Kinetics and mechanism of benzylamine additions to ethyl α-acetyl-β-phenylacrylates in acetonitrile

Article abstract of DOI:10.1039/b401239a

Kinetic studies of the addition of benzylamines to a noncyclic dicarbonyl group activated olefin, ethyl α-acetyl-β-phenylacrylate (EAP), in acetonitrile at 25.0°C are reported. The rates are lower than those for the cyclic dicarbonyl group activated olefi

Full text of DOI:10.1039/b401239a

View Article Online / Journal Homepage / Table of Contents for this issue
Kinetics and mechanism of benzylamine additions to ethyl
ꢀ-acetyl-ꢁ-phenylacrylates in acetonitrile
Hyuck Keun Oh,a In Kon Kim,a Dae Dong Sungb and Ikchoon Lee*c
a Department of Chemistry, Chonbuk National University, Chonju, 560-756, Korea
b Department of Chemistry, Dong-A University, Busan, 604-714, Korea
c Department of Chemistry, Inha University, Inchon, 402-751, Korea.
E-mail: ilee@inha.ac.kr; Fax: ϩ82-32-865-4855
Received 26th January 2004, Accepted 24th February 2004
First published as an Advance Article on the web 15th March 2004
Kinetic studies of the addition of benzylamines to a noncyclic dicarbonyl group activated olefin, ethyl α-acetyl-β-
phenylacrylate (EAP), in acetonitrile at 25.0 ЊC are reported. The rates are lower than those for the cyclic dicarbonyl
group activated olefins. The addition occurs in a single step with concurrent formation of the Cα–N and Cβ–H bonds
through a four-center hydrogen bonded transition state.
The kinetic isotope effects (kH/kD > 1.0) measured with deuterated benzylamines (XC6H4CH2ND2) increase with a
stronger electron acceptor substituent (δσX > 0) which is the same trend as those found for other dicarbonyl group
activated series ( 1–4), but is in contrast to those for other ( noncarbonyl) group activated series (5–9). For the
dicarbonyl series, the reactivity-selectivity principle (RSP) holds, but for others the anti-RSP applies. These are
interpreted to indicate an insignificant imbalance for the former, but substantial lag in the resonance delocalization
in the transition state for the latter series.
for the reactions with substantial imbalances (vide infra). Since
Introduction
we noted that the dicarbonyl activated olefins,1e–g Z, ZЈ = (CO)2 ؒ
Nucleophilic addition of amines (XRNH2) to olefins (YC6H4-
R1R2, especially with the cyclic structure, 11e and 2,1f belong
CH᎐CZZЈ) activated by electron-withdrawing groups (Z, ZЈ)
᎐
to such a class with smaller kH/kD values and insignificant
imbalances, it is of much interest to verify that the cyclic di-
carbonyl structure of the activating group is a prerequisite to
the negligible imbalance in the amine additions in acetonitrile.
Although we found the same behavior with a noncyclic di-
carbonyl activated olefin, 31g we test further in this work with
another acyclic dicarbonyl activated group, 4, ethyl α-acetyl-β-
phenylacrylates. The purpose of this work is to examine (i)
whether such a negligible imbalance is limited to the cyclic
dicarbonyl activated olefins ( 1 and 2) or not, and (ii) why do
dicarbonyl, or cyclic dicarbonyl, activated olefins1e–g behave
differently from other activated olefins?1a–d
is known to proceed in acetonitrile by concerted formation of
the Cα–N and Cβ–H bonds in a single-step process to a neutral
product,1 eqn. 1.
