Effects of HMPA on the Structure and Reactivity of the Lithium Enolate of Cyclopentanone in THF: The Dimer is Responsible for Alkylation and Proton Exchange Reactions

Article abstract of DOI:10.1246/bcsj.77.259

The structure and reactivity of the lithium enolate of cyclopentanone are strongly influenced by coexisting HMPA molecules. Detailed low-temperature ⁷Li, ³¹P, and ¹³C NMR studies of the 0.04-0.2 M THF solutions indicate that excess quantities of HMPA act primarily to form a bis-HMPA coordinated dimer. Tetrameric, trimeric, and monomeric enolates were not detected by HMPA titration. The convergency on the dimeric aggregate has been monitored by the successive displacement of THF with HMPA molecules. Even at a high concentration of HMPA, the formation of monomeric enolate was not observed. Kinetic experiments, combined with the structural information, showed that alkylation of the enolate by methyl iodide proceeds via the dimer with the assistance of HMPA. Proton exchange between the enolate and 2-methylcyclopentanone also occurs via the dimer, but without the participation of additional HMPA. These findings explain the role of HMPA in the selective monoalkylation of lithium enolate.

Full text of DOI:10.1246/bcsj.77.259

ꢀ
2004 The Chemical Society of Japan
Bull. Chem. Soc. Jpn., 77, 259–268 (2004)
259
Headline Articles
Effects of HMPA on the Structure and Reactivity of the Lithium Enolate
of Cyclopentanone in THF: The Dimer is Responsible for Alkylation
and Proton Exchange Reactions
ꢀ
1
ꢀ;1
Masaaki Suzuki, Hiroko Koyama, and Ryoji Noyori
Division of Regeneration and Advanced Medical Science, Graduate School of Medicine, Gifu University,
Yanagido, Gifu 501-1193
1Department of Chemistry and Research Center for Materials Science, Nagoya University,
Chikusa, Nagoya 464-8602
Received August 20, 2003; E-mail: suzukims@biomol.gifu-u.ac.jp
The structure and reactivity of the lithium enolate of cyclopentanone are strongly influenced by coexisting HMPA
7
31
13
molecules. Detailed low-temperature Li, P, and C NMR studies of the 0.04–0.2 M THF solutions indicate that excess
quantities of HMPA act primarily to form a bis-HMPA coordinated dimer. Tetrameric, trimeric, and monomeric enolates
were not detected by HMPA titration. The convergency on the dimeric aggregate has been monitored by the successive
displacement of THF with HMPA molecules. Even at a high concentration of HMPA, the formation of monomeric enolate
was not observed. Kinetic experiments, combined with the structural information, showed that alkylation of the enolate by
methyl iodide proceeds via the dimer with the assistance of HMPA. Proton exchange between the enolate and 2-methyl-
cyclopentanone also occurs via the dimer, but without the participation of additional HMPA. These findings explain the
role of HMPA in the selective monoalkylation of lithium enolate.
Lithium enolates are organic nucleophiles that are indispen-
sable for the synthesis of many theoretically and practically im-
found that an HMPA additive had significant effects.1,6c,d The
reaction of 3 with 5 molar amounts of methyl iodide in THF
1
ꢁ
portant compounds. Synthetic, structural, and kinetic studies
over the past several decades have provided enormous progress
at 0 C resulted in a 79% yield of 2-methylcyclopentanone
(4), together with 15% polymethylated ketones (Scheme 1) that
resulted from undesired enolate/ketone proton exchange reac-
tions. The presence of 3 molar amounts of HMPA, however,
1–5
in this chemistry. Our interest in this subject stemmed from
6
the three-component prostaglandin synthesis. The key to our
ꢁ
success was the modulation of the reactivity of a cyclopenta-
none enolate, generated in situ by conjugate addition of a lower
side-chain organometallic to a protected 4-hydroxy-2-cyclo-
pentenone 1 (Chart 1). Under certain controlled conditions,
the regio-defined enolate can be trapped with an upper side-
chain iodide, stereoselectively producing 2 without ꢀ-proton
exchange; under these conditions, ketones also coexist in the
significantly accelerated the alkylation at ꢂ78 C to give a
94% yield of 4 accompanied by only 3.3% dimethylcyclopen-
tanones (2,5/2,2 = 7).6 The main purpose of this study is the
elucidation of the role of HMPA in this reaction system.2
The reactivity and selectivity of lithium enolates are pro-
foundly influenced by the structures and reaction conditions, in-
c
e,8
cluding particularly the solvents, the concentrations, and addi-
2–5,9–11
6,7
reaction system. The high selectivity for alkylation over pro-
ton exchange is crucial for the efficient vicinal carba-condensa-
tion of ꢀ,ꢁ-unsaturated ketones. When we investigated the be-
havior of a simple cyclopentanone enolate 3 in this reaction, we
tives interacting with Li.
A milestone in this significant
Scheme 1. Methylation of cyclopentanone lithium enolate by
methyl iodide.
Chart 1.
Published on the web February 10, 2004; DOI 10.1246/bcsj.77.259

2
60 Bull. Chem. Soc. Jpn., 77, No. 2 (2004)
HEADLINE ARTICLES
subject was marked by Seebach,3,4 who elucidated the X-ray
structures of aggregates of lithium enolates crystallized with
various solvents and ligands. This classic work implied the pos-
sibility of participation of these aggregates in the reactions of
3, which had been prepared by reaction of 1-(trimethylsiloxy)-
cyclopentene and butyllithium.16 While its THF solvate was
observed to form a tetrameric structure possessing a Li4O4 cube
4
in the crystalline state, the melting-point depression measure-
ꢁ
9
lithium enolates. Although Jackman hypothesized that a tetra-
ment in 0.47 M THF solution at ꢂ107 C suggested an average
meric aggregate of the isobutyrophenone lithium enolate is di-
rectly involved in both the C- and O-alkylation reactions with
dimethyl sulfate, which take place in a 0.16 M (1 M = 1
aggregation number of 2.8.17 Thus, the addition of HMPA to
the solution would be expected to result in its dissociation to
a dimer with a Li2O2 square and, possibly, to a monomeric
enolate2e having a contact or solvent-separated ion-paired
structure.5c,14b
ꢂ3
mol dm ) solution of 1,3-dioxolane or 1,2-dimethoxyethane,
the actual structure of the reactive enolate remained unclear.12
To attempt to correlate reactivity with aggregation state,
Streitwieser recently designed lithium enolates possessing a
We examined the enolate structures in THF in the presence
7
31
13
of HMPA using low-temperature Li, P, and C NMR, by
2
7
31
UV-detectable chromophore. By performing detailed spectro-
measuring characteristic chemical shifts and the Li– P cou-
pling constant.5 The lithium enolate 3 was found to have es-
tablished the dynamic equilibrium in ethereal solvents illustrat-
ed in Scheme 2. According to this Scheme, for both the tetramer
T and the dimer D, the first subfix represents the number of at-
tached HMPAs and the second subfix denotes the number of co-
c,14
scopic and rate studies in dilute THF solutions, they found that
a monomeric lithium enolate equilibrating with its dimer or tet-
ramer reacts with benzyl bromide in a concentration range of
ꢂ4
ꢂ3
2b,c
6
:2 ꢃ 10 M to 1:1 ꢃ 10 M.
