On the mechanism of addition of lithium pinacolone enolate to benzaldehyde: Polar or electron transfer?

Article abstract of DOI:10.1021/ja971262c

The carbonyl-carbon kinetic isotope effect (KIE) and the substituent effect were measured for the reaction of lithium pinacolone enolate (CH₂=C(OLi)C(CH₃)₃) with benzaldehyde, and the results were compared with those for other lithium reagents such as MeLi, PhLi, and CH₂=CHCH₂Li. Ab initio MO calculations (HF/6-31+G*) were carried out to estimate the equilibrium IE on the addition to benzaldehyde. A carbonyl addition reaction, in general, proceeds by way of either a polar direct nucleophilic attack (PL) in a one-step or a two-step process going through a radical ion intermediate (eq 1). The carbonyl-carbon KIE is of primary nature for the PL or the RC rate-determining ET mechanism, while it is considered to be secondary for the ET rate-determining mechanism. The reaction of lithium pinacolone enolate with benzaldehyde gave a small positive KIE (¹²k/¹³k = 1.019), which is larger than the theoretical equilibrium IE (¹²K/¹³K = 1.006) determined by the MO calculations. Thus, there is a reaction-coordinate contribution to the observed KIE. This is in sharp contrast to the absence of KIE (¹²k/¹⁴k = 1.000) measured previously for the MeLi addition. Dehalogenation and enone-isomerization probe experiments showed no evidence of a single electron transfer to occur during the course of the reaction. The primary carbonyl-carbon KIE together with the substituent effect and chemical probe experiments led to the conclusion that the reaction of lithium pinacolone enolate with benzaldehyde proceeds via the polar mechanism.

Full text of DOI:10.1021/ja971262c

J. Am. Chem. Soc. 1997, 119, 9975-9979
9975
On the Mechanism of Addition of Lithium Pinacolone Enolate
to Benzaldehyde: Polar or Electron Transfer?
Hiroshi Yamataka,*,§ Daizo Sasaki,† Yoshiyuki Kuwatani,‡
Masaaki Mishima,*,† and Yuho Tsuno†
Contribution from the Institute of Scientific and Industrial Research, Osaka UniVersity,
Ibaraki, Osaka 567, Institute for Fundamental Research of Organic Chemistry,
Kyushu UniVersity, Higashi-ku, Fukuoka, 812-81, and Department of Chemistry,
Faculty of Science, Tokyo Metropolitan UniVersity, Hachioji, Tokyo 192-03, Japan
ReceiVed April 21, 1997X
Abstract: The carbonyl-carbon kinetic isotope effect (KIE) and the substituent effect were measured for the reaction
of lithium pinacolone enolate (CH2dC(OLi)C(CH3)3) with benzaldehyde, and the results were compared with those
for other lithium reagents such as MeLi, PhLi, and CH2dCHCH2Li. Ab initio MO calculations (HF/6-31+G*)
were carried out to estimate the equilibrium IE on the addition to benzaldehyde. A carbonyl addition reaction, in
general, proceeds by way of either a polar direct nucleophilic attack (PL) in a one-step or a two-step process going
through a radical ion intermediate (eq 1). The carbonyl-carbon KIE is of primary nature for the PL or the RC
rate-determining ET mechanism, while it is considered to be secondary for the ET rate-determining mechanism.
The reaction of lithium pinacolone enolate with benzaldehyde gave a small positive KIE (12k/13k ) 1.019), which is
larger than the theoretical equilibrium IE (12K/13K ) 1.006) determined by the MO calculations. Thus, there is a
reaction-coordinate contribution to the observed KIE. This is in sharp contrast to the absence of KIE (12k/14k )
1.000) measured previously for the MeLi addition. Dehalogenation and enone-isomerization probe experiments
showed no evidence of a single electron transfer to occur during the course of the reaction. The primary carbonyl-
carbon KIE together with the substituent effect and chemical probe experiments led to the conclusion that the reaction
of lithium pinacolone enolate with benzaldehyde proceeds via the polar mechanism.
Introduction
transfer during the reaction by observing an EPR spectrum and
by analyzing kinetic behavior of the spectrum.3
Lithium enolate is an important reagent in its versatile
synthetic utility, especially for enantioselective reactions. De-
sign of the reaction, which is critical to attain high selectivity,
requires the knowledge of the mechanistic details. However,
understanding of the mechanism of this fundamental transfor-
mation is still incomplete. The well-known tendency of the
organolithium reagent to form aggregates1 is a source of
difficulty in understanding the mechanism, but a more funda-
mental question remained to be solved is the course of the
reaction.