(1)
In contrast, the reactions in aqueous solution are reported
to occur through a zwitterionic intermediate, T , with an
imbalanced transition state (TS) in which the development of
resonance into the activating groups (Z, ZЈ) lags behind Cα–N
bond formation.2 The imbalance in the TS is mainly caused
by the poorly developed resonance into Z, ZЈ and solvation
with the negative charge largely localized on carbon (Cβ), an
exaggerated form of which can be given as I. The imbalance
phenomenon was pronounced in the amine additions in
aqueous solution, which was expressed semi-quantitatively by
using structure–reactivity coefficients such as Brønsted α and
β values.2 In contrast, the imbalance was hardly observable in
the amine addition reactions in acetonitrile, mainly due to the
concerted formations of Cα–N and Cβ–H bonds.1
Results and discussion
The reactions studied in the present work followed a simple
second-order rate law given by eqns 2 and 3
Recently, however, we have found that a stronger localized
anionic charge on Cβ due to the lag in the development of
resonance induces a larger kinetic isotope effect involving
deuterated amines (XRND2), kH/kD > 2.3, which is caused by a
larger N–H bond stretching (II) in the concerted single step
addition in acetonitrile.1a–d In the absence of, or insignificantly
Ϫ d[EAP]/dt = kobs [EAP]
kobs = k2 [BA]
(2)
(3)
small, imbalance, the isotope effects were smaller (kH/kD
<
where [EAP] and [BA] are the concentrations of substrate, 4,
and benzylamine, respectively. No catalysis by a second amine
molecule was detected. The second-order rate constants, k2,
2.0)1e–g and the trends of change in kH/kD with substituents X in
the nucleophile and Y in the ring were exactly opposite to those
T h i s j o u r n a l i s © T h e R o y a l S o c i e t y o f C h e m i s t r y 2 0 0 4
O r g . B i o m o l . C h e m . , 2 0 0 4 , 2, 1 2 1 3 – 1 2 1 6
1213

View Article Online
Table 1 The second order rate constants, k2 × 102 dm3 molϪ1 sϪ1 for the addition reactions of ethyl α-acetyl-β-phenylacrylates with X-benzylamines
in acetonitrile at 25.0 ЊC
Y
a
X
p-OMe
p-Me
H
p-Cl
p-Br
ρY
0.767
0.559b
0.408c
0.612
2.38
1.69
1.18
1.77
p-OMe
1.02
1.47
2.74
1.03 0.06
p-Me
H
0.825
0.581
1.14
0.815
1.99
1.33
0.95 0.05
0.90 0.05
0.439
1.22
0.295
0.212
0.151
Ϫ0.83
( 0.03)
0.79
0.749
0.554
0.404
Ϫ0.99
( 0.05)
0.95
p-Cl
0.372
0.496
0.807
0.83 0.03
d
ρX
Ϫ0.89
( 0.01)
0.84
Ϫ0.94
( 0.02)
0.90
Ϫ1.05
( 0.06)
1.01
ρXY e = Ϫ 0.38 ( 0.13)
f
βX
( 0.03)
( 0.02)
( 0.01)
( 0.02)
( 0.03)
a The σ values were taken from ref. 11a. Correlation coefficients were better than 0.995 in all cases. b At 15.0 ЊC. c At 5.0 ЊC. d The source of σ is the
same as for footnote a. Correlation coefficients were better than 0.997 in all cases. e Correlation coefficients was 0.997. f The pKa values were taken
from ref. 11b. Correlation coefficients were better than 0.999 in all cases. pKa = 9.67 was used for X = p-CH3O. (ref. 11c).
Table 2 Comparison of rates (k2) and other parameters for the benzylamine additions to 1–4 in acetonitrile
k2/ЊCa
∆H≠ b
Ϫ∆S≠ c
kH/kD
βX
ρXY
Ref.
d
e
BMA (1)
BID (2)
BDM (3)
EAP (4)
86.6(20)
∼ 4.0
∼ 6.2
∼ 6.4
∼ 4.9
∼Ϫ37
∼Ϫ37
∼Ϫ48
∼Ϫ52
1.45 ∼ 1.74
1.25 ∼ 1.81
1.52 ∼ 2.07
1.78 ∼ 2.38
0.23
1.09
0.88
0.90
Ϫ0.33
Ϫ0.33
Ϫ0.45
Ϫ0.38
1e
1f
1g
1.48(25)
2.48 × 10Ϫ2(20)
0.815 × 10Ϫ2(25)
This work
a For X = Y = H: MϪ1 sϪ1 b kcal molϪ1 c cal molϪ1 KϪ1 d With deuterated benzylamine (XC6H4CH2ND2). e For Y = H.
. . .
were obtained from the slopes of the linear plots of kobs vs.