Moreover, an HMPA ad-
dend modulated the aggregation system, resulting in monomer-
ꢂ4
18
ic lithium enolates at a 10 M concentration, and the resulting
monomers were observed to undergo further HMPA coordina-
tion to form a solvent-separated ion pair that participated in the
ordinated THF ligands. The equilibrium point is thermody-
namically determined by the enolate concentration and the
quantity of HMPA. In the presence of excess HMPA, the eno-
late largely exists as a bis-HMPA coordinated dimer D2,2 in
2
e
alkylation.
THF.1 Tetramers,
9,20
19,20
trimers,
5c,21
and monomers,
5c,14b
how-
NMR spectroscopy provides valuable information about the
nature of lithium enolates in a higher concentration and about
the equilibria in which they participate. Unlike lithium amides,
ever, were not detected.
7
Li- and 31P-NMR spectra of 0.16 M THF solutions of 3 at
which demonstrate 6 Li– N coupling, there have been few
=7
15
13
ꢂ100 C are shown in Fig. 1, A and B, respectively. Without
HMPA, the enolate exists as a mixture of T0,4 and D0,4 in a ratio
ꢁ
22
5
reports on lithium enolates, due to the lack of Li–O coupling.
Understanding this problem, Jackman cleverly used C spin–
13
17
of 2:3. These aggregates are in rapid equilibrium even at
ꢁ
7
lattice relaxation time and Li quadrupole-splitting constants
to analyze the structure of the enolate of isobutyrophenone,
concluding that this enolate forms a tetrameric structure at a
ꢂ100 C, giving a broad singlet at ꢂ0:36 ppm (Fig. 1 Aa).
The addition of 6 molar amounts of HMPA generated D2,2 ex-
clusively,23 giving a 1:1 Li doublet at ꢂ0:37 ppm (Fig. 1 Ah)
7
22
ꢂ1
M concentration.5a Furthermore, a study on lithium 3,5-
31
1
dimethylphenolate as a model of lithium enolates, using Li–
0
and a 1:1:1:1 P quartet at 25.9 ppm (Fig. 1 Bh) with the same
7
2
7
31
2
31
7
coupling constant J( Li– P) = J( P– Li) = 9.5 Hz.
By careful monitoring the change of intensity of quartet sig-
nals in 31P NMR as a function of added HMPA, we were able to
analyze transient and terminated enolate structures. Upon addi-
31
2
7
31
14
P spin–spin long range coupling, J( Li– P), showed that,
in a 0.2 M ethereal solution containing 5 molar amounts of
HMPA, the phenolate has an HMPA-coordinated tetrameric
5
c
31
structure.
tion of 0.25 molar amount of HMPA (Fig. 1 Bb), the P quartet
2
31
7
These findings suggested that the aggregation state of lithi-
um enolates in solution is highly dependent on their organic
backbones, intrinsic basicity and concentration, the steric fac-
tors of the coexisting solvent and other donor compounds,
due to D1,3 appeared at 26.1 ppm, with J( P– Li) = 10.3 Hz.
3
1
When >0:5 molar amount of HMPA were added, a P quartet
signal assignable to D2,2 was observed at 25.9–26.0 ppm with
J( P– Li) = 9.5 Hz. Addition of 2 molar amounts of HMPA
2
31
7
3,10,11
and the temperature of the reaction.
Obviously, each ag-
reversed the relative ratio of D1,3 and D2,2 (Fig. 1 Bf), and the
addition of 6 molar amounts of HMPA generated D2,2 exclu-
sively (Fig. 1 Bh). Thus the tetramer/dimer equilibrium in
the absence of HMPA is shifted to a dimeric structure by the
coordination of HMPA, establishing the D1,3/D2,2 equilibrium.
This is further supported by the appearance of free HMPA as a
broad singlet at 26.1 ppm after the addition of 0.75 and 1.0 mo-
lar amounts of HMPA and the convergence on D2,2 by increas-
ing amounts of HMPA.
gregate has a different nucleophilicity and basicity, resulting
in complicated selectivity profiles.15
In this study, we demonstrate the participation of a dimeric
enolate in an alkylation and proton exchange reaction. Using
NMR, we have directly characterized the enolate responsible
for this reaction. The role of HMPA was also clarified by both
structural and kinetic methods. To our knowledge, this is the
first such report showing the role of a dimeric enolate in this re-
action.
The profiles of the HMPA titration in THF were also quan-
titatively determined (Fig. 2), by calculating the relative
amounts of each enolate species and plotting them based on
Results and Discussion
3
1
Structures of Cyclopentanone Lithium Enolate in the
Presence of HMPA. Enolate structures in ethereal solution
are highly sensitive to their organic frameworks, whether cyclic
or acyclic, sterically congested or uncongested, or electronical-
the P signal intensities of D1,3, D2,2 and free HMPA observed
in Fig. 1B (for the details of the calculation, see the Experimen-
tal Section). The HMPA-free enolate species are initially re-
placed by D1,3 and subsequently by D2,2. Here, the D1,3/D2,2
equilibrium is controlled by the higher coordination ability of
3,10,11
ly conjugated or nonconjugated.
6
In connection with our
prostaglandin synthesis, we had selected the lithium enolate
24
HMPA to lithium than THF by approximately 100 times.

M. Suzuki et al.
Bull. Chem. Soc. Jpn., 77, No. 2 (2004)
261
Scheme 2. Equilibrium of cyclopentanone lithium enolate in the presence of HMPA.
The monomers were undetectable in both Li and 31P spec-
tra. Even at an enolate concentration as low as 0.04 M, or at as
high as 8 molar amounts of HMPA, no high-field broad singlets
arising from rapid exchange with free HMPA,5 or quintets in
the ꢂ0:8 to ꢂ0:4 ppm region due to ion-paired species, were
7
the corresponding bis-HMPA coordinated dimer was independ-
ent of the quantity of added HMPA, but was influenced by the
enolate concentration, confirming the establishment of the
c,14
19
equilibrium between T4,0 and D2,2 as shown in Scheme 2.
Further, the ratio of the corresponding tetramer and dimer
changes in the order of 5:95, <1:99, and 0:100 in diethyl ether,
dimethyl ether, and THF, respectively, depending on the basic-
ity of S.19 Thus, in more solvative THF, only the dimer D2,2 was
observed.
7
5c,14
observed in Li NMR.
In addition, the profile of the conver-
gence on D2,2 in the HMPA titration of the enolate and the
clear-cut structural assignment gave no indication of any tetra-
mers The rather broad Li and P signals
are due to the exchange processes under the experimental con-
19,20
5c,21
7
31
and trimers.
Full spectral data for the cyclopentanone lithium enolate ag-
gregates in THF are summarized in Table 1.
