Correlation analysis of carbonyl carbon 13C NMR chemical shifts, IR absorption frequencies and rate coefficients of nucleophilic acyl substitutions. A novel explanation for the substituent dependence of reactivity

Article abstract of DOI:10.1039/a900189a

Rate coefficients of nucleophilic acyl substitutions, carboxylate carbon ¹³C NMR chemical shift values and v(C=O) frequencies of several series of aryl and acyl substituted aryl acetates or alkyl benzoates have been investigated. An increasing electron-withdrawal by the acyl or aryl substituents results in higher reaction rates, upfield ¹³C NMR chemical shifts and higher frequencies of the C=O stretching. Good correlations are observed for the log k versus δ_C(C=O) plots. The increase of the reaction rate with increased electron density at the C=O carbon (as proved by ¹³C NMR shifts) contradicts the previous concept of increased electrophilicity of the carbonyl carbon by electron-withdrawing substituents. The rate increase is now attributed to the decrease of the ester ground state resonance stabilization caused by electron-withdrawing substituents. The use of log k versus δ_c(C=O) correlations is presented as a practical method to evaluate rate coefficients especially for compounds for which Hammett type correlations cannot be used.

Full text of DOI:10.1039/a900189a

View Article Online / Journal Homepage / Table of Contents for this issue
13
Correlation analysis of carbonyl carbon C NMR chemical shifts,
IR absorption frequencies and rate coefficients of nucleophilic acyl
substitutions. A novel explanation for the substituent dependence
of reactivity
Helmi Neuvonen* and Kari Neuvonen
Department of Chemistry, University of Turku, FIN-20014 Turku, Finland
Received (in Cambridge) 6th January 1999, Accepted 26th April 1999
Rate coefficients of nucleophilic acyl substitutions, carboxylate carbon 13C NMR chemical shift values and ν(C᎐O)
᎐
frequencies of several series of aryl and acyl substituted aryl acetates or alkyl benzoates have been investigated. An
increasing electron-withdrawal by the acyl or aryl substituents results in higher reaction rates, upfield 13C NMR
chemical shifts and higher frequencies of the C᎐O stretching. Good correlations are observed for the log k versus
᎐
δ (C᎐O) plots. The increase of the reaction rate with increased electron density at the C᎐O carbon (as proved by 13
C
᎐
᎐
C
NMR shifts) contradicts the previous concept of increased electrophilicity of the carbonyl carbon by electron-
withdrawing substituents. The rate increase is now attributed to the decrease of the ester ground state resonance
stabilization caused by electron-withdrawing substituents. The use of log k versus δ (C᎐O) correlations is presented
᎐
C
as a practical method to evaluate rate coefficients especially for compounds for which Hammett type correlations
cannot be used.
The significance of acyl compounds is well-established in
organic chemistry and in biochemistry, and the understanding
of the factors affecting their reactivity is of primary import-
ance. Most acyl transfer reactions are thought to occur via a
tetrahedral intermediate as shown in Scheme 1. Reactivity of
ity of the reaction to the influence of the substituents. Although
reaction conditions, such as solvent or temperature, also affect
ρ, the most important factor controlling its value is the sus-
ceptibility of the reaction to electronic effects. Because of the
definition of σ and σ* values, the reaction constants ρ and ρ*
are positive for reactions facilitated by electron-withdrawing
substituents and negative for those reactions facilitated by elec-
tron donors. Substituent effects on the rate of a reaction reflect
the effect of substituents on the activation energy and they
can therefore reflect the substituent effect on the ground state
and/or transition state.2 It is usually considered that a positive ρ
means the development of a negative charge (or reduction of a
positive charge) at the reaction site in the transition state and
correspondingly a negative ρ means that a positive charge
is developed (or a negative charge is lost). For example, for
alkaline hydrolysis of ethyl benzoates the ρ value 2.27 is con-
sistent with a mechanism where a negative charge is developed
at the reaction site in the transition state. Furthermore, it is
usually considered that the electron-withdrawing substituents
facilitate the nucleophilic attack by increasing the electrophilic-
O
N
O
C
R
X
R
X
N
O
C
R
N
X
Scheme 1
the carbonyl carbon of RCOX towards a nucleophilic attack is
known to depend on the electron-withdrawing ability of R or
X as reflected by the linear free energy correlations often
observed.1 The Hammett equation [eqn. (1)] is most often used
ity of the C᎐O carbon through the reduction in its electron
᎐
density.1b,c
log (k/ko) = ρσ
(1)
13C NMR spectroscopy is a useful tool. Especially for aro-
᎐
᎐
matic compounds containing a C᎐O, C᎐C, C᎐C, C᎐N or C᎐N
᎐
᎐
᎐
᎐
᎐
to correlate reaction rates and equilibria of side-chain reactions
of para- and meta-substituted aromatic compounds. The
standard reaction by which the substituent constants σ have
been defined is the aqueous dissociation equilibrium of para-
and meta-substituted benzoic acids. In eqn. (1), k is a rate
coefficient for the para- or meta-substituted compound, ko that
for the unsubstituted derivative, and ρ is the Hammett reaction
constant.1b–d For aliphatic compounds an analogous relation-
ship [eqn. (2)] is known as the Taft equation, where the polar
substituent constants σ* for groups R have been determined
with the aid of alkaline and acid hydrolysis of esters RCOORЈ.
