.1 Hz, 1H; O2N-ArH), 8.40−8.45 (m, 2H; O2N-ArH),
8.93 (s, 1H; O2N-ArH); 13C NMR (75 MHz, CDCl3, 20 °C): δ = 55.5
(ArOCH3), 78.4 (Ar2CH), 114.2, 124.8, 127.1, 127.7, 128.3, 128.8,
129.0, 129.9, 131.9, 132.2, 135.6, 139.9, 148.5, 159.7 (Ar), 163.8 (C
O) ppm. MALDI-TOF/TOF MS Calcd for C21H17NO5 [M + K+]:
402.0738; found: 402.0746.
(IEFPCM, solvent = water). Frequency calculations were performed at
the same level of theory, and thermal corrections were calculated at 1
atm and 298 K. Cartesian coordinates for all optimized geometries and
corresponding energies are given in the Supporting Information.
ASSOCIATED CONTENT
* Supporting Information
■
S
Kinetic Methods. Solvolysis rate constants were measured
conductometrically. Freshly prepared solvents (30 mL) were thermo-
statted ( 0.1 °C) at a given temperature for several minutes prior to
addition of the substrate. Typically, 30−60 mg of substrate was
dissolved in 0.10−0.15 mL of dichloromethane and injected into the
solvent. An increase of the conductivity during solvolysis was
monitored automatically by means of a WTW LF 530 conductometer,
using a Radiometer two-pole conductivity cell (CDC641T). Individual
rate constants were obtained by least-squares fitting of the conductivity
data to the first-order kinetic equation for three to four half-lives. The
rate constants were averaged from at least three measurements. To
achieve a complete ionization of a liberated acid, a proton sponge base
[1,8-bis(dimethylamino)naphthalene] was added in a range of
concentration for each given aqueous binary mixture presented in
Table S3 in Supporting Information. Calibration showed the linear
response of conductivity in the presented ranges of concentrations of
the proton sponge base and benzoic acids liberated in examined
solvolyses.
Correlations of log k vs Ef in aq ethanol solutions and aq
acetone solutions, correlations between experimental and
calculated ΔG⧧ for heterolysis and ΔH⧧ for 2-oxyethyl-LG
epoxydation, Hammett plots, calculated ΔG⧧ and rate
constants for solvolysis of benzoates in aqueous ethanol
1
solutions and their Nf parameters, intrinsic barriers, H NMR
and 13C NMR spectra, optimized geometries (Cartesian
coordinates), and corresponding energies for studied benzoates.
This material is available free of charge via the Internet at
http://pubs.acs.org.
AUTHOR INFORMATION
Corresponding Author
*E-mail: bdenegri@pharma.hr; kronja@pharma.hr.
■
Notes
The authors declare no competing financial interest.
Computational Methods. All calculations were carried out using
the Gaussian 09 program suit.14 Unless otherwise noted geometry
optimizations were performed without any symmetry constraints at
the B3LYP/6-311+G(2d,p) level using the integral equation formalism
model in a polarizable continuum (IEFPCM) representing water as
solvent.15 For charged anionic model-substrates, water has been
chosen as a solvent because of its high bulk dielectric constant (ε =
78.3553). Because of solvation effects, charged systems demand
additional structural relaxation in solution with respect to the gas
phase. For that reason, geometry optimizations were performed by
using the polarizable continuum model, which mimics a solvent, rather
than in the gas phase with subsequent PCM single-point calculation.