The convergence of the dimeric structure in the presence of
7
31
ditions. In fact, the Li and P signals observed in diethyl ether
ꢁ
and dimethyl ether at ꢂ100 to ꢂ110 C are much sharper than
those in THF due to the slower S/L ligand exchange.1
9,20
In
excess amounts of HMPA was also confirmed by the C-NMR
13
1
3
such ethereal solvents, the HMPA-coordinated tetramers de-
picted in Scheme 2 were observed as minor components.1
Here, the ratio of the tetra-HMPA coordinated tetramer and
signal of their oxygen-bearing carbon. Thus, the C-NMR
spectrum of THF-d8 containing 6 molar amounts of HMPA
showed only a single broad band at 169.5 ppm, indicating that
9,20

2
62 Bull. Chem. Soc. Jpn., 77, No. 2 (2004)
HEADLINE ARTICLES
ꢁ
7
31
Fig. 1. HMPA titration of a 0.16 M THF solution of cyclopentanone lithium enolate 3 at ꢂ100 C. A: Li-NMR spectra. B: P-NMR
spectra.
and 0.657 s, respectively. Although efforts have been made to
correlate QSC values with enolate aggregate structure for some
lithium enolates5
a,25,26
and phenoxides,
5b,c,25
no general rule has
emerged from these attempts. For any aggregate, it is difficult to
estimate the influence of a ligand solvating to Li, or of the bulk-
iness of the enolate anion, on QSC values.28 Lithium 2,6-di-
methylphenoxide is the only lithium enolate-related compound
known to have dimeric structure in dioxolane, THF, and pyri-
dine, with QSC values of 147, 169, and 116 kHz, respectively.5b
Since the dimeric structure of the cyclopentanone lithium eno-
late D2,2 was deduced unambiguously by the HMPA titration,
the QSC value obtained in this study is the first for the lithium
enolate with a dimeric structure.
In summary, cyclopentanone lithium enolate 3 in THF
0.04–0.2 M) undergoes successive S/L ligand exchanges by
increasing HMPA to form D1,3 and D2,2, and eventually con-
verges on the bis-HMPA coordinated dimer D2,2 with excess
HMPA.29
Fig. 2. The percentage of each aggregate species as func-
tions of added HMPA for 0.16 M lithium enolate 3 in THF.
(
only D2,2 is formed under such a condition.
Kinetics of Alkylation of the Lithium Enolate. (a) Effects
of HMPA on Rates: The reactivity and selectivity of the
methylation of enolates are markedly influenced by the HMPA
addend. In THF, enolate 3 was reacted slowly with methyl io-
7
25,26
The values of Li quadrupole-splitting constants (QSC)
for a 0.2 M THF-d8 solution of cyclopentanone lithium enolate
3
calculated as 116 kHz, using the equation 7.5(T1( C)/
in the presence of 5 molar amounts of HMPA27 at 30 C were
ꢁ
1
3
ꢁ
dide at temperatures below ꢂ50 C. Addition of 0.64 M HMPA
7
1=2
4
13
T1( Li)) ꢃ 10 (Hz), with observed C (hydrogen-bearing
enhanced the rate of methylation of a 0.06 M solution of 3 by a
factor of 7500 ((ꢂd[CH3I]/dt)/[CH3I] = 8:67 ꢃ 10ꢂ8
7
ꢂ1
vinylic carbon) and Li spin–lattice relaxation times of 1.56
s
vs

M. Suzuki et al.
Bull. Chem. Soc. Jpn., 77, No. 2 (2004)
263
Table 1. 7Li and 31P Chemical Shifts and Coupling Con-
stants for the HMPA Solvated Dimer of Cyclopentanone
a)
Lithium Enolate in THF
7
b)
7
b)
Li (s )
c)
Li (d )
7
mol. amt. of
HMPA
c)
2
31 d)
ꢂ
ꢂ ( J( Li– P) )
D1,3
D1,3
D2,2
0
0
0
1
2
4
6
.25
.5
.75
.0
.0
.0
f
f
e)
g)
ꢂ0:32
ꢂ0:34
ꢂ0:30
ꢂ0:36 (10.7)
e)
e)
g)
ꢂ0:37 (10.5)
g)
ꢂ0:36 (10.1)
ꢂ0:37 (9.5)
ꢂ0:37 (9.5)
b)
.0
3
1
P (q )
c)
2
31
7
d)
ꢂ
( J( P– Li) )
D1,3
D2,2
0
0
0
1
2
4
6
.25
.5
.75
.0
.0
.0
26.1 (10.3)
26.1 (10.3)
26.0 (10.3)
26.0 (10.3)
26.0 (10.3)
26.0 (9.5)
25.9 (9.5)
25.9 (9.5)
25.9 (9.5)
25.9 (9.5)
25.9 (9.5)
.0
ꢁ
a) Spectra were recorded at ꢂ100 C. b) s = singlet, d = dou-
blet, and q = quartet. c) The chemical shifts are reported rela-
tive to external standard as described in the Experimental Sec-
tion. d) All coupling constants J are reported in hertz. e) The
signal was superimposed on the lower field peak of doublet
due to HMPA-coordinated Li. f) Unassignable broad signal.
g) The signal involves HMPA-coordinated Li atoms in both
D1,3 and D2,2.
Fig. 3. Log–log plots of initial rates/[CH3I] and initial
rates/[4] as a function of the concentrations of the dimeric
lithium enolate D2,2 in THF for alkylation (A) and proton
abstraction reaction (B), in the presence of a fixed concen-
tration of HMPA. The ꢀ (ꢀ ) values derived from the
slopes were 1.0 for both reactions.
0
ꢂ4 ꢂ1
ꢁ
30
6
tion of HMPA accelerated proton abstraction from 2-methylcy-
:53 ꢃ 10
s
at ꢂ75 C). In contrast, the same concentra-
sary side reactions such as polyalkylation and the aldol reac-
tion. Thus alkylation was carried out using 0.034–0.067 M of
clopentanone (4), resulting in polymethylation, by only a factor
ꢂ5 ꢂ1
ꢁ
ꢂ5 ꢂ1
D2,2 in THF at ꢂ75 C under pseudo-first-order conditions,
of five ((ꢂd[4]/dt)/[4] = 1:80 ꢃ 10
s
vs 9:46 ꢃ 10
s
ꢁ
with constant concentrations of CH3I (8 mM) and HMPA
(0.64 M). The rate was determined within 10% conversion after
rapid quenching of the unreacted D2,2 with chlorotrimethylsi-
lane. Similarly, the proton exchange rate between D2,2
at ꢂ50 C). These combined effects of HMPA resulted in sig-
nificantly increased selectivity of monomethylation.6c
(b) Aggregation State of the Reacting Enolate: We used
enolate 3 concentrations of 0.06 to 0.13 M, and HMPA concen-
trations of 0.36 to 0.64 M, in THF. Under these concentrations,
exists only as a bis-HMPA coordinated dimer D2,2. Since con-
centrations of the other aggregates in the equilibrium
Scheme 2) are negligible, the rates for the alkylation and pro-
(
0.028–0.067 M) and the methyl ketone 4 (10 mM) in THF at
ꢁ
32
ꢂ50 C was determined by monitoring the consumption of
3
4
by rapid silylation of the enolates or the formation of the eno-
0
late of 4. Consequently, the ꢀ and ꢀ values were determined to
be 1.0 for both the alkylation and proton abstraction reactions
from the slopes shown in Fig. 3, indicating that the average ki-
netic aggregation number for each reaction is 2.0. These find-
ings indicate that the dimeric enolate is responsible for both re-
actions. This result also eliminates the possibility that undetect-
ed, but highly reactive monomeric enolate species, participate
in these reactions. This conclusion is in contrast to that of
Streitwieser’s work on different Li enolates, clarifying that
(1) the monomeric enolates are responsible for alkylation,
and (2) proton exchange is promoted by both the monomeric
2d
and dimeric enolates.
(
ton abstraction reactions are simplified to Eqs. 1 and 2, respec-
2
tively.