The rate coefficient ko corresponds to the reaction for R = CH3.
In eqn. (2), ρ* is the polar reaction constant.1b–d
system in their side chain, the substituent chemical shifts of
these unsaturated carbons have been shown to be systematic
and electronic in origin reflecting mainly the changes in the
π-electron density of the carbon in question.3–5 The upfield shift
of the carbon resonance corresponds to the increasing electron
shielding of the carbon. The π-electron density calculations
performed for some series of aromatic side chain derivatives
are consistent with the NMR data.3b,d,e,4
Inspection of the previous works of Bromilow et al.,3c
O’Connor et al.6 and Dell’Erba et al.,7 shows that electron-
withdrawing substituents in the aromatic acyl or alcoholic
moiety of esters cause upfield shifts of the carbonyl carbon
resonance. This is inconsistent with the usual concept that the
electrophilicity of the carbonyl carbon increases with electron-
withdrawing substituents. That contradiction has not been
discussed previously. The aim of our work was to study system-
atically the correlations between the reactivity of carboxylic
log (k/ko) = ρ*σ*
(2)
The value of the reaction constant ρ or ρ* shows the sensitiv-
J. Chem. Soc., Perkin Trans. 2, 1999, 1497–1502
1497

View Article Online
Table 1 Carbonyl carbon 13C NMR chemical shifts, IR C᎐O stretch-
ing frequencies and rate coefficients of the neutral hydrolysis in 20%
acetonitrile–water at 298.2 K for phenyl substituted phenyl dichloro-
acetates CHCl2COOC6H4X
acid esters towards a nucleophilic attack and the 13C NMR
chemical shift of the carboxylate carbon of the ground state
molecules to clarify the origin of the effect in question exerted
by the substituents in the aromatic ring and also by the sub-
stituents in the aliphatic acyl group. In addition, we see the use
᎐
X
δ(C᎐O)a/ppm
ν(C᎐O)/cmϪ1
ko b/10Ϫ3 sϪ1
᎐
᎐
of reactivity versus δ (C᎐O) correlations as a potential method
᎐
C
to estimate the reactivity of compounds with known C᎐O
᎐
4-NO2
3-NO2
4-CN
4-Cl
3-Cl
4-Br
162.13
162.43
162.20
162.78
162.65
162.70
163.02
163.20
163.37
1788; 1767
1775c
1781
14.4 0.2
8.43 0.09
8.74 0.09
1.36 0.04
1.88 0.03
1.27 0.03
0.476 0.008
0.273 0.002
0.349 0.006
carbon 13C NMR chemical shifts.
13C NMR shift, ν(C᎐O) frequency and reactivity measure-
᎐
1762c
1783
ments have been performed with the series 1 and 2. Further-
more, we have correlated existing reactivity and spectroscopic
results for compounds of series 3 and 4. In these cases, one set
of data used is always obtained from one source.
1767
1777
1765
1763
H
4-CH3
4-OCH3
CHCl2COOC6H4X
a In CDCl3. b The ko values are averages of four determinations and the
error shown is a standard deviation. c Ref 8c.
1
X = 4-NO2, 4-CN, 4-Cl, 4-Br, H, 4-Me, 4-OMe, 3-NO2 or 3-Cl
1
The spectra were measured with a H broad-band decoupling
technique. The chemical shifts are expressed in ppm relative to
TMS used as an internal reference. The infrared spectra were
recorded for capillary films or KBr discs on a Mattson Galaxy
6020 FTIR spectrometer.
RCOOC6H4NO2
2
R = CH3, CH2Cl, CHCl2, CCl3 or CF3
Kinetics
Reaction rates of the neutral hydrolysis of substituted phenyl
dichloroacetates in 20% (v/v) acetonitrile–water were deter-
mined spectrophotometrically by observing the phenol release
at a suitable wavelength (4-nitrophenol 310 nm, 3-nitrophenol
328 nm, 4-cyanophenol 250 nm, 4-chlorophenol 280 nm, 3-
chlorophenol 280 nm, 4-bromophenol 280 nm, phenol 270 nm,
4-methylphenol 280 nm, 4-methoxyphenol 290 nm). The reac-
tion of imidazole with 4-nitrophenyl trichloroacetate in
acetonitrile was followed by detecting the increase in the
absorption of 4-nitrophenol (310 nm) in first-order conditions
(at least 80-fold excess of imidazole). A Gilford 2600 spectro-
photometer equipped with a Gilford Thermoset Temperature
Controller was used. The temperature was accurate to 0.1 ЊC.