[BA], eqn 3, and are summarized in Table 1. The Hammett
ρX and ρY values and Brønsted βX values are also shown in
Table 1 together with the cross-interaction constant ρXY, which
is defined as eqns 4 and 5.3
forming a four-membered type (II), or alternatively hydrogen
bonding to a carbonyl oxygen forming a six-membered type,
(III), structure, albeit the latter possibility has been shown to be
less likely1g (vide infra). The kH/kD values determined in this
work are collected in Table 3. We note that the kH/kD values are
smaller (1.78 ∼ 2.24) for a stronger nucleophile (δσX < 0) than
for a weaker nucleophile (1.82 ∼ 2.38) i.e., δσX < 0
δ (kH/kD)
log (kXY/kHH) = ρXσX ϩ ρYσY ϩ ρXYσXσY
ρXY = ∂ρX/ ∂σY = ∂ρY/ ∂σX
(4)
(5)
< 0. Admittedly the differences are marginal but the same trend
is clear Exactly the same tend was found for the benzylamine
additions to other dicarbonyl activated series in acetonitrile,
irrespective of whether the dicarbonyl group has a ring
structure (BMA (1) and BID (2))1e,f or not (BDM (3)1g and
EAP (4)).
The βX values were determined by the plots of log k2 (MeCN)
against pKa (H2O) of benzylamines. This procedure was found
to be reliable, since the pKa (MeCN) varies in parallel with the
pKa (H2O) with a reasonably constant difference of ∆pKa (= pKa
(MeCN) Ϫ pKa (H2O)) ≅ 7.5.4 The rates and other relevant
parameters for the benzylamine additions to the dicarbonyl
activated benzylidene series of substrate (1–4) in acetonitrile are
compared in Table 2. We note that the rates are significantly
faster for the cyclic dicarbonyl activated substrates (1 and 2)
than the noncyclic dicarbonyl activated ones (3 and 4). This is
clearly the π-overlap effect of the ring structure which alleviates
the rate lowering due to the lag in the resonance stabilization.5
In noncyclic systems the resonance develops late along the
reaction coordinate but in the ring system the π-overlap is
already maximally built into the olefin so that the structural
reorganization that may be needed to achieve this overlap in
noncyclic systems is not required and hence the activation
barrier is lowered. The sign of ρXY is negative in all cases, which
is consistent with those reported for nucleophilic bond form-
ation processes.3 The magnitudes of ρXY for the cyclic series
(Ϫ0.33) are somewhat smaller than those for the noncyclic
series (Ϫ0.45 and Ϫ0.38), which could be an indication of an
earlier TS for the cyclic compounds. This is supported by
smaller kH/kD values for 1 and 2 than for 3 and 4. The kinetic
isotope effect, kH/kD, involving deuterated nucleophiles
(XC6H4CH2ND2) reflects bond stretching of the N–H (N–D)
bond of the amine in the TS due to the hydrogen bonding of
the amine proton toward the anionic center developing on Cβ
In nucleophilic substitution reactions, the two product
stabilizing factors are (i) a stronger nucleophile (δσX < 0) and
(ii) a better leaving group (δσZ > 0), for which the TS shift
is predicted based on thermodynamic models,6 such as the
Hammond postulate,6a the Bell–Evans–Polanyi6b (BEP) prin-
ciple etc. According to these models, a stronger nucleophile
(δσX < 0) leads to an earlier TS with a lower degree of bond
formation (and bond cleavage of the leaving group), which in
the present case should give a lower degree of Cα–N bond form-
ation with a low degree of progress in the hydrogen bonding by
the N–H (D) proton i.e., a smaller kH/kD value for a stronger
nucleophile (δσX < 0
δ (kH/kD) < 0) as was observed with
all the dicarbonyl activated compounds, 1–4. This means that
the TS structures, or the TS positions along the reaction co-
ordinate, are largely determined by the product stability, and a
O r g . B i o m o l . C h e m . , 2 0 0 4 , 2, 1 2 1 3 – 1 2 1 6
1214

View Article Online
Table 3 Kinetic isotope effects on the second-order rate constants (k2) for the reactions of ethyl α-acetyl-β-phenylacrylates with deuterated
X-benzylamines in acetonitrile at 25.0 ЊC
X
Y
kH × 102/MϪ1sϪ1
kD × 102/MϪ1sϪ1
kH/kD
p-OMe
p-OMe
p-OMe
p-OMe
p-Cl
p-Cl
p-Cl
p-OMe
p-Me
H
p-Cl
p-OMe
p-Me
H
0.767( 0.005)
1.02( 0.004)
1.47( 0.01)
0.341( 0.002)
0.481( 0.005)
0.753( 0.006)
1.34( 0.02)
0.124( 0.001)
0.170( 0.001)
0.251( 0.003)
0.412( 0.003)
2.24 0.02a
2.12 0.02
1.95 0.02
1.78 0.03
2.38 0.02
2.19 0.02
1.98 0.03
1.82 0.02
2.38( 0.02)
0.295( 0.001)
0.372( 0.002)
0.496( 0.004)
0.749( 0.005)
p-Cl
p-Cl
a Standard deviations.