There are in general two possible reaction routes, the polar
addition (PL) mechanism and the electron transfer (ET) - radical
coupling (RC) sequence, eq (1), in nucleophilic addition
reactions, and the mechanism is highly nucleophile dependent.
Recently Arnett reported that the reaction of lithium pinacolone
enolate with benzaldehyde proceeds via the polar mechanism
on the basis of the comparison of the experimental activation
energy with the energy required for ET determined electro-
chemically.1a,2 The conclusion was different from that for the
reaction of the same nucleophile with benzophenone drawn by
Ashby, who clearly indicated the occurrence of an electron
We have previously reported that mechanistic criteria such
as kinetic isotope effect (KIE), substituent effect, and chemical
probe experiments are useful in clarifying the reaction pathway
of nucleophilic addition reactions of RLi, RMgX, and Wittig
reagents.4-6 For example, the addition reactions of MeLi, PhLi,
and CH2dCHCH2Li with benzophenone and benzaldehyde gave
in all cases a very small carbonyl-carbon KIE and a small
Hammett F value, except for the reaction of highly hindered
mesityl phenyl ketone. This suggested that most of these
reactions proceed via a rate-determining ET mechanism. Mesi-
tyl phenyl ketone reacted with MeLi much slower than other
† Kyushu University.
‡ Tokyo Metropolitan University.
§ Osaka University.
(3) Ashby, E. C.; Agyropoulos, J. N. J. Org. Chem. 1986, 51, 472.
(4) Yamataka, H.; Matsuyama, T.; Hanafusa, T. J. Am. Chem. Soc. 1989,
111, 4912.
X Abstract published in AdVance ACS Abstracts, October 1, 1997.
(1) For lithium enolate aggregation, see: (a) Arnett. E. M.; Palmer, C.
A. J. Am. Chem. Soc. 1990, 112, 7354. (b) Seebach, D. Angew. Chem., Int.
Ed. Engl. 1988, 27, 1624. (c) Arnett, E. M.; Fisher, F. J.; Nichols, M. A.;
Ribeiro, A. A. J. Am. Chem. Soc. 1990, 112, 801. (d) Seebach, D.; Amstutz,
R.; Dunitz, D. HelV. Chim. Acta 1981, 64, 2622. (e) Jackman, L. M.;
Szeverenyi, N. M. J. Am. Chem. Soc. 1977, 99, 4954. (f) Jackman, L. M.;
Scarmoutzos, L. M.; DeBrosse, C. W. J. Am. Chem. Soc. 1987, 109, 5355.
(2) Palmer, C. A.; Ogle, C. A.; Arnett, E. M. J. Am. Chem. Soc. 1992,
114, 5619.
(5) (a) Yamataka, H.; Fujimura, N.; Kawafuji, Y.; Hanafusa, T. J. Am.
Chem. Soc. 1987, 109, 4305. (b) Yamataka, H.; Kawafuji, Y.; Nagareda,
K.; Miyano, N.; Hanafusa, T. J. Org. Chem. 1988, 54, 4706.
(6) (a) Yamataka, H.; Nagareda, K.; Hanafusa, T.; Nagase, S. Tetrahedron
Lett. 1989, 30, 7187. (b) Yamataka, H.; Nagareda, K.; Ando, K.; Hanafusa,
T. J. Org. Chem. 1992, 57, 2865. (c) Yamataka, H.; Nagareda, K.;
Takatsuka, T.; Ando, K.; Hanafusa, T.; Nagase, S. J. Am. Chem. Soc. 1993,
115, 8570.
S0002-7863(97)01262-6 CCC: $14.00 © 1997 American Chemical Society

9976 J. Am. Chem. Soc., Vol. 119, No. 42, 1997
Yamataka et al.
Table 1. Kinetic and Equilibrium Isotope Effects for the Reaction
of Lithium Pinacolone Enolate and Benzaldehyde
substituted benzophenone derivatives and showed a sizable
carbonyl-carbon KIE, and the results were ascribed to a slower
RC step for this sterically hindered ketone and hence a shift of
rate-determining step from ET (for benzophenone) to RC (for
mesityl phenyl ketone).5 In sharp contrast, the reaction of
RMgX gave medium sized KIEs in most cases, and this was
concluded to be due to rate-determining RC mechanism.4
Our conclusion that the reaction of MeLi, PhLi, and
CH2dCHCH2Li with the aromatic carbonyl compounds proceed
via an ET process was different from that by Arnett for the
reaction of lithium pinacolone enolate. The discrepancy should
be due to either the difference of reagents or the difference of
the mechanistic criteria. In the present study, the KIE, sub-
stituent effect, and chemical probe experiments were carried
out for the addition reaction of lithium pinacolone enolate with
benzaldehyde. The results clearly show that the mechanism
for the enolate is different to other RLi and is polar.