In a typical experiment 250 mm3 of the reaction solution was
prethermostatted in the optical cell (300 mm3). In hydrolysis
reactions 3 mm3 of 0.03 M and in the reaction of imidazole
with 4-nitrophenyl trichloroacetate 0.75–1.4 mm3 of 0.01 M
acetonitrile solution of the ester was added into the cell, the
solution was shaken throughly and the cell was placed into the
thermostatted cell compartment. The reactions were followed
for six half-lives and the final value was observed after ten half-
lives if needed.
The first-order rate coefficients of the neutral hydrolysis were
calculated by the method of Guggenheim.12 The standard devi-
ations of the individual rate coefficients were ≤0.3%. It was
checked that good linear plots of ln(A∞ Ϫ At) versus time were
obtained (r ≥ 0.9999 over 3 t1/2 ). For the reaction of imidazole
with 4-nitrophenyl trichloroacetate strict first-order kinetics
were not observed. The rate coefficients were calculated by the
plots of ln(A∞ Ϫ At) versus time. The correlation coefficients of
the plots were only satisfactory (r = 0.996–0.999) and the stand-
ard deviations of the calculated rate coefficients were 1.5–3%.
However, from previous studies it is known that the reactions of
imidazole with 4-nitrophenyl acetate, chloroacetate, dichloro-
acetate and trifluoroacetate in acetonitrile follow first-order
kinetics.8a,b The reason for this difference is under investigation.
The rate coefficients given in Tables 1 and 2 are mean values of
three or four determinations and the error given is a standard
deviation.
CH3COOC6H4X
3
X = 4-NO2, 4-COMe, 4-CN, 4-CO2Et, 4-F, 4-Br, 4-Cl, H, 4-Me, 4-OMe, 4-NH2,
3-NO2, 3-CN, 3-CO2Et, 3-F, 3-Cl, 3-Br, 3-Me, 3-COMe or 3-OMe
XC6H4COOCH3
4
X = 4-NO2, 4-Br, H, 4-Me, 4-OMe, 3-NO2, 3-Br, 3-Me, 3-OMe or 3-N(Me)2
Experimental
Materials
The preparation and purification of the 4-nitrophenyl esters of
acetic,8a chloroacetic8b and dichloroacetic8b acids as well as
those of 3-nitro-, 4-chloro- and 4-methoxyphenyl dichloro-
acetates8c were as described in previous works. The other esters
of dichloroacetic acid were prepared from dichloroacetyl chlor-
ide and the appropriate phenol by the dropwise addition of an
equivalent amount of pyridine to a stirred mixture of dichloro-
acetyl chloride and the appropriate phenol in dry diethyl ether
at room temperature.9 The esters were recrystallized from
hexane. The melting points of the compounds were as follows:
4-methylphenyl dichloroacetate 53.5–54.5 ЊC, phenyl dichloro-
acetate 46–47 ЊC (lit.10 45–47 ЊC) and 4-bromophenyl dichloro-
acetate 72–73 ЊC. 4-Cyanophenyl dichloroacetate was obtained
as an oil. 3-Chlorophenyl dichloroacetate was distilled under
reduced pressure 140–141 ЊC/13 mmHg. 4-Nitrophenyl tri-
chloroacetate was prepared from 4-nitrophenol and trichloro-
acetyl chloride in dry benzene9,11 and it was recrystallized twice
from a diethyl ether–petroleum ether (1:2 v/v) mixture, mp 63–
64 ЊC (lit.11 64 ЊC). Acetonitrile (Aldrich, spectrograde) used in
the preparation of the reaction solutions and the stock solu-
tions of the esters as well as deuteriochloroform (Sigma, 99.8
atom%) employed in NMR measurements were used as
received.
Spectroscopic measurements
Results and discussion
The 13C NMR spectra were recorded at 125.78 MHz with a 45Њ
pulse and a digital resolution of 0.92 Hz on a JEOL JNM-A500
spectrometer in CDCl3 at 27 ЊC at a concentration of 0.1 mol
dmϪ3. The deuterium of the solvent was used as a lock signal.
Tables 1 and 2 give the carboxylate carbon 13C NMR chemical
shifts, ν(C᎐O) frequencies and the kinetic data, for series 1 and
᎐
2, respectively. The appearance of the ν(C᎐O) as a doublet
᎐
1498
J. Chem. Soc., Perkin Trans. 2, 1999, 1497–1502

View Article Online
Table 2 Carbonyl carbon 13C NMR chemical shifts, IR C᎐O stretching frequencies and rate coefficients for the nucleophilic reaction of imidazole
᎐
for acyl substituted 4-nitrophenyl acetates RCOOC6H4-NO2 in acetonitrile at 298.2 K
δ(C᎐O)a/
ppm
ν(C᎐O)/
k1/10Ϫ1 dm3
k2/dm6
᎐
᎐
R
cmϪ1
molϪ1 sϪ1
molϪ2 sϪ1
CH3
168.38
165.05
162.13
159.73
155.1d
1761
1772
0.0185b 0.0007
1.54c 0.11
1.19c 0.46
1.0 2.0
0.0143b 0.0004
7.66c 0.16
544c
CH2Cl
CHCl2
CCl3
1767, 1788
1782
7
250 30
CF3
1806e
6.3b 1.2
3360b 70
a In CDCl3. b Ref. 8a. c Ref. 8b. d Ref. 13a. e Ref. 13b.