ꢀ
p
ꢂ d½CH3Iꢄ=dt ¼ kR½D2,2ꢄ ð½HMPAꢄ ꢂ 2½D2,2ꢄÞ ½CH3Iꢄ ð1Þ
0
0
p
ꢀ
ꢂ d½4ꢄ=dt ¼ kH½D2,2ꢄ ð½HMPAꢄ ꢂ 2½D2,2ꢄÞ ½4ꢄ
The superscript ꢀ (or ꢀ ) is the reaction order, which is equal to
nk=na, where na is the average equilibrium aggregation number
ð2Þ
0
2
2
b,c
2b,c,e,f
and nk is the average kinetic aggregation number. The super-
0
script p (or p ) is the number of free HMPA molecules that
modulate the reaction of the aggregate.
31
0
The reaction orders (ꢀ and ꢀ ) can be determined on the ba-
sis of the reaction rates and the concentrations of the dominant
dimer. The reaction conditions were chosen to be consistent
with the structural study, as well as to minimize any unneces-
(c) Kinetic Roles of HMPA in Alkylation and Proton
Abstraction: Rate studies in the presence of HMPA revealed
that free HMPA molecules behave differently during the meth-
ylation2 and proton abstraction reactions.
e
31

2
64 Bull. Chem. Soc. Jpn., 77, No. 2 (2004)
HEADLINE ARTICLES
Fig. 5. Plots of the initial rates of alkylation in THF against
HMPA concentrations. The line equation is expressed by
ꢂ2
ꢂ1 ꢂ1
ꢂ2
rate/[D2,2][CH3I] = 1:35 ꢃ 10
M
s
+ 4:54 ꢃ 10
ꢂ2 ꢂ1
M
s
([HMPA] ꢂ 0.36 M) at 0.03 M D2,2 and 8 mM
CH3I.
Fig. 4. HMPA dependence of the rates of alkylation (A) and
proton abstraction reaction (B) in THF. The slopes, which
correspond to the reaction order, indicate that one molecule
of HMPA participates in the alkylation, while the proton
abstraction rates are independent of HMPA concentration.
The alkylation reaction conducted in THF in the presence of
–9 molar amounts of HMPA (0.36–0.54 M), with the constant
6
concentrations of D2,2 (0.03 M) and CH3I (8 mM). A plot of
log((ꢂd[CH3I]/dt)/[D2,2][CH3I]) vs log([HMPA] ꢂ 2[D2,2])
gave approximate first-order behavior for free HMPA (Fig.
4
A). Thus, the alkylation rate was calculated from Eq. 3.
ꢂ d½CH3Iꢄ=dt ¼ kR½D2,2ꢄð½HMPAꢄ ꢂ 2½D2,2ꢄÞ½CH3Iꢄ
ð3Þ
The observed initial rate is increased consistently by the in-
crease of HMPA addend at HMPA concentrations more than
.36 M (Fig. 5), where only D2,2 is observed by NMR
Scheme 3. Possible transition state structures for alkylation
and proton abstraction reactions.
0
33
(
1
Fig. 1). The resulting linear correlation is expressed by
ꢂ2
ꢂ1 ꢂ1
ꢂ2
ꢂ2 ꢂ1
ꢂ3
:35 ꢃ 10
M
s
+ 4:54 ꢃ 10
M
s
.36 M), indicating that the initial rate constant at 0.36 M
([HMPA] ꢂ
This yielded a second-order rate constant, kH ¼ 2:54 ꢃ 10
ꢂ1 ꢂ1
ꢁ
32
0
M
s
These results indicate that both the methylation and proton
, at ꢂ50 C.
ꢂ2
ꢂ1 ꢂ1
HMPA is 1:35 ꢃ 10
M
s
and the third-order rate con-
ꢁ
ꢂ2
ꢂ2 ꢂ1
stant (kR) is 4:54 ꢃ 10
M
s
(ꢂ75 C) at 0.03 M D2,2
exchange reactions of 3 involve the dimeric enolate, but that
the number of HMPA molecules involved in the transition
structures is different. This rate-law difference explains the ki-
netic role of HMPA in altering the selectivity of monoalkyla-
tion.1
and 8 mM CH3I. Thus, we concluded that an additional HMPA
molecule is involved in the methylation of D2,2 at HMPA con-
centration higher than 0.36 M.
,6c
In contrast, the rate of proton abstraction of 4 (10 mM) in
THF was independent of HMPA concentration (0.36–0.54 M)
(Fig. 4B). Thus the rate of this reaction is expressed by Eq. 4
in the same range of the HMPA concentration (>0:36 M) as
alkylation.
Possible Transition Structures. Various transition states
are conceivable for reactions of the lithium enolate 3. The fore-
going structural and rate studies urge us to propose the transi-
tion state structures, TSR and TSH, for the alkylation and pro-
ton abstraction reactions, respectively (Scheme 3). The enolate
dimer D2,2, which possesses a Li2O2 four-membered ring,
ꢂ d½4ꢄ=dt ¼ kH½D2,2ꢄ½4ꢄ
ð4Þ

M. Suzuki et al.
Bull. Chem. Soc. Jpn., 77, No. 2 (2004)
265
7
31
13
would equilibrate with the open form 5, with the assistance of a
solvent molecule or a ketonic substrate. The Li/O=C associa-
tion in 5 (S = ketone or solvent) would thus enhance the acidity
of the carbonyl ꢀ proton, and, at the same time, would increase
the basicity of the enolate moiety. As a consequence, proton ab-
straction would take place via the cyclic transition structure
194.25 MHz ( Li), 202.35 MHz ( P), or 125.65 MHz ( C). The
7Li and 31P chemical shifts were referenced to external standards,
0.41 M LiCl/THF-d8 (ꢂ 0.0) and 1.0 M P(C6H5)3/THF-d8 (ꢂ
ꢁ
ꢂ6:0), respectively, at ꢂ100 C. Shiming was performed on the
ꢁ
7
31
external reference at ꢂ100 C and Li- and P-NMR spectra were
1
3
then taken without locking. The C chemical shifts were refer-
enced to tetramethylsilane/THF-d8 (ꢂ 0.0) as an external standard
TSH, involving two Li atoms.34,35
Further, action of HMPA
ꢁ
at ꢂ100 C. Digital resolutions were 0.59, 0.79, and 0.23 Hz for
on D2,2 generates the open dimer 6, which possesses a more
highly polarized Li–O bond than 5, due to the additional HMPA
coordination that enhances the nucleophilicity of the enolate.
The open dimer 6 would then undergo SN2 displacement of
CH3I via TSR.
7
31
13
Li, P, and C, respectively.
7
13
7
13
The Li and C relaxation times (T1( Li) and T1( C)) were
measured with the inversion-recovery method, and data were proc-
essed by nonlinear least-squares programs.
Product Analysis. Gas chromatographical (GC) analyses were
performed on a Shimadzu GC-14B instrument with a flame ioniza-
tion (FID) and a capillary column (TC-1, 0.25 mm ꢃ 60 m, Shi-
Conclusions
This study shows the significance of dimeric lithium enolates
in organic reactions. In the presence of an excess amount of
HMPA, cyclopentanone lithium enolate 3 exists as a dimer in
THF. This dimer participates in alkylation with methyl iodide
and proton exchange with a cyclopentanone derivative. While
HMPA markedly enhances the reactivity of the enolate in alky-
lation, it does not affect proton exchange. This conclusion,
however, is limited to the cyclopentanone enolate 3 under the
above stated conditions.29 Lithium enolates can exist in various
forms, and each form has a different nucleophilicity and basic-
2
madzu Inc.); carrier gas, N2 (pressure of 0.5 kg/cm ); split ratio,
ꢁ
30:1; initial column temperature, 70 C for 8 min; final column
ꢁ
temp., 250 C; progress rate, 20 C/min; detection temp., 280
ꢁ
ꢁ
C; retention time (tR) of cyclopentanone, 8.5 min; tR of 2-methyl-
cyclopentanone, 11.7 min; tR of 1-(trimethylsiloxy)cyclopentene,
15.2 min.