The most stable conformation found for a given ground-state structure
and a conformation of the corresponding transition-state structure
were used for calculation of the enthalpies of activation for both types
of reactions: anchimerically assisted heterolytic dissociations of 2-
oxyethyl benzoates and thermoneutral methyl exchange reactions. The
transition-state structures were located using relaxed potential energy
surface (PES) scans for heterolytic bond cleavage, after which the
structure with the highest energy on the given PES was fully
optimized. Stationary points were characterized either as minima
(NImag = 0) or as saddle points (NImag = 1) by B3LYP/6-
311+G(2d,p) level frequency calculations, which were also used to
calculate the thermal corrections to enthalpies and free energies at 1
atm and 298 K. Values of imaginary frequencies for the anchimerically
assisted heterolytic dissociation, which correspond to the oxygen
attack on the α-carbon and simultaneous translation of the leaving
group, vary between −550 and −560 cm−1 for all examined 2-oxyethyl
benzoates. The imaginary frequencies for thermoneutral methyl
exchange reactions are associated with the translation of the methyl
group between two identical benzoate leaving groups, and their values
also vary between −550 and −560 cm−1. Intrinsic reaction coordinate
(IRC) calculations have been employed to verify that the transition-
state structures are associated with the given processes. The complexes
between epoxide and benzoate leaving groups were not located despite
a downward energy path with further elongation of distance between
the α-carbon and leaving groups behind the TS. SCRF-PCM
calculations, because of the stabilization energy of the highly polar
solvent, separate ion−molecule complexes that follow the TS which
results in missing the product minimum in the displacement reaction
energy path.10a Hence, products of the heterolytic reactions were
considered as free species.
ACKNOWLEDGMENTS
■
We gratefully acknowledge the financial support of this research
by the Ministry of Science, Education and Sport of the Republic
of Croatia (grant no. 006-0982933-2963). Computational
resources provided by Isabella cluster (isabella.srce.hr) at
Zagreb University Computing Centre (Srce) were used for
this research.
REFERENCES
■
(1) (a) Bentley, T. W.; Schleyer, P. v. R. Adv. Phys. Org. Chem. 1977,
14, 1−67. (b) Bentley, T. W.; Llewellyn, G. Prog. Phys. Org. Chem.
1990, 17, 121−159. (c) Bentley, T. W.; Llewellyn, G.; Ryu, Z. H. J.
Org. Chem. 1998, 63, 4654−4659. (d) Bentley, T. W.; Dau-Schmidt, J.
P.; Llewellyn, G.; Mayr, H. J. Org. Chem. 1992, 57, 2387−2392.
(e) Kevill, D. N. Advances in Quantitative Structure-Property Relation-
ships; Charton, M., Ed.; JAI Press: Greenwich, CT, 1996; Vol. 1, pp
81−115.
(2) (a) Brown, H. C.; Fletcher, R. S. J. Am. Chem. Soc. 1949, 71,
1845−1854. (b) Bingham, R. C.; Schleyer, P. v. R. J. Am. Chem. Soc.
1971, 93, 3189−3199. (c) Slutsky, J.; Bingham, R. C.; Schleyer, P. v.
R.; Dickason, W. C.; Brown, H. C. J. Am. Chem. Soc. 1974, 96, 1969−
1970.
(3) (a) Schleyer, P. v. R.; Jemmis, E. D.; Spitznagel, G. W. J. Am.
Chem. Soc. 1985, 107, 6393−6394. (b) Richard, J. P.; Amyes, T. L.;
Rice, D. J. J. Am. Chem. Soc. 1993, 115, 2523−2524. (c) Apeloig, Y.;
Biton, R.; Abu-Freih, A. J. Am. Chem. Soc. 1993, 115, 2522−2523.
(d) Kirmse, W.; Goer, B. J. Am. Chem. Soc. 1990, 112, 4556−4557.
(4) Streidl, N.; Denegri, B.; Kronja, O.; Mayr, H. Acc. Chem. Res.
2010, 43, 1537−1549.
́
(5) (a) Denegri, B.; Streiter, A.; Juric, S.; Ofial, A. R.; Kronja, O.;
Mayr, H. Chem.Eur. J. 2006, 12, 1648−1656; Chem.Eur. J. 2006,
12, 5415−5415. (b) Denegri, B.; Minegishi, S.; Kronja, O.; Mayr, H.