Table 3 Correlations and cross-correlations of the 13C NMR chemical shifts (δCO) or SCS values of the carbonyl carbons and for the logarithms of
the rate coefficients of the neutral (ko) or alkaline (kOH) hydrolysis or of the rate coefficients of the nucleophilic reaction of imidazole (kIm) for
compounds of series 1–4 and 15
Line
Type of correlation
Slope
sa
rb
nc
1
2
3
4
5
6
7
8
9
log ko(series 1)d vs. σ
1.65
Ϫ1.46
Ϫ1.53
Ϫ0.76
Ϫ1.22
Ϫ0.79
0.78
0.10
0.14
0.07
0.06
0.11
0.07
0.01
0.04
0.04
0.9864
0.9693
0.9897
0.9701
0.9688
0.9755
0.9994
0.9990
0.9998
9
9
13
14
10
9
8
4
4
log ko(series 1)d vs. SCS(series 1)e
log kIm(series 3)f vs. SCS(series 3)g
log kOH(series 3)h vs. SCS(series 3)g
log kOH(series 3)i vs. SCS(series 3)g
log kOH(series 4)j vs. SCS(series 4)k
SCS(series 1)e vs. SCS(series 3)g
δCO (series 2)l vs. δCO (series 15)m
δCO (series 2)l vs. σ*
1.27
Ϫ3.27
a Standard deviation. b Correlation coefficient. c Number of correlation points. d In 20% acetonitrile–water at 298.2 K (Table 1). e In CDCl3 (Table 1).
f In aqueous solution at 298.2 K; the values are from ref. 21. g In CDCl3; the values are from ref. 3c. h In 3% ethanol–water at 298.2 K; the values are
from ref. 19b. i In 56% acetonitrile–water at 274.2 K; the values are from ref. 19a. j In 85% methanol–water at 298.2 K; the values are from ref. 22. k In
CDCl3; the values are from ref. 23. l In CDCl3 (Table 2). m In CD2Cl2; the values are from ref. 25.
absorption for 4-nitrophenyl dichloroacetate is thought to be
due to a conformational equilibrium,14 although the possibility
of a Fermi resonance cannot be excluded. The IR spectrum for
that compound was recorded as a capillary film.
discussed elsewhere.3c,d,f,4 There are also the early theoretical
calculations by Pople and Gordon16 which predicted that
unsaturated bonds tend to polarize by the introduction of a
polar substituent leading to alternation of charge.
An alternative explanation states that the non-linear
Hammett correlations of the carbon shifts are due to the
variation of the average excitation energy (∆E) term in the
Karplus–Pople equation.17 The latter describes in a simplified
form the paramagnetic term of the shielding constant of a
nucleus. However, the assumption that ∆E is not constant in a
substituent series has not been proved. In any case, the pro-
posed ∆E approach17 has not been able to explain the system-
atic reverse 13C NMR shift correlations [negative ρ values in
eqn. (3)] observed in numerous cases.3,4,5,7,18
On the C᎐O shifts
᎐
Electron-withdrawing substituents in both the acyl and aryl
groups of the esters studied cause upfield shifts of the C᎐O
carbon signal suggesting increased shielding (Tables 1 and 2).
Analogous “reverse” substituent chemical shift (SCS) effects
᎐
have been detected previously for several unsaturated carbons
᎐
of aromatic side chains possessing a C᎐O, C᎐C, C᎐C, C᎐N or
᎐
᎐
᎐
᎐
3,4a,5–7
᎐
C᎐N system.
The excellent correlations obtained by
᎐
eqn. (3) indicate the electronic origin of these relationships.3–5,15
SCS = ρIσI ϩ ρRσR
(3)
Reactivity and the C᎐O carbon shifts
᎐
As expected,1 electron-withdrawing substituents in the acyl or
aryl moiety of the ester increase the rate of a nucleophilic acyl
substitution of substituted phenyl acetates as demonstrated by
neutral hydrolysis or nucleophilic reaction of imidazole (Tables
1 and 2). A positive Hammett ρ value of 1.65 is observed for the
neutral hydrolysis of series 1 (Table 3, line 1). The effect of
varying the leaving group on the neutral hydrolysis of aryl
esters has not been widely studied. However, with the aid of
the rate coefficients of 4-nitro- and 4-methoxyphenyl dichloro-
acetates Fife and McMahon have previously calculated the ρ
value of 1.3 in aqueous solution at 303 K.9b For the neutral
hydrolysis of substituted phenyl trifluoroacetates in 3.89 M
water in acetonitrile we have previously observed the ρ value of
2.45.13a For alkaline hydrolysis of substituted phenyl acetates
ρ values from 1.0 to 1.8 in Hammett or Hammett type equa-
tions in water or in different acetone–water mixtures have been
reported19a,20 and for the nucleophilic reaction of imidazole
with substituted phenyl acetates in aqueous solution the ρ value
of 1.76 has been observed.21
The DSP equation (3) dissects the substituent effects into
inductive (ρIσI) and resonance (ρRσR) contributions. SCS means
the 13C chemical shift (in ppm) for a substituted compound
relative to the unsubstituted one. The reverse trend of SCS
effects (ρI, ρR < 0) has been explained by the π-polarization
mechanism: the substituent dipole in the aromatic ring is
thought to polarize each π-unit as a localized system as shown
in 5 for substituted phenyl acetates.3c,d The polarization of the
δ+
O
δ+ δ–
H3C
C
δ–
O
X
5
C᎐O group in the side chain by an electron-withdrawing sub-
᎐
stituent then results in an increase in the π-electron density and
the upfield shift. The chemical shift dependence on electron
density is normal but the substituent effect on the electron
density is reversed. The π-polarization has been extensively
Fig.