labeled position
k/k* a
K/K* b
kH/kD5
1.031 ( 0.004
0.988 ( 0.002
1.019 ( 0.004c
1.094
13
k / k
D5
12k/13k
1.006
a Experimental kinetic isotope effect in THF at 0 °C. b Calculated
equilibrium isotope effect on the addition process at 0 °C. c Calcd from
kH/kD5 and kD5/13k.
carbonyl carbon KIE is expected to be large for the PL and the
rate-determining RC mechanism because the carbonyl carbon
is heavily involved in these C-C bond forming processes. On
the other hand, the carbon KIE is expected to be small for the
rate-determining ET mechanism since there is no contribution
from the reaction-coordinate motion. Distinction between PL
and rate-determining RC is more difficult, and kinetic methods
like KIE are almost ineffective because the two transition states
behave in a similar manner in kinetic criteria. Another way to
differentiate these two mechanisms is needed which examines
the possible intervention of a radical ion-pair intermediate in
the ET-RC sequence, and this will be discussed later.
Isotope effects were determined for C6D5CHO vs C6H5CHO
and C6H513CHO vs C6D5CHO in order to avoid the interference
of 13C natural abundance in intensity measurement. The
observed KIEs at 0 °C were listed in Table 1. In discussing
the observed KIEs, it is useful to know the magnitudes of IEs
on the corresponding equilibria. For this purpose, theoretical
isotope effects on the addition equilibrium were computed by
using ab initio MO methods. Thus, benzaldehyde and benzal-
dehyde-methyl lithium adduct were fully optimized at HF/6-
31+G*, and their vibrational frequencies were calculated. MeLi
was used here instead of lithium pinacolone enolate for
simplicity assuming that the equilibrium IEs are not much
different for the MeLi addition and the lithium enolate addition.
Equilibrium IEs for carbonyl-13C, aldehyde-d1, and phenyl-d5
species were computed from the frequencies of isotopomers by
using eq 2 and listed in Table 1. The HF frequencies were
scaled down by a factor of 0.89 in the IE calculations.
Results and Discussion
Lithium pinacolone enolate was prepared from the reaction
of pinacolone and lithium diisopropylamide in THF at -78 °C
and allowed to react with benzaldehyde at 0 °C. Substituent
effects were determined by competition experiments as previ-
ously reported,4-6 and carbonyl carbon-13 and C6D5 KIEs were
measured as described in the Experimental Section. We used
in the present study FT ion-cyclotron resonance mass spec-
trometry (FT-ICR MS) to determining the carbon KIE of 13C-
labeled compound rather than the conventional liquid scintil-
lation counting method of radioactive 14C compounds. This
new method can be a powerful tool in mechanistic study by
allowing us to measure KIEs of nonradioactive elements, such
as 18O, 15N, D, etc. in addition to 13C. The FT-ICR method
has previously been used to determine equilibrium IEs success-
fully.10 Two chemical probe experiments, enone isomerization
and dehalogenation probes, were carried out as described
previously.6c,7 Ab initio MO calculations were carried out to
estimate the magnitude of equilibrium IEs for the polar process
by using the Gaussian 92 and 94 packages of program.8,9
Carbon KIE. Among the three possible mechanistic alterna-
tives, distinction of the rate-determining ET mechanism from
PL and rate-determining RC can be achieved in a rather
straightforward manner by using the carbonyl-carbon KIE as a
probe. A carbon KIE is larger if the labeled carbon atom is
involved to a greater extent in the reaction-coordinate vibrational
motion at the rate-determining transition state,11,12 and thus the
1 - e-u
1 - e-u
3n-6
3n-6
1i
∏ u2i ∏
3n-6
k1
k2
νqL1
u1i
(
(
)
)
exp[
(u - u )/2]
1i 2i
2i
∏
)
(2)
q
q
3n-7
q
[ ]
1 - e-u
1 - e-u
3n-7
3n-7
1i
∏ u2i ∏
νL2
exp[
(uq - uq )/2]
∏
1i
2i
q
uq1i
2i
(7) Yamataka, H.; Yamaguchi, K.; Takatsuka, T.; Hanafusa, T. Bull.
Chem. Soc. Jpn. 1992, 65, 1157.