Angew. Chem., Int. Ed. 2004, 43, 2302−2305.
(6) (a) Denegri, B.; Kronja, O. J. Org. Chem. 2007, 72, 8427−8433.
(b) Denegri, B.; Kronja, O. J. Org. Chem. 2009, 74, 5927−5933.
(c) Matic,
6024. (d) Juric,
3851−3854. (e) Juric,
2011, 25, 147−152.
(7) Matic, M.; Juric,
13, 2012−2024.
́
M.; Denegri, B.; Kronja, O. Eur. J. Org. Chem. 2010, 6019−
́
S.; Denegri, B.; Kronja, O. J. Org. Chem. 2010, 75,
́
S.; Denegri, B.; Kronja, O. J. Phys. Org. Chem.
The reference 2-oxyethyl benzoates and the corresponding
transition-state structures were also optimized using the M06-2X
functional17 and 6-311+G(2d,p) basis set in a polarizable continuum
́
́
S.; Denegri, B.; Kronja, O. Int. J. Mol. Sci. 2012,
(8) Denegri, B.; Kronja, O. J. Phys. Org. Chem. 2009, 22, 495−503.
8997
dx.doi.org/10.1021/jo3013308 | J. Org. Chem. 2012, 77, 8986−8998

The Journal of Organic Chemistry
Article
(9) (a) Boyd, D. B. J. Org. Chem. 1985, 50, 885−886. (b) Marshall,
D. R.; Thomas, P. J.; Stirling, C. J. M. J. Chem. Soc., Chem. Commun.
1975, 940−941.
(10) (a) Castejon, H.; Wiberg, K. B. J. Am. Chem. Soc. 1999, 121,
2139−2146. (b) Castejon, H.; Wiberg, K. B.; Sklenak, S.; Hinz, W. J.
Am. Chem. Soc. 2001, 123, 6092−6097. (c) Martinez, A. G.; Vilar, E.
T.; Barcina, J. O.; Cerero, S. D. J. Org. Chem. 2005, 70, 10238−10246.
(d) Chen, X.; Regan, C. K.; Craig, S. L.; Krenske, E. H.; Houk, K. N.;
Jorgensen, W. L.; Brauman, J. I. J. Am. Chem. Soc. 2009, 131, 16162−
16170. (e) Yamataka, H.; Aida, M. Chem. Phys. Lett. 1998, 289, 105−
109. (f) Ren, Y.; Yamataka, H. Chem.Eur. J. 2007, 13, 677−682.
(g) Melo, A.; Alfaia, A. J. I.; Reis, J. C. R.; Calado, A. R. T. J. Phys.
Chem. B 2006, 110 (4), 1877−1888.
(11) Aggarwal, V. K.; Harvey, J. N.; Robiette, R. Angew. Chem., Int.
Ed. 2005, 44, 5468−5471.
(12) (a) Rodriquez, C. F.; Williams, I. H. J. Chem. Soc., Perkin Trans.
̌
́ ̌ ́
S.; Vrcek, V.; Gjuranovic, Z.;
2 1997, 4, 959−965. (b) Malnar, I.; Juric,
Mihalic,
́
Z.; Kronja, O. J. Org. Chem. 2002, 67, 1490−1495.
(c) Schreiner, P. R.; Schleyer, P. v. R.; Schaefer, H. F. J. Org. Chem.
1997, 62, 4216−4228.
(13) Wang, C.; Fu, Y.; Guo, Q.-X.; Liu, L. Chem.Eur. J. 2010, 16,
2586−2598.
(14) Frisch, M. J.; Trucks, G. W.; Schlegel, H. B.; Scuseria, G. E.;
Robb, M. A.; Cheeseman, J. R.; Scalmani, G.; Barone, V.; Mennucci,
B.; Petersson, G. A.; Nakatsuji, H.; Caricato, M.; Li, X.; Hratchian, H.