1 shows the good correlation between the rate
coefficients of the neutral hydrolysis of substituted phenyl
J. Chem. Soc., Perkin Trans. 2, 1999, 1497–1502
1499

View Article Online
reactivity, the 13C chemical shifts and ν(C᎐O) frequencies is
᎐
discussed below.
An explanation for the reactivity versus C᎐O shift and í(C᎐O)
᎐
᎐
frequency behaviour: ester ground state destabilization
The cross-correlation between the C᎐O carbon shifts of phenyl
᎐
substituted phenyl dichloroacetates and phenyl acetates is good
(Table 3, line 7). The slope 0.78 reflects the lower susceptibility
of the C᎐O bond polarization to para-substituents in the case
᎐
of phenyl dichloroacetates and indicates the higher C᎐O double
᎐
bond character of phenyl dichloroacetates as compared to
phenyl acetates. This concept is confirmed by the increasing
carbonyl stretching frequencies with increasing electron-
withdrawal by acyl substituents in a series of substituted 4-
nitrophenyl acetates (Table 2). In a series of acyl-halogenated
ethyl acetates, ν(C᎐O) also gradually shifts towards higher
᎐
frequencies with increasing electron withdrawal. This has been
explained by the electron-withdrawing effect of the halogen
Fig. 1 A plot of the logarithms of the rate coefficients for the neutral
hydrolysis of phenyl substituted phenyl dichloroacetates in 20% (v/v)
acetonitrile–water versus the substituent chemical shift values (SCS
means the 13C NMR chemical shift for a substituted compound relative
to the unsubstituted compound) for the carboxylate carbon of the
esters in CDCl3. The values are from this work.
substituents which suppress the polarization of the C᎐O
᎐
bond.14a In a series of phenyl substituted phenyl dichloro-
acetates (Table 1) or phenyl trifluoroacetates,13a the C᎐O
᎐
stretching frequency also shifts towards higher wavenumbers
with increasing electron-withdrawal, indicating an increase in
the strength of the C᎐O bond. Satisfactory linear correlations
᎐
dichloroacetates and the substituent chemical shift (SCS) values
between the carbonyl stretching frequency and substitution
(νCO versus σ or σ*) have been reported in some cases20,24
although the significance of the correlation has hardly been
discussed. These results can be understood by considering the
ester resonance structures (6–8).
for the C᎐O carbon of the esters. A negative slope of Ϫ1.5 is
᎐
observed (Table 3, line 2). The upfield shift of the carbon reson-
ance and an increase in the reaction rate are accompanied by
an increase in the electron-withdrawing power of the phenyl
substituent. The upfield shift indicates an increased π-electron
shielding of the carboxylate carbon. This contradicts the usual
concept of the increased electrophilicity, i.e., decreased electron
density, of the carboxylate carbon as a reason for the increase
of the rate of nucleophilic acyl substitutions by electron-
withdrawing substituents.1 An analogous behaviour is seen for
the rate coefficients of the nucleophilic reaction of imidazole
with a series of substituted phenyl acetates in aqueous solution
and the SCS values of the carboxylate carbons of the substrate
esters in CDCl3 (Table 3, line 3). Negative slopes are also
observed for the respective correlations for phenyl substituted
phenyl acetates (Table 3, lines 4 and 5) and for benzoyl substi-
tuted methyl benzoates (Table 3, line 6), the correlation being in
both cases between the rate coefficient of the alkaline hydrolysis
O–
C
O–
C
O
R1
C
OR2
R1
OR2
R1
OR2
6
7
8
The electron-withdrawing substituents in the acyl (R1) and
aryl (R2) groups of the ester destabilize the resonance form 7
and those in the aryl group also destabilize the resonance form
8 and thereby increase the relative contribution of the structure
6. The increasing importance of structure 6 is evidenced by the
increasing shielding, i.e., by the upfield 13C NMR shifts, of the
carbonyl carbon and by the higher ν(C᎐O) frequencies. Accord-
᎐
ingly, in agreement with the 13C NMR and IR results the
increase in the reaction rate of nucleophilic acyl substitutions
by electron-withdrawing substituents, which means positive
reaction constants ρ or ρ* in Hammett or Taft equations,
respectively, can be explained by the ground state destabiliz-
ation of the substrate. This ground state destabilization is due
to the diminished overall resonance stabilization. So, reaction
rates increase despite the decrease in the electrophilicity of the
acyl carbon. The above interpretation is in accordance with the
13C NMR spectroscopic study of Dell’Erba et al.7b The com-
parison of the sensitivity of the carbonyl carbon resonance to
the benzoyl substituent in four series of the nitro-substituted
phenyl benzoates with methyl benzoates (13) and aceto-
phenones (14) showed along the 9–12 series a gradual shift from
a 13-like to a 14-like behaviour. Therefore, increasing electron-
withdrawal by the aryl group was proposed to decrease the
significance of the normal ester resonance.