General Procedure for Preparation of Cyclopentanone Lith-
ium Enolate. 1-(Trimethylsiloxy)cyclopentene (78.2 mg, 0.500
mmol) was placed into a 10-mL test tube with positive Ar pressure.
THF (1.5 mL) was added, and then the solution was transferred
through a stainless cannula into a 20-mL Schlenk tube. To the so-
3,10
ity. The structures and reactivities of these enolates are high-
ly dependent on their concentrations and steric and electronic
properties, as well as on the additives, solvents, and tempera-
ture of reaction.
lution was added butyllithium (1.43 M hexane solution, 0.344 mL,
ꢁ
0
.500 mmol) at 0 C via a syringe under Ar purge. The mixture was
left to stand at room temperature for 1.0 h under Ar atmosphere.
Preparation of Samples for Li and 31P NMR Spectroscopic
Analyses (Fig. 1). A THF solution of the lithium enolate was pre-
pared as described above. The solution was cooled to 0 C with an
ice bath, and 0.25 molar amount of HMPA (22.4 mg, 0.125 mmol)
was added, followed by additional THF to give a total volume of
7
Experimental
ꢁ
General. All apparatus used in the reactions were dried in an
ꢁ
oven (100 C) overnight and baked out with a heat gun under re-
duced pressure to remove air and moisture, and then filled with ar-
gon (Ar) after they had cooled to room temperature. The air and
moisture sensitive materials were manipulated under Ar using a
glove box, vacuum line, and syringe techniques.
Solvent and Materials. THF was freshly distilled over sodium
diphenylketyl. Solvents used for the NMR measurements (THF
and THF-d8) were distilled over sodium diphenylketyl, and then
transferred into a receiver containing a sodium–potassium alloy
and distilled under vacuum. These solvents were degassed prior
to use. Hexamethylphosphoric triamide (HMPA) was obtained
from Tokyo Kasei and distilled over CaH2 under reduced pressure.
3
.1 mL. A 5-mm NMR tube was charged with this solution (750
mL); the filled tube was cooled to 77 K and then sealed with a flame
under vacuum. Samples to which 0.50, 0.75, 1.0, 2.0, 4.0, and 6.0
molar amounts of HMPA were added were prepared according to
ꢁ
this procedure. NMR samples were kept at ꢂ80 C. Spectral con-
fidence was confirmed by repreating the same preparation proce-
ꢁ
dure. NMR samples were kept at ꢂ80 C and quickly used for
the NMR measurement. The solutions of some samples changed
from colorless to yellow after one week and were contaminated
7
by several undesirable peaks as judged by Li NMR.
Procedure to Deduce the Percentage of HMPA-Free
Aggregates, Mono- and Bis-HMPA Coordinated Dimers
1
grade (97%). NMR studies were done using the enol silyl ether
-(Trimethylsiloxy)cyclopentene was purchased as commercial
(
Fig. 2). When a molar amount of HMPA is added, the relation
ꢁ
of 99.8% purity after distillation (bp 156 C, 101 kPa) equipped
between the total quantity of HMPA molecules (hHMPAitotal)
total
with a Hempel-type distilling column. HMPA and the enol silyl
ether were stored in ampoules under Ar. Butyllithium in hexane
was purchased from Nacalai Tesque, Inc., and stored in a Schlenk
and the total quantity of Li atoms (hLii ) is given by Eq. 5:
hHMPAitotal ¼ ahLiitotal
ð5Þ
ꢁ
tube equipped with a Young’s tap under Ar at 4 C. Diphenylacetic
acid used as an indicator was recrystallized from methanol and
In the presence of HMPA, cyclopentanone lithium enolate exists as
a mixture of HMPA-free aggregates ((LiOEn) ), D1,3, and D2,2. In
this case, the total quantity of Li atoms is expressed as Eq. 6:
ꢁ
dried at 60 C under high vacuum. Methyl iodide obtained from
n
Nacalai was distilled over P4O10. 2-Methylcyclopentanone ob-
tained from Aldrich was purified by treatment with aqueous
KMnO4 and distilled over molecular sieves 13X. Nonane used as
the internal standard for the kinetic studies was distilled over mo-
lecular sieves 3A. Chlorotrimethylsilane and triethylamine were
distilled over CaH2.
hLiitotal ¼ ꢁn½ðLiOEnÞ ꢄ þ 2½D1,3ꢄ þ 2½D2,2ꢄ ðn ¼ 1; 2; . . .Þ
n
¼ fLiOEng þ 2½D1,3ꢄ þ 2½D2,2
ꢄ
ð6Þ
where {LiOEn} means the formal concentration of HMPA-free
enolate. Likely, the total quantity of HMPA molecules is given
by the sum of the quantity of free and coordinated HMPA.
NMR Spectroscopy. All the multinuclear NMR experiments
were recorded on a JEOL JNM ꢀ-500 spectrometer operated at

2
66 Bull. Chem. Soc. Jpn., 77, No. 2 (2004)
hHMPAitotal ¼ ½HMPAꢄ þ ½D1,3ꢄ þ 2½D2,2ꢄ
The following value p, which is obtained by the ratio of inten-
HEADLINE ARTICLES
Table 2. Calculated p and q Values and Percentages of Eno-
ð7Þ
late Aggregates in THF
sities of two 31P quartet signals, corresponds to the relative quan-
tity of the coordinated HMPA for D1,3 and D2,2 (Fig. 1B).
mol. amt. of
pa)
T0,4 and D0,4
D1,3
D2,2
q
HMPA (a)
%
%
%
b)
b)
b)
0
0
0
0
0
1
2
4
6
.25
.50
.75
.80
.90
.0
.0
.0
.0
0.00
0.16
0.26
0.33
0.38
0.46
3.3
0.00
0.10
0.71
0.74
0.91
0.93
—
44
20
19
17
17
9
56
74
71
71
70
74
37
3
0
6
p ¼ 2½D2,2ꢄ=½D1,3ꢄ
The value q, which is obtained by the integral ratio of a singlet and
two quartet signals, corresponds to the relative quantity between
free and coordinated HMPA.
b)
10
12
13
17
63
97
98
q ¼ ½HMPAꢄ=ð½D1,3ꢄ þ 2½D2,2ꢄÞ
From Eqs. 5–7, p and q, the relative concentration between
c)
0
0
c)
78
98
—
—
{LiOEn} and [D1,3] is derived (Eq. 8).
c)
0
2
fLiOEng=½D1,3ꢄ ¼ ðp þ 1Þðq þ 1Þ=a ꢂ ðp þ 2Þ
ð8Þ
a) Determined based on the relative intensities of right-side sig-
nals in quartets. b) In a range of added HMPA quantity less
than 0.5 mol. amt., the quantity of free HMPA cannot be esti-
mated directly because of its signal superimposition on that of
coordinated HMPA. Therefore, the p and q values were deter-
mined according to Eqs. 9–11 based on A/na, B, and C values
obtained by the extrapolation of the defined curve in a range of
Thus, the relative concentrations between D1,3 and D2,2 and be-
tween HMPA-free LiOEn and D1,3 are determined from p=2 and
ðp þ 1Þðq þ 1Þ=a ꢂ ðp þ 2Þ, respectively.