The experimental carbonyl-13C KIE was small positive (12k/
13k ) 1.019), which corresponds to 1.039 of 14C KIE. This
value is similar to carbonyl-14C KIEs previously observed for
polar nucleophilic additions to benzophenone; 12k/14k ) 1.035
(BH3/THF), 1.021 (AlH3/Et2O), LiAlH4/THF (1.017), LiAlH4/
Et2O (1.024), LiBH4/Et2O (1.043), and NaBH4/2-PrOH (1.066).13
The theoretical equilibrium IE at carbonyl 13C for the addition
to benzaldehyde was found to be small (12k/13k ) 1.006). The
fact that the experimental carbonyl-13C KIE is larger than the
corresponding equilibrium IE indicates that the carbonyl-carbon
KIE is of primary nature, suggesting that the bonding to the
carbon is changing at the rate-determining transition state. The
IE results are consistent with the PL mechanism for the addition
reaction, although the IE results by themselves do not eliminate
the possibility that the reaction proceeds via fast ET followed
by a slow rate-determining RC step.
(8) Frisch, M. J.; Trucks, G. W.; Head-Gordon, M.; Gill, P. M.; Wong,
M. W.; Foresman, J. B.; Johnson, B. G.; Schlegel, H. B.; Robb, M. A.;
Replogle, E. S.; Gomperts, R.; Andress, J. L.; Rachavachari, K.; Binkley,
J. S.; Gonzalez, C.; Martin, R. L.; Fox, D. J.; Defrees, D. J.; Baker, J.;
Stewart, J. J. P.; Pople, J. A. GAUSSIAN 92; Gaussian Inc.: Pittsburgh,
PA, 1992.
(9) Frisch, M. J.; Trucks, G. W.; Schlegel, H. B.; Gill, P. M. W.; Johnson,
B. G.; Robb, M. A. ; Cheeseman, J. R.; Keith, T.; Petersson, G. A.;
Montgomery, J. A.; Raghavachari, K.; Al-Laham, M. A.; Zakrzewski, V.
G.; Ortiz, J. V.; Foresman, J. B.; Cioslowski, J.; Stefanov, B. B.;
Nanayakkara, A.; Challacombe, M.; Peng, C. Y.; Ayala, P. Y.; Chen, W.;
Wong, M. W.; Andress, J. L.; Replogle, E. S.; Gomperts, R.; Martin, R.
L.; Fox, D. J.; Binkley, J. S.; Defrees, D. J.; Baker, J.; Stewart, J. P.; Head-
Gordon, M.; Gonzalez, C.; Pople, J. A. GAUSSIAN 94, Revision C2;
Gaussian, Inc.: Pittsburgh, PA, 1995.
(10) Yamataka, H.; Mishima, M.; Kuwatani, Y.; Tsuno, Y. J. Am. Chem.
Soc. 1995, 117, 5829.
(11) Melander, L.; Saunders, W. H., Jr. Reaction Rates of Isotopic
Molecules; Wiley-Interscience: New York, 1980.
(12) Sims, L. B.; Lewis, D. E. In Isotopes in Organic Chemistry; Buncel,
E., Lee, C. C., Eds.; Elsevier: Amsterdam, 1984; Vol. VI, Chapter 4.
Yamataka, H.; Ando, T. J. Phys. Chem. 1981, 85, 2281. Yamataka, H.;
Tamura, S.; Hanafusa, T.; Ando, T. J. Am. Chem. Soc. 1985, 107, 5429.
(13) Yamataka, H.: Hanafusa, T. J. Am. Chem. Soc. 1986, 108, 6643.
Yamataka, H.; Hanafusa, T. J. Org. Chem. 1988, 53, 772.