and the SCS values of the ester C᎐O carbon.
᎐
The reactions for which rate coefficients and SCS values are
correlated (Table 3, lines 2–6) represent different types of
reaction mechanisms and therefore different transition states:
rate-limiting formation of a tetrahedral addition intermediate
(alkaline hydrolysis),1a,b rate-limiting breakdown of the addi-
tion intermediate (reaction of imidazole with phenyl acetates in
an aprotic solvent)8a or rate-limiting formation of the addition
intermediate with the C–O bond breaking possibly contribut-
ing to the rate of the reaction (neutral hydrolysis).13 The rate
coefficients and the 13C chemical shifts in Table 2 show that also
with acyl substituted 4-nitrophenyl acetates (series 2) the
increasing electron-withdrawal is accompanied both with an
increase in the rate of the nucleophilic reaction of imidazole (k1
refers to the uncatalysed nucleophilic reaction of imidazole and
k2 to the general base catalysed nucleophilic reaction of imid-
azole8a,b) and with an upfield shift of the carboxylate carbon
13C NMR chemical shift. Linear correlations between the rate
coefficient and the C᎐O shift are, however, not observed in that
Analogy with acyl chlorides. There are only a few relevant
᎐
case, obviously because of steric effects.
There is little discussion about the correlation of the reaction
studies concerning aliphatic substituent effects on the C᎐O
᎐
carbon 13C NMR chemical shifts. However, the chemical shift
behaviour of chloro substituted acetyl chlorides (series 15) is
known:25 the 13C chemical shifts of the acyl carbons of acetyl,
chloroacetyl, dichloroacetyl and trichloroacetyl chlorides in
CD2Cl2 are 170.6, 168.1, 165.9 and 163.7 ppm, respectively.
Further, the complexation behaviour of these compounds with
AlCl3, and the efficiency of these complexes in an aromatic
rates with the C᎐O carbon shifts in the literature, and the
᎐
relevance of the reverse behaviour has not been discussed at
all.6 π-Polarization can explain the reverse behaviour of the
C᎐O shifts, but what explains the increase in the reaction rates?
᎐
Without neglecting the contribution of the stability of the tran-
sition state, an explanation for the substituent dependence of
1500
J. Chem. Soc., Perkin Trans. 2, 1999, 1497–1502

View Article Online
O
C
quantum mechanical gas phase calculations performed by
Kallies and Mitzner27 show that the lower reactivity of N,NЈ-
dimethylurea (DMU) as compared to N-methyl acetamide
(MAM) reflects the higher overall resonance stabilization of the
former compound despite the fact that the lone pair of one
nitrogen is less delocalized with DMU than MAM.
X
OAr
9
Ar =
10 Ar
=
NO2
Estimation of rate coefficients with the aid of C᎐O shifts
᎐
The utilization of the reactivity versus carbonyl carbon 13C
chemical shift correlation to predict rate coefficients was tested
for the neutral hydrolysis of substituted phenyl dichloro-
acetates. With the equation log ko = (Ϫ1.5 0.1) × SCS Ϫ
(3.2 0.1), obtained from the phenyl dichloroacetates other
than the 4-bromo substituted example (Table 1), the calculated
value (1.9 0.7) × 10Ϫ3 sϪ1 was obtained for ko of 4-bromo-
phenyl dichloroacetate in 20% acetonitrile–water at 298.2 K.
This value is a satisfactory estimate of the experimental value
1.3 × 10Ϫ3 sϪ1 (Table 1). The use of log k versus δ (C᎐O) plots is
a convenient way to obtain at least a first approximation for a
rate coefficient. This can be especially important for substituted
derivatives for which the substituent constants are not known
and therefore any Hammett type correlations cannot be used
for the evaluation of the rate coefficient.
O2N
O2N
O2N
11 Ar
=
12 Ar =
NO2
NO2
O
O
C
᎐
C
R
C
Cl
R
Cl
15
16
R = CH3, CH2Cl, CHCl2, CCl3
acylation, have been studied.25 During the interaction of acyl
chlorides with AlCl , the C᎐O bond is weakened and the
᎐
3
Conclusions
carbonyl carbon chemical shifts are displaced downfield while
the carbonyl oxygen 17O chemical shifts are displaced upfield.