The numbers A, B, and C, are the proportions of {LiOEn},
[D1,3], and [D2,2], respectively, and are calculated by the following
Eqs. 9–11.
0
.75–1.0 mol. amt. of added HMPA. c) The enolate percentag-
A ¼ fLiOEng=ðfLiOEng þ ½D1,3ꢄ þ ½D2,2ꢄÞ
fðp þ 1Þðq þ 1Þ=a ꢂ ðp þ 2Þg
fðp þ 1Þðq þ 1Þ=a ꢂ ðp þ 2Þ=2g
B ¼ ½D1,3ꢄ=ðfLiOEng þ ½D1,3ꢄ þ ½D2,2ꢄÞ
1=fðp þ 1Þðq þ 1Þ=a ꢂ ðp þ 2Þ=2g
C ¼ ½D2,2ꢄ=ðfLiOEng þ ½D1,3ꢄ þ ½D2,2ꢄÞ
p=2=fðp þ 1Þðq þ 1Þ=a ꢂ ðp þ 2Þ=2g
es are calculated by assuming that HMPA-free enolate would
be absent in these HMPA quantities.
¼
=
ð9Þ
ð10Þ
ð11Þ
B + C) ꢃ 100, respectively. Thus, the calculated p and q values
and resulting relative amounts of T0,4, D0,4, D1,3, and D2,2 are sum-
marized in Table 2. Overall, the percentages of the enolate species
are plotted as functions of HMPA concentrations, as shown in
Fig. 2.
General Procedure for Kinetic Studies. (a) Determination
of the Reaction Order of the Lithium Enolate for Alkylation
(Fig. 3A): The lithium enolate was prepared in THF as described
above. To the resulting solution, HMPA (573 mg, 3.2 mmol) was
added, followed by additional THF to give a total volume of 5.0
¼
¼
Actually, the 0.16 M solution of the lithium enolate is submitted
for NMR analysis and therefore, Eq. 12 is established under such
condition.
mL. After the solution was left to stand at room temperature for
ꢁ
0
:16 M ¼ fLiOEng þ 2½D1,3ꢄ þ 2½D2,2ꢄ
which is rewritten by Eq. 13.
D1,3ꢄ þ ½D2,2ꢄ ¼ ð0:16 M ꢂ fLiOEngÞ=2
ð12Þ
1
.0 h, the solution was cooled in a CryoCool-controlled ꢂ75 C
bath. Aliquots of a stock THF solution of methyl iodide (0.4 M,
.10 mL, 0.04 mmol) containing nonane as GC standard (7.18
0
½
ð13Þ
mg, 0.056 mmol) were added via a syringe. After the mixture
had been stirred for 1.0 min, chlorotrimethylsilane (514 mg, 4.7
mmol) and triethylamine (479 mg, 4.7 mmol) were added, and then
a saturated aqueous NaHCO3 solution (1 mL) was poured into the
solution. The organic layer was separated and the aqueous layer
was extracted twice with ether (2 mL). The combined organic solu-
tions were dried over Na2SO4. The reaction mixtures were submit-
ted to GC analysis. The reactions for 1.5, 2.0, and 2.5 min were re-
peated by a similar procedure. The initial rates were measured
based on the yield of 2-methylcyclopentanone. Kinetic studies
were carried out with the change of the concentration of the lithium
enolate from 0.07 to 0.13 M as described above. The resulting rates
are summarized in Table 3. The plotting in Fig. 3A was conducted
based on these kinetic data.
(b) Determination of the Reaction Order of the Lithium
Enolate for Proton Abstraction Reaction (Fig. 3B): The lithium
enolate was prepared in THF as described above. To the solution
was added HMPA (573.4 mg, 3.2 mmol) and this was followed
by additional THF to give a total volume of 5.0 mL. After the solu-
tion was left to stand at room temperature for 1.0 h, the solution
was cooled in a CryoCool-controlled ꢂ50 ꢁC bath. A 0.5 M
THF solution of 2-methylcyclopentanone (0.10 mL, 0.05 mmol)
Substitution of Eq. 13 into Eq. 9 gives Eq. 14 in which the for-
mal concentration of HMPA-free enolate in expressed with the de-
fined number A.
fLiOEng ¼ 0:16A=ð2 ꢂ AÞ
The cyclopentanone lithium enolate is known to exist as a mix-
ture of the tetramer T0,4 and dimer D0,4 in THF.16 Assuming that
ð14Þ
the equilibrium constant for the HMPA-free enolate, K ¼
2
½
D0,4ꢄ =½T0,4ꢄ, could be adapted in a range of our enolate concen-
trations, each concentration of HMPA-free aggregates, T0,4 and
D0,4, at a given enolate concentration is determined using K
(
¼ 0:15 ꢅ 0:02 M) deduced from the melting point depression
16
in THF. Here, an average aggregation number (na; na
=
ꢁ
n2[(LiOEn)n]/ꢁn[(LiOEn)n]) is given by Eq. 15.
na ¼ ð4 ꢃ 4½T0,4ꢄ þ 2 ꢃ 2½D0,4ꢄÞ=ð4½T0,4ꢄ þ 2½D0,4ꢄÞ
ð15Þ
Thus, the actual concentration of HMPA-free aggregates is written
by {LiOEn}/na, then its proportion is A/na. According to the
above procedure, the relative amounts (%) of HMPA-free aggre-
gates (T0,4 and D0,4), D1,3, and D2,2 are determined by A/na/(A/
na + B + C) ꢃ 100, B/(A/na + B + C) ꢃ 100, and C/(A/na +

M. Suzuki et al.
Bull. Chem. Soc. Jpn., 77, No. 2 (2004)
267
Table 3. Initial Rates of Alkylation at Several Concentra-
tions
Table 5. Initial Rates of Alkylation at Several Concentra-
tions
ꢂ1
yb)
[HMPA]/M xa) (ꢂd[CH3I]/dt)/[D2,2][CH3I] (M
ꢂ1 ꢂ1
) yb)
s
[
D2,2]/M
xa)
(ꢂd[CH3I]/dt)/[CH3I] (s
)
ꢂ4
ꢂ2
0.034
0.039
0.045
0.050
0.056
0.061
0.067
ꢂ1:47
ꢂ1:41
ꢂ1:35
ꢂ1:30
ꢂ1:25
ꢂ1:21
ꢂ1:17
6:52 ꢃ 10
ꢂ3:19
ꢂ3:14
ꢂ3:10
ꢂ3:03
ꢂ2:30
ꢂ2:92
ꢂ2:88
0.360 ꢂ0:523
0.390 ꢂ0:481
0.402 ꢂ0:466
0.420 ꢂ0:444
0.450 ꢂ0:409
0.480 ꢂ0:377
0.510 ꢂ0:347
1:29 ꢃ 10
1:46 ꢃ 10
1:56 ꢃ 10
1:73 ꢃ 10
1:82 ꢃ 10
1:81 ꢃ 10
2:08 ꢃ 10
2:13 ꢃ 10
ꢂ1:89
ꢂ1:84
ꢂ1:81
ꢂ1:76
ꢂ1:74
ꢂ1:74
ꢂ1:68
ꢂ1:67
ꢂ4
ꢂ2
ꢂ2
ꢂ2
ꢂ2
ꢂ2
ꢂ2
ꢂ2
7:16 ꢃ 10
ꢂ4
7:96 ꢃ 10
ꢂ4
9:27 ꢃ 10
ꢂ3
1:03 ꢃ 10
ꢂ3
1:21 ꢃ 10
ꢂ3
1:30 ꢃ 10
0
.540 ꢂ0:319
a) x: log½D2,2ꢄ. b) y: log((ꢂd[CH3I]/dt)/[CH3I]).
a) x: log([HMPA] ꢂ 2[D2,2]). b) y: log((ꢂd[CH3I]/dt)/
[
D2,2][CH3I]).