Table 4 Activation parametersa for the reactions of ethyl α-acetyl-β-
greater reactivity is accompanied by a lower selectivity, i.e., the
reactivity–selectivity (RSP)7 holds for the dicarbonyl activated
olefins.
phenylacrylates with X-benzylamines in acetonitrile
X
Y
∆H≠/kcal molϪ1
Ϫ∆S≠/cal molϪ1 KϪ1
Contrary to this, a larger kH/kD value was observed with a
stronger nucleophile (δσX < 0
δ (kH/kD) > 0) for the benzyl-
p-OMe
p-OMe
p-Cl
p-OMe
p-Br
p-OMe
p-Br
4.9
5.3
4.9
4.6
52
48
54
53
amine additions in acetonitrile to olefins activated by other
(than dicarbonyl) groups: BMN (5), benzylidenemalononitrile
(YC H CH᎐C(CN) )1b; NS (6), β-nitrostyrene (YC H CH᎐
p-Cl
᎐
᎐
6
4
2
6
4
CH(NO2))1a NSB (7), β-nitrostilbene (YC H CH᎐C(C H )
a Calculated by the Eyring equation. The maximum errors calculated
(by the method of Wiberg12) are 0.6 kcal molϪ1 and 3 e.u. for ∆H≠
and ∆S≠, respectively.
᎐
6
4
6
5
NO )1c; CNS (8), β-cyano-4Ј-nitrostilbene (YC H CH᎐C(CN)ؒ
᎐
2
6
4
C6H4NO2)1c ECC (9), ethyl-α-cyanocinnamate, (YC H CH᎐
᎐
6
4
C(CN)COOEt).1d A later TS with a greater degree of bond
formation with a greater extent of hydrogen bonding by the
N–H (D) proton leading to a larger value of kH/kD (δ(kH/kD) > 0)
with a stronger nucleophile (δσX < 0) represents an anti-
Hammond effect6a or an intrinsic effect,3b i.e., a greater
reactivity leads to a greater selectivity, and an anti-RSP.3b This
is no doubt related to the stronger anionic charge development
on Cβ with a stronger nucleophile due to the lag in the delocal-
ization of charge onto the activating groups, which are not
dicarbonyls. This kind of charge imbalance causes a reduction
in the intrinsic rate2,5 (ko), and hence the effect on selectivity
should be intrinsic so that the anti-Hammond effect,6a or anti-
RSP, should hold as we have observed with the trend of
changes in the magnitude of kH/kD for 5–9, as well as the
substantially larger magnitude (kH/kD ≅ 2.3 ∼ 2.8) than for the
dicarbonyl series (kH/kD ≅ 1.2 ∼ 2.4). It is therefore clear that
the dicarbonyl group, irrespective of whether it has a ring
structure (as in 1 and 2) or not (3 and 4), has very small TS
imbalance effect in the amine additions in acetonitrile. This is
evident since for the dicarbonyl activated series the thermo-
dynamic models, which are based on product stability, i.e., the
Hammond effect and the BEP principle,6 hold. This means that
due to strong resonance delocalization through dicarbonyl
groups in the TS the charge imbalance effect becomes small.
The resonance delocalization is especially efficient with ring
structure5 as we noted above.
activation barrier,5 as we noted above (Table 2). In fact the
noncyclic dicarbonyls such as dimethyl malonate, 10 (CH2–
(COOCH3)2), and 1,3-cyclohexadione, 11 (CH2(COCH2CH3)2),
have nonplanar twisted structures ranging 35–86o (calculated at
the B3LYP/6-311ϩG** level).9 When these are deprotonated
the planar structures, IV, are formed so that there should be
some reorganization energies required. This should increase the
activation barrier. Thus the anionic center developing on Cβ in
the noncyclic dicarbonyl activated series (3 and 4) should have
lower rate than the cyclic dicarbonyl series (1 and 2) as observed
in Table 2.