Interestingly, acetyl chloride is much more reactive than tri-
chloroacetyl chloride in the AlCl3 promoted acylation.25 The
complexation of trichloroacetyl chloride with AlCl3 is much
weaker than that of acetyl chloride and therefore a much
smaller change as a consequence of the complexation was
observed in the 13C and 17O chemical shifts of trichloroacetyl
chloride as compared to acetyl chloride. This, together with the
Electron-withdrawing substituents increase the rates of nucleo-
philic acyl substitutions of esters. The electron-withdrawal by
the substituents also results in an upfield shift of the 13C NMR
resonance of the C᎐O carbon and a higher wavenumber of the
᎐
ν(C᎐O) frequency. The upfield chemical shifts and the higher
᎐
wavenumbers of the C᎐O stretching reflect the increased
double bond character of the C᎐O bond. As for the rate
᎐
᎐
increase in the nucleophilic acyl substitutions, the results sug-
gest a significant decrease in the ester ground state stability due
to the decreased resonance stabilization, in contrast to the con-
ventional concept according to which electron-withdrawing
substituents increase the reaction rate by the increased electro-
philicity of the C᎐O carbon and by the stabilization of the
negative charge developed at the transition state. As a first
approximation, the use of log k versus δ (C᎐O) correlations is
an alternative way to evaluate rate coefficients especially for
substituted derivatives for which the substituent constants are
not known and therefore Hammett type correlations cannot be
used.
C᎐O 13C chemical shift behaviour in the series of chloro-
᎐
substituted acetyl chlorides, indicates the more fixed C᎐O
᎐
double bond character in the case of trichloroacetyl chloride as
compared to acetyl chloride. According to our postulation the
increased double bond character can be attributed to the
decreased resonance contribution of form 16 relative to 15 due
to the electron-withdrawal by acyl substituents. In the Friedel–
Crafts acylation this effect results in competition. While the
chloro substituents make the acyl chloride more prone to a
nucleophilic attack, they make it less susceptible to complex-
᎐
᎐
C
ation with AlCl because of diminished C᎐O bond polarization
᎐
3
and diminished basicity of the carbonyl oxygen toward a Lewis
acid. Consequently, the chloro substituents decrease the total
reactivity.
References
1 (a) A. J. Kirby, in Comprehensive Chemical Kinetics, eds. C. H.
Bamford and C. F. H. Tipper, Elsevier Publishing Company,
Amsterdam, 1980, vol. 10, chapter 2, p. 161; (b) K. A. Connors,
Structure Reactivity Relationships: the Study of Reaction Rates in
Solution, VCH Publishers, New York, 1990, chapter 7, p. 311; (c)
G. W. Klumpp, Reactivity in Organic Chemistry, Wiley, New York,
1982, chapter 3, p. 103; (d) O. Exner, in Advances in Linear Free
Energy Relationships, Plenum Press, London, 1972, chapter 1, p. 1.
2 K. A. Connors, Structure Reactivity Relationships: the Study of
Reaction Rates in Solution, VCH Publishers, New York, 1990,
chapter 5, p. 187.
3 (a) J. Bromilow, R. T. C. Brownlee and D. J. Craik, Aust. J. Chem.,
1977, 30, 351; (b) R. T. C. Brownlee and D. J. Craik, Org. Magn.
Reson., 1981, 15, 248; (c) J. Bromilow, R. T. C. Brownlee, D. J.
Craik, P. R. Fiske, J. E. Rowe and M. Sadek, J. Chem. Soc., Perkin
Trans. 2, 1981, 753; (d) D. J. Craik and R. T. C. Brownlee, Prog.
Phys. Org. Chem., 1983, 14, 1 and references therein; (e) R. T. C.
Brownlee, M. Sadek and D. J. Craik, Org. Magn. Reson., 1983, 21,
616; ( f ) W. F. Reynolds, Prog. Phys. Org. Chem., 1983, 14, 165.
4 (a) G. K. Hamer, I. R. Peat and W. F. Reynolds, Can. J. Chem.,
1973, 51, 897, 915; (b) W. J. Hehre, R. W. Taft and R. D. Topsom,
Prog. Phys. Org. Chem., 1976, 12, 159.
Interestingly, an excellent correlation is seen (r = 0.999)
between the C᎐O carbon shifts of acyl substituted 4-nitro-
᎐
phenyl acetates and acetyl chlorides (Table 3, line 8). Further,
the C᎐O carbon 13C chemical shifts of 4-nitrophenyl acetates
᎐
versus the Taft’s polar σ* parameters show an excellent correl-
ation when the clearly deviating point of the trifluoroacetyl
group is excluded (ρ* = Ϫ3.27, Table 3, line 9). This correlation
also shows that in this aliphatic substitution series electronic
effects dominate the substituent chemical shift effects. Further-
more, the negative slope suggests that the electron-withdrawing
acyl substituents suppress the ester resonance and increase the
carbonyl carbon shielding. We are not aware of similar previous
correlations. At the present time it is not known why the CF3
group does not fit the correlation line.