P.; Izmaylov, A. F.; Bloino, J.; Zheng, G.; Sonnenberg, J. L.; Hada, M.;
Ehara, M.; Toyota, K.; Fukuda, R.; Hasegawa, J.; Ishida, M.; Nakajima,
T.; Honda, Y.; Kitao, O.; Nakai, H.; Vreven, T.; Montgomery, J. A., Jr.
; Peralta, J. E.; Ogliaro, F.; Bearpark, M.; Heyd, J. J.; Brothers, E.;
Kudin, K. N.; Staroverov, V. N.; Kobayashi, R.; Normand, J.;
Raghavachari, K.; Rendell, A.; Burant, J. C.; Iyengar, S. S.; Tomasi,
J.; Cossi, M.; Rega, N.; Millam, J. M.; Klene, M.; Knox, J. E.; Cross, J.
B.; Bakken, V.; Adamo, C.; Jaramillo, J.; Gomperts, R.; Stratmann, R.
E.; Yazyev, O.; Austin, A. J.; Cammi, R.; Pomelli, C.; Ochterski, J. W.;
Martin, R. L.; Morokuma, K.; Zakrzewski, V. G.; Voth, G. A.; Salvador,
P.; Dannenberg, J. J.; Dapprich, S.; Daniels, A. D.; Farkas, O.;
Foresman, J. B.; Ortiz, J. V.; Cioslowski, J.; Fox, D. J. Gaussian 09,
Revision A.02; Gaussian, Inc., Wallingford, CT, 2009.
(15) Tomasi, J.; Mennucci, B.; Cammi, R. Chem. Rev. 2005, 105,
2999−3093.
(16) Kevill, D. N.; Kolwyck, K. C.; Shold, D. M.; Kim, C. B. J. Am.
Chem. Soc. 1973, 95, 6022−6027.
(17) (a) Zhao, Y.; Truhlar, D. G. Theor. Chem. Acc. 2008, 120, 215−
241. (b) Zhao, Y.; Truhlar, D. G. Acc. Chem. Res. 2008, 41, 157−167.
(18) Sigma values were taken from Lowry, T. H.; Richardson, K. S.
Mechanism and Theory in Organic Chemistry, 3rd ed.; Harper and Row
Publishers: New York, 1987.
(19) Marcus, R. A. J. Phys. Chem. 1968, 72, 891−899.
(20) (a) Richard, J. P. Tetrahedron 1995, 51, 1535−1573. (b) Richard,
J. P.; Amyes, T. L.; Williams, K. B. Pure Appl. Chem. 1998, 70, 2007−
2014. (c) Richard, J. P.; Amyes, T. L.; Toteva, M. M. Acc. Chem. Res.
2001, 34, 981−988.
(21) (a) Bernasconi, C. F. Adv. Phys. Org. Chem. 1992, 27, 119−238.
(b) Bernasconi, C. F. Acc. Chem. Res. 1992, 25, 9−16.
(22) Wurthwein, E.-U.; Lang, G.; Schappele, L. H.; Mayr, H. J. Am.
̈
Chem. Soc. 2002, 124, 4084−4092.
(23) (a) Amyes, T. L.; Stevens, I. W.; Richard, J. P. J. Org. Chem.
1993, 58, 6057−6066. (b) Richard, J. P.; Amyes, T. L.; Bei, L.;
Stubblefield, V. J. Am. Chem. Soc. 1990, 112, 9513−9519.
(24) (a) Horn, M.; Mayr, H. Chem.Eur. J. 2010, 16, 7469−7477.
(b) Horn, M.; Mayr, H. Chem.Eur. J. 2010, 16, 7478−7487.
(25) Wiberg, K. B. J. Org. Chem. 2002, 67, 4787−4794.
8998
dx.doi.org/10.1021/jo3013308 | J. Org. Chem. 2012, 77, 8986−8998