Although the trends in the change in kH/kD with variation of
substituent Y in the substrate for the dicarbonyl series (1–4)1e,g
and others (5–9)1a–d are opposite, it is rather ambiguous as
to the effect of Y on the RSP. An electron donor Y (δσY < 0)
should increase negative charge on Cα leading to a lower
rate of attack by the nucleophile, benzylamine, but should be
weakly stabilizing the neutral product by donating electrons to
electron-withdrawing activating groups.
The activation parameters, ∆H≠and ∆S≠, are shown in Table 4.
Since the reaction proceeds by a concurrent bond formation of
N–Cα and H–Cβ bonds, the ∆H≠ values are rather low, but the
Ϫ∆S≠ values are large due to a constrained hydrogen bonded
TS structure, V. In this structure the amine hydrogen is
bonded to Cβ not to a carbonyl oxygen as in the six-membered
TS structure, III. The charge on Cβ in the TS should be stronger
due to the lag (although it may be small, it is not absent entirely)
in the resonance delocalization of anionic charge into the
activating groups, and hence should lead to a larger kH/kD value
with a greater degree of N–H (D) bond cleavage than expected
from a synchronous resonance delocalization. The hydrogen
bonding to a carbonyl oxygen will be very weak since in the TS
delocalization will not be complete.
Dicarbonyl groups attached to a carbanionic center are
known to have enormous charge transfer stabilization energies
due to the two strong vicinal nC
π*C᎐O charge transfer inter-
actions,8 Ϫ∆ECT > 250 kcal molϪ1 at the NBO-B3LYP/6-
311ϩG** level.9 Due to the delocalization of anionic charge
on the carbanionic center, Cβ, into the two carbonyl groups, the
dicarbonyl moiety becomes planar, IV.
However, if it has a built-in planar structure as in the cyclic
dicarbonyls (1 and 2), the reorganization required to form such
a planar structure is not required, which should result in a lower
O r g . B i o m o l . C h e m . , 2 0 0 4 , 2, 1 2 1 3 – 1 2 1 6
1215

View Article Online
Lastly the timing of the two processes, anionic charge
development on Cβ and hydrogen bonding to the Cβ, is rather
difficult to clearly envisage. These two would not take place in
the absence of N–Cα bond formation. The electronic shift to
form an anionic center should follow immediately and then a
proton shift to Cβ. Since both shifts, electron and proton, are
fast, at best these two could conceivably occur synchronously
with bond formation.
reduced, and two new peaks for CH–CH in the product,
p-CH3OC6H4(MeOC6H4CH2NH)CH–CH(COCH3)CO2C2H5,
grew at 3.46 and 4.77 ppm as the reaction proceed. No other
peaks or complications were found during the reaction except
the 3 peak height changes indicating that the reaction proceeds
with no other side reactions.
Acknowledgements
This work was supported by a Korea Research Foundation
Grant (KRF-2002-070-C00061).
Experimental
Materials
References
Merck GR acetonitrile was used after three distillations.
The benzylamine nucleophiles, Aldrich GR, were used after
recrystallization.
1 (a) H. K. Oh, J. H. Yang, D. D. Sung and I. Lee, J. Chem. Soc.,
Perkin Trans. 2, 2000, 101; (b) H. K. Oh, J. H. Yang, H. W. Lee
and I. Lee, J. Org. Chem., 2000, 65, 2188; (c) H. K. Oh, T. S. Kim,
H. W. Lee and I. Lee, J. Chem. Soc., Perkin Trans. 2, 2002, 282;
(d ) H. K. Oh, J. H. Yang, Y. H. Hwang, H. W. Lee and I. Lee, Bull.
Korean Chem. Soc., 2002, 23, 221; (e) H. K. Oh, T. S. Kim, H. W. Lee
and I. Lee, Bull. Korean Chem. Soc., 2003, 24, 193; ( f ) H. K. Oh,
J. H. Yang, H. W. Lee and I. Lee, J. Org. Chem., 2000, 65, 5391;
(g) H. K. Oh, I. K. Kim, H. W. Lee and I. Lee, submitted to J. Org.
Chem.
2 (a) C. F. Bernasconi, Acc. Chem. Res., 1987, 20, 301; (b) C. F.
Bernasconi, Tetrahedron, 1989, 45, 4017.
3 (a) I. Lee, Chem. Soc. Rev., 1990, 19, 317; (b) I. Lee, Adv. Phys. Org.
Chem., 1992, 27, 57; (c) I. Lee and H. W. Lee, Coll. Czech. Chem.