Table 4. Initial Rates of Proton Abstraction at Several Con-
centrations
Table 6. Initial Rates of Proton Abstraction at Several Con-
centrations
ꢂ1
yb)
[
D2,2]/M
xa)
(ꢂd[4]/dt)/[4] (s
)
ꢂ5
0.028
0.034
0.039
0.045
0.056
0.059
0.067
ꢂ1:55
ꢂ1:47
ꢂ1:41
ꢂ1:35
ꢂ1:25
ꢂ1:23
ꢂ1:17
7:97 ꢃ 10
ꢂ4:10
ꢂ4:02
ꢂ3:98
ꢂ3:90
ꢂ3:81
ꢂ3:78
ꢂ3:72
[
HMPA]/M
xa)
(ꢂd[4]/dt)/[D2,2][4] (M
ꢂ1 ꢂ1
s )
yb)
ꢂ5
9:46 ꢃ 10
ꢂ4
ꢂ3
1:05 ꢃ 10
0.360
0.390
0.420
0.450
0.480
ꢂ0:523
ꢂ0:481
ꢂ0:444
ꢂ0:409
ꢂ0:377
ꢂ0:347
ꢂ0:319
2:52 ꢃ 10
ꢂ2:60
ꢂ2:61
ꢂ2:59
ꢂ2:56
ꢂ2:61
ꢂ2:61
ꢂ2:64
ꢂ4
ꢂ3
1:25 ꢃ 10
2:44 ꢃ 10
ꢂ4
ꢂ3
1:55 ꢃ 10
2:58 ꢃ 10
ꢂ4
ꢂ3
1:65 ꢃ 10
2:76 ꢃ 10
ꢂ4
ꢂ3
1:92 ꢃ 10
2:48 ꢃ 10
ꢂ3
0
0
.510
.540
2:48 ꢃ 10
a) x: log½D2,2ꢄ. b) y: log((ꢂd[4]/dt)/[4]).
ꢂ3
2:30 ꢃ 10
a) x: log([HMPA] ꢂ 2[D2,2]). b) y: log((ꢂd[4]/dt)/[D2,2][4]).
containing nonane as GC standard (7.18 mg, 0.056 mmol) was add-
ed via a syringe. After the mixture had been stirred for 4.0 min,
chlorotrimethylsilane (1.37 g, 13.0 mmol) and triethylamine
in Fig. 4A was conducted based on these kinetic data.
(d) Determination of the Reaction Order of HMPA for Pro-
ton Abstraction (Fig. 4B): Kinetic studies were carried out with
the 0.06 M lithium enolate under the change of the concentration of
HMPA from 0.36 to 0.54 M as described above. The resulting rates
are summarized in Table 6. The plotting in Fig. 4B was conducted
based on these kinetic data.
(
1.28 g, 13.0 mmol) were added, and then saturated aqueous
NaHCO3 solution (1 mL) was poured into the mixture. The organic
layer was separated, and the aqueous layer was extracted twice
with ether (2 mL). The combined organic extracts were submitted
to GC analysis. The reactions for 8.0 and 12.0 min were repeated
by a similar procedure. Initial rates were measured based on the
yield of 2-methylcyclopentanone. Kinetic studies were carried
out with the change of the concentration of the lithium enolate
from 0.06 to 0.13 M as described above. The resulting rates are
summarized in Table 4. The plotting in Fig. 3B was conducted
based on these kinetic data.
This work was supported in part by Grant-in-Aids for Crea-
tive Scientific Research 13NP0401 (M.S.) and 14GS0214
R.N.) from the Japan Society for the Promotion of Science.
We thank Dr. M. Kataoka for helpful guidance in NMR meas-
urements in Nagoya University Chemical Instrument Center.
(
(
c) Determination of the Reaction Order of HMPA for Alky-
lation (Fig. 4A): The lithium enolate (0.3 mmol, 27.0 mg) was
prepared in THF (1.5 mL) as described above. To the resulting so-
lution was added HMPA (322 mg, 1.8 mmol), and followed by ad-
ditional THF to give a total volume of 5.0 mL. After the solution
was left to stand at room temperature for 1.0 h, the solution was
References
1
a) R. L. Augustine, ‘‘Carbon–Carbon Bond Formation,’’
Marcel Dekker, New York (1979), Vol. 1. b) C. H. Heathcock,
‘‘Comprehensive Organic Synthesis,’’ ed by B. M. Trost and I.
Fleming, Pergamon Press, Oxford, U. K. (1991), Vol. 3. c) D. A.
Evans, ‘‘Asymmetric Synthesis,’’ ed by J. D. Morrison, Academic
Press, New York (1984), Vol. 3, pp. 1–110.
ꢁ
cooled in a CryoCool-controlled ꢂ75 C bath. A 0.4 M THF solu-
tion of methyl iodide (0.10 mL, 0.04 mmol) containing nonane as
GC standard (7.18 mg, 0.056 mmol) was added via a syringe. After
the mixture had been stirred for 5.0 min, chlorotrimethylsilane
(
514 mg, 4.7 mmol) and triethylamine (479 mg, 4.7 mmol) were
2
a) A. Abbotto and A. Streitwieser, J. Am. Chem. Soc.,
added, and then a saturated aqueous NaHCO3 solution (1 mL)
was poured into the mixture. The organic layer was separated,
and the aqueous layer was extracted twice with ether (2 mL).
The combined organic extracts were analyzed by GC. The reac-
tions for 10.0 and 15.0 min were repeated by the same procedure.
Initial rates were measured based on the yield of 2-methylcyclo-
pentanone. Kinetic studies were carried out with the change of
the concentration of HMPA from 0.36 to 0.54 M as described
above. The resulting rates are summarized in Table 5. The plotting
117, 6358 (1995). b) F. Abu-Hasanayn, M. Stratakis, and A.
Streitwieser, J. Org. Chem., 60, 4688 (1995). c) A. Abbotto, S.
S.-W. Leung, A. Streitwieser, and K. V. Kilway, J. Am. Chem.
Soc., 120, 10807 (1998). d) A. Streitwieser and D. Z.-R. Wang,
J. Am. Chem. Soc., 121, 6213 (1999). e) S. S.-W. Leung and A.
Streitwieser, J. Org. Chem., 64, 3390 (1999). f) D. Z. Wang,
Y.-J. Kim, and A. Streitwieser, J. Am. Chem. Soc., 122, 10754
(2000).
3
Review: D. Seebach, Angew. Chem., Int. Ed. Engl., 27,

2
68 Bull. Chem. Soc. Jpn., 77, No. 2 (2004)
HEADLINE ARTICLES
rameric and dimeric enolates in acyclic ethereal solvents, see: M.
1
624 (1988).
4
R. Amstutz, W. B. Schweizer, D. Seebach, and J. D. Dunitz,
Helv. Chim. Acta, 64, 2617 (1981).
a) L. M. Jackman and N. M. Szeverenyi, J. Am. Chem. Soc.,
9, 4954 (1977). b) L. M. Jackman and C. W. DeBrosse, J. Am.