Other evidence. The concept of a noticeable contribution of
the ground state resonance stabilization to the dependence of
reaction rates on the substrate structure in nucleophilic acyl
substitution is supported by the high values of the localization
energies (14–18 kcal molϪ1 for esters) estimated for addition
of nucleophiles to acyl derivatives.26 Furthermore, the recent
5 (a) K. Neuvonen, F. Fülp, H. Neuvonen and K. Pihlaja, J. Org.
Chem., 1994, 59, 5895; (b) K. Neuvonen, F. Fülöp, H. Neuvonen,
M. Simeonov and K. Pihlaja, J. Phys. Org. Chem., 1997, 10, 55.
J. Chem. Soc., Perkin Trans. 2, 1999, 1497–1502
1501

View Article Online
6 C. J. O’Connor and T. D. Lomax, Aust. J. Chem., 1983, 36, 917;
C. J. O‘Connor, D. J. McLennan, D. J. Calvert, T. D. Lomax,
A. J. Porter and D. A. Rogers, Aust. J. Chem., 1984, 37, 497.
7 (a) C. Dell’Erba, F. Sancassan, M. Novi, G. Petrillo, A. Mugnoli,
D. Spinelli, G. Consiglio and P. Gatti, J. Org. Chem., 1988, 53, 3564;
(b) C. Dell’Erba, F. Sancassan, G. Leandri, M. Novi, G. Petrillo and
A. Mele, Gazz. Chim. Ital., 1989, 119, 643; (c) C. Dell’Erba,
A. Mele, M. Novi, G. Petrillo, F. Sancassan and D. Spinelli,
J. Chem. Soc., Perkin Trans. 2, 1990, 2055.
8 (a) H. Neuvonen, J. Chem. Soc., Perkin Trans. 2, 1987, 159; (b)
H. Neuvonen, J. Chem. Soc., Perkin Trans. 2, 1995, 951; (c)
H. Neuvonen, Ber. Bunsenges. Phys. Chem., 1996, 100, 1231.
9 (a) T. H. Fife, J. Am. Chem. Soc., 1965, 87, 4597; (b) T. H. Fife and
D. M. McMahon, J. Am. Chem. Soc., 1969, 91, 7481.
10 K. Koehler, R. Skora and E. H. Cordes, J. Am. Chem. Soc., 1966,
88, 3577.
11 R. Buyle, Helv. Chim. Acta, 1964, 47, 2449.
12 A. A. Frost and R. G. Pearson, Kinetics and Mechanisms, Wiley,
New York, 1961, p. 27–55.
13 (a) H. Neuvonen, J. Chem. Soc., Perkin Trans. 2, 1986, 1141; (b)
H. Neuvonen, Dissertation, University of Turku, Turku, 1985.
14 (a) M. Avram and GH. D. Mateescu, Infrared Spectroscopy:
Applications in Organic Chemistry, Wiley, New York, 1972, chapter
8, p. 416; (b) Y. Mido and M. Hashimoto, J. Mol. Struct., 1985, 131,
71.
15 D. F. Ewing, in Correlation Analysis in Chemistry: Recent Advances,
eds. N. B. Chapman and J. Shorter, Plenum Press, New York, 1978,
p. 357.
16 J. A. Pople and M. Gordon, J. Am. Chem. Soc., 1967, 89, 4253.
17 H. Dahn and P.-A. Carrupt, Magn. Reson. Chem., 1997, 35, 577.
18 R. Noto, L. Lamartina, C. Arnone and D. Spinelli, J. Chem. Soc.,
Perkin Trans. 2, 1987, 689; 1988, 887.
19 (a) J. J. Ryan and A. A. Humffray, J. Chem. Soc. B, 1966, 842;
(b) T. Nishioka, T. Fujita, K. Kitamura and M. Nakajima, J. Org.
Chem., 1975, 40, 2520.
20 L. A. Cohen and S. Takahashi, J. Am. Chem. Soc., 1973, 95, 443.
21 A. Williams and R. A. Naylor, J. Chem. Soc. B, 1971, 1967.
22 A. D. Campbell, S. Y. Chool, L. W. Deady and R. A. Shanks, Aust.
J. Chem., 1970, 23, 203.
23 M. Budesinsky and O. Exner, Magn. Reson. Chem., 1989, 27, 585.
ˇˇ
24 R. D. Topsom, Prog. Phys. Org. Chem., 1987, 16, 193; S. Takahashi,
L. A. Cohen, H. K. Miller and E. G. Peake, J. Org. Chem., 1971, 36,
1205; A. S. Shawali and N. F. Eweiss, Can. J. Chem.,1977, 55, 3967.
25 F. Bigi, G. Casnati, G. Sartori and G. Predieri, J. Chem. Soc., Perkin
Trans. 2, 1991, 1319.
26 J. Fastrez, J. Am. Chem. Soc., 1977, 99, 7004.
27 B. Kallies and R. Mitzner, J. Chem. Soc., Perkin Trans. 2, 1996, 397.
Paper 9/00189A
1502
J. Chem. Soc., Perkin Trans. 2, 1999, 1497–1502