Commun., 1999, 64, 1529.
4 (a) C. D. Ritchie, in Solute-Solvent Interactions, eds J. F. Coetzee
and C. D. Ritchie, Marcel Dekker, New York, 1969, ch. 4; (b) J. F.
Coetzee, Prog. Phys. Org. Chem., 1967, 4, 54; (c) W. J. Spillane,
G. Hogan, P. McGrath, J. King and C. Brack, J. Chem. Soc., Perkin
Trans. 2, 1996, 2099; (d ) I. Lee, C. K. Kim, I. S. Han, H. W. Lee,
W. K. Kim and Y. B. Kim, J. Phys. Chem. B, 1999, 103, 7302.
5 C. F. Bernasconi and M. W. Stronach, J. Am. Chem. Soc., 1991, 113,
2222.
6 (a) A. Pross, Theoretical and Physical Aspects of Organic Reactivity,
Wiley, New York, 1995, ch. 5; (b) M. J. S. Dewar and
R. C. Dougherty, The PMO Theory of Organic Chemistry, Plenum,
New York, 1975, ch. 5.
Preparation of ethyl ꢀ-acetyl-ꢁ-phenylacrylates
The ethyl α-acetyl-β-phenylacrylates were prepared by the
literature method of Horning et al.10 Equimolecular amounts
of benzaldehyde (10 mmol) and ethyl acetoacetate(10 mmol)
were dissolved in the minimal amount of pyridine and refluxed
for 1 h. Solvent was removed under reduced pressure and
product was separated by column chromatography (silica gel,
10% ethylacetate-n-hexane) (yield >85%). IR (Nicolet 5BX
1
FT-IR) and H and 13C NMR (JEOL 400 MHz) data were
found to agree well with the literature values.10
Kinetic measurements
The reaction was followed spectrophotometrically by monitor-
ing the decrease in the concentration of ethyl α-acetyl-β-phenyl-
acrylate, [EAP], at λmax of the substrate to over 80% com-
pletion. The reaction was studied under pseudo-first-order
conditions, [EAP] = 8.0 × 10Ϫ5 M and [BA] = 0.02–0.05 M at
25.0 0.1 ЊC. The pseudo first-order rate constant, kobs, was
determined from the slope of the plot (r > 0.995) ln[EAP]
(2.303 log [EAP ] vs. time. Second-order rate constants, kN, were
obtained from the slope of a plot (r > 0.993) of kobs vs. benzyl-
amine with more than four concentrations of more than three
runs and were reproducible to within 3%.
7 (a) A. Pross, Adv. Phys. Org. Chem., 1997, 14, 69; (b) E. Buncel and
H. Wilson, J. Chem. Educ., 1987, 64, 475.
8 I. Lee, I. S. Han, C. K. Kim and H. W. Lee, Bull. Korean Chem. Soc.,
2003, 24, 1141.
9 (a) A.E. Reed, L.A. Curtis and F. Weinhold, Chem. Rev., 1988, 88,
899; (b) N. D. Epiotis, W. R. Cherry, S. Shaik, R. Yates and
F. Bernardi, Structural Theory of Organic Chemistry, Springer-
Verlag, Berlin, 1977, part IV.
Product analysis
The analysis of final products was difficult due to partial
decomposition during product separation and purification. We
therefore analysed the reaction mixture by NMR (JEOL 400
MHz) at appropriate intervals under exactly the same reaction
conditions as the kinetic measurement in MeCN at 25.0 ЊC.
Initially we found a peak for CH in the reactant, p-CH3OC6-
10 E. C. Horning, M. S. Fish and G. N. Walker, Org. Synth., 1983, 4,
408.
11 (a) C. Hansch, A. Leo and R. W. Taft, Chem. Rev, 1991, 91, 165;
(b) A. Fischer, W. J. Galloway and J. Vaughan, J. Chem. Soc., 1964,
3588; (c) H. K. Oh, J. Y. Lee and I. Lee, Bull. Korean Chem. Soc.,
1998, 19, 1198.
12 K. B. Wiberg, Physical Organic Chemistry, Wiley, New York, 1964,
p. 378.
H CH᎐C(COCH )CO C H , at 7.48 ppm, which was gradually
᎐
4
3
2
2
5
O r g . B i o m o l . C h e m . , 2 0 0 4 , 2, 1 2 1 3 – 1 2 1 6
1216