Suzuki, H. Koyama, and R. Noyori, Tetrahedron (2004), in press.
20 Some of the results were reported in Abstracts; M. Suzuki,
H. Koyama, and R. Noyori, ‘‘48th Symposium on Organometallic
Chemistry, Japan,’’ Kinki Chemical Society, Yokohama National
University, Kanagawa, Japan (2001), pp. 114–115; M. Suzuki,
H. Koyama, and R. Noyori, ‘‘50th Symposium on Organometallic
Chemistry, Japan,’’ Kinki Chemical Society, Osaka University,
Osaka, Japan (2003), pp. 44–45.
5
9
Chem. Soc., 105, 4177 (1983). c) L. M. Jackman and X. Chen, J.
Am. Chem. Soc., 114, 403 (1992).
6
a) M. Suzuki, A. Yanagisawa, and R. Noyori, J. Am. Chem.
Soc., 107, 3348 (1985). b) M. Suzuki, A. Yanagisawa, and R.
Noyori, J. Am. Chem. Soc., 110, 4718 (1988). c) Y. Morita, M.
Suzuki, and R. Noyori, J. Org. Chem., 54, 1785 (1989). d) M.
Suzuki, Y. Morita, H. Koyano, M. Koga, and R. Noyori, Tetrahe-
dron, 46, 4809 (1990).
21 J. L. Rutherford and D. B. Collum, J. Am. Chem. Soc., 121,
10198 (1999); J. L. Rutherford and D. B. Collum, J. Am. Chem.
Soc., 123, 199 (2001).
7
22 A doublet signal of Li in Fig. 1 Ah became broad singlet
ꢁ
7
a) P. A. Tardella, Tetrahedron Lett., 10, 1117 (1969). b) H.
Nishiyama, K. Sakuta, and K. Itoh, Tetrahedron Lett., 25, 223
1984). c) H. Nishiyama, K. Sakuta, and K. Itoh, Tetrahedron Lett.,
5, 2487 (1984).
For HMPA effects on the structures of organolithium com-
pounds, see: a) H. J. Reich and D. P. Green, J. Am. Chem. Soc.,
11, 8729 (1989). b) H. J. Reich, D. P. Green, M. A. Medina,
signal at ꢂ50 C arising from rapid ligand exchange between
THF and HMPA.
23 The spectra no longer changed with increasing HMPA
(8 molar amounts).
(
2
8
24 The degree of coordination ability was determined by the
ratio of free and coordinated HMPA in the presence of equimolar
amount of HMPA. For the solvation energy, see: H. J. Reich and K.
J. Kulicke, J. Am. Chem. Soc., 118, 273 (1996).
25 a) L. M. Jackman, L. M. Scarmoutzos, and C. W. DeBrosse,
J. Am. Chem. Soc., 109, 5355 (1987). b) L. M. Jackman and D.
C¸ izmeciyan, Magn. Reson. Chem., 34, 14 (1996).
26 J. Q. Wen and J. B. Grutzner, J. Org. Chem., 51, 4220
(1986).
1
¨
W. S. Goldenberg, B. O. Gudmundsson, R. R. Dykstra, and N.
H. Phillips, J. Am. Chem. Soc., 120, 7201 (1998). c) H. J. Reich
and W. H. Sikorski, J. Org. Chem., 64, 14 (1999). d) W. H.
Sikorski and H. J. Reich, J. Am. Chem. Soc., 123, 6527 (2001).
9
L. M. Jackman and B. C. Lange, J. Am. Chem. Soc., 103,
494 (1981).
L. M. Jackman and B. D. Smith, J. Am. Chem. Soc., 110,
829 (1988).
4
3
1
0
27 D2,2 is formed exclusively under these conditions.
28 QSC values in THF for ꢀ-tetralone and acetaldehyde26
25a
1
1991).
1
E. M. Arnett and K. D. Moe, J. Am. Chem. Soc., 113, 7288
lithium enolate with tetrameric structures are 64 and 65 kHz at
ꢁ
ꢁ
(
30 C and ꢂ1 C, respectively, while the value of tetrameric iso-
butyrophenone lithium enolate5 is 136 kHz.
a
1
2
For aggregated reactants such as those proposed for aldol
reactions, see: a) D. Seebach, R. Amstutz, and J. D. Dunitz, Helv.
Chim. Acta, 64, 2622 (1981). b) C. H. Heathcock and J. Lampe, J.
Org. Chem., 48, 4330 (1983). c) P. G. Williard and M. J. Hintze, J.
Am. Chem. Soc., 109, 5539 (1987). d) E. M. Arnett, F. J. Fisher, M.
A. Nichols, and A. A. Ribeiro, J. Am. Chem. Soc., 112, 801 (1990).
29 Spectra of other lithium enolates such as p-phenylisobu-
tyrophenonate,2 pinacolonate,4 and isobutyrophenonate,5a were
complicated and difficult to analyze.
30 Actual alkylation reactions in the absence of HMPA were
ꢁ
conducted at temperatures from 0 to ꢂ50 C. Therefore, the rate
ꢁ
1
1
3
4
D. B. Collum, Acc. Chem. Res., 26, 227 (1993).
a) H. J. Reich and J. P. Borst, J. Am. Chem. Soc., 113, 1835
at ꢂ75 C was deduced from the equation of 8:51 ꢂ 4:94 ꢃ
3
10 =T obtained by the Arrhenius plot (ln{(d[CH3I]/dt)/[CH3I]}
against 1=T).
(
1991). b) H. J. Reich, J. P. Borst, R. R. Dykstra, and D. P. Green, J.
Am. Chem. Soc., 115, 8728 (1993).
A. Streitwieser, Y.-J. Kim, and D. Z.-R. Wang, Org. Lett.,
, 2599 (2001).
Although the lithium enolate is often generated by the reac-
31 X. Sun and D. B. Collum, J. Am. Chem. Soc., 122, 2452
(2000).
1
5
3
32 The same dimer D2,2 is involved in the reaction in a wide
ꢁ
1
6
range of temperature including ꢂ50 C as judged by Arrhenius
3
tion of a ketonic substrate with lithium diisopropylamide in an
ethereal solvent, we used an amine-free pure enolate in this study
to avoid the structural complication arising from the participation
of the amine ligand.
equation (lnfðd½4ꢄ=dtÞ=½4ꢄg ¼ 5:98 ꢂ 1:91 ꢃ 10 =T) obtained
ꢁ
22
from the reaction conducted at the range of ꢂ20 to ꢂ75 C.
33 At free HMPA concentrations lower than 0.30 M in THF,
alkylation rates are not linear, due to incomplete convergence on
the dimer.
1
1984).
7
W. Bauer and D. Seebach, Helv. Chim. Acta, 67, 1972
(
34 a) E. Kaufmann, P. v. R. Schleyer, K. N. Houk, and Y.-D.
Wu, J. Am. Chem. Soc., 107, 5560 (1985). b) M. P. Bernstein
and D. B. Collum, J. Am. Chem. Soc., 115, 789 (1993).
35 F. E. Romesberg and D. B. Collum, J. Am. Chem. Soc., 117,
2166 (1995), and references cited therein.
1
8
The coordination number of lithium cation is defined as
four, according to general empirical formula. This consistently
simplifies the following structural analyses and discussion.
1
9
For the details including the formation of a mixture of tet-