Addition of Lithium Pinacolone Enolate to Benzaldehyde
J. Am. Chem. Soc., Vol. 119, No. 42, 1997 9977
Table 2. Z-E Isomerization during the Reaction of Z-Enone with
Lithium Pinacolone Enolate in THF at 0 °C
reagent
enolate
time, min
recovered Z:E
1
60
120
60
99.6:0.4
98.7:1.3
97.9:2.1
99.8:0.2
blank
it is important to consult with other experimental criteria to
distinguish the PL mechanism from the RC mechanism in which
the radical coupling process following a fast ET step is rate
determining. One way to determine whether the reaction
proceeds via the PL or RC mechanism is to detect the presence
of radical ion pair intermediate during the reaction. Two
chemical probe experiments, dehalogenation7 and enone isomer-
ization,14 have been shown to be useful in detecting the
intermediate for the additions of Grignard and Wittig reagents.6,7
These are probes that measure whether a reagent has enough
ability to transfer an electron to halobenzophenone and enone,
respectively, and therefore, if these two probes were positive,
then there would be a good possibility that the reaction proceeds
via the ET-RC sequence. The basis idea of these probes were
discussed previously in detail.6c In the dehalogenation probe
experiment, o-iodobenzophenone was mixed with lithium pi-
nacolone enolate in THF at 0 °C, and after appropriate reaction
time the reaction mixture was analyzed by GC. The reaction
gave no dehalogenated product and only the starting o-
iodobenzophenone was recovered, and thus the dehalogenation
probe was negative for the lithium enolate. In the enone
isomerization, (Z)-2,2,6,6-tetramethylhept-4-ene-3-one was mixed
with lithium pinacolone enolate in THF at 0 °C, and the reaction
mixture was analyzed by GC. As listed in Table 2, slight enone
isomerization was detected only for a prolonged reaction time,
and thus the enone-isomerization probe was negative for the
lithium enolate. In summary, these two probe experiments
suggest that ET is not involved during the reaction of lithium
pinacolone enolate with benzaldehyde.
Mechanistic Consideration. On the basis of the present
experiment, we could conclude that the reaction of lithium
pinacolone enolate with benzaldehyde proceeds via the polar
mechanism. In Table 3 were summarized the results of
mechanistic study on the addition reactions of several organo-
lithium reagents with benzaldehyde and benzophenone. There
can be seen a clear mechanistic gap between the lithium enolate
and the other lithium reagents. It is worthy to note that the
mechanistic difference correlates with the stability of the
reagents measured by the intrinsic acidity of the conjugate acids
of R anions in the gas phase. The intrinsic acidity of R-H can
be estimated from the gas phase heat of formation (eq 3) from
the data compiled in the literature:15 416.8 kcal/mol (R ) Me),
400.8 (Ph), 390.7 (CH2dCHCH2), and 368.0 ((CH3)3CC(dO)-
CH2). The enolate reagent whose conjugate acid is the most
Figure 1. Variations of reactivity with σ values for the reaction of
lithium pinacolone enolate and substituted benzaldehydes in THF at 0
°C.
The magnitude of the carbonyl carbon KIE for the present
reaction is clearly different from those observed for nucleophilic
additions to benzophenone of other RLi reagents; all the
reactions of MeLi, PhLi, and CH2dCHCH2Li with benzophe-
none as well as the reaction of PhLi with benzaldehyde gave
very small carbonyl carbon KIEs, from which the rate-
determining ET mechanism was concluded.
Since D5 KIE is a secondary effect, the contribution from
reaction-coordinate vibration to the IE is negligible or very small
if any. The magnitude of the KIE should thus be in between
1.0 and the equilibrium IE (1.094), depending on the position
of the TS along the reaction coordinate. The observed D5 KIE
of 1.032 is one-third of the equilibrium IE, which suggests that
the TS is reached at about one-third along the reaction coordinate
if the reaction proceeds via the PL mechanism.
Substituent Effects. Substituent effects on the reactivity of
substituted benzaldehydes with lithium pinacolone enolate were
measured in THF at 0 °C by competition experiment. Figure 1
shows the relationship between log(kX/kH) and substituent
constant σ. Since the σ constants of the ortho substituents were
not available, the log(kX/kH) values for the ortho derivatives were
plotted against the corresponding para-substituent constants and
are indicated by closed circles, and the F value was calculated
from the meta- and para-substituted derivatives. The observed
medium-sized F value (1.16 ( 0.31) indicates that the substituent
electronic effect is significant in the reaction. The results are
in sharp contrast to those observed for the reactions of MeLi,
PhLi and CH2dCHCH2Li, in which much smaller F values were
observed. Thus, the character of the rate-determining TS is
different for the lithium enolate and for MeLi, PhLi, and
CH2dCHCH2Li.
If the addition reaction proceeds via the PL mechanism, the
structure of benzaldehyde should vary monotonically on going
from the reactant state to the product state. The ratio of the
Hammett F value on rate to that on equilibrium may be taken
as a measure of the position of the TS along the reaction
coordinate. Arnett determined the substituent effects on the heat
of reaction for the addition equilibrium of the lithium enolate
and para-substituted benzaldehydes in THF by the thermo-
chemical method.1a Assuming that the entropy contribution to
the equilibrium is the same for a family of substituted benzal-
dehydes, the reported value gives F ) 2.70 on the log(k/k0) scale
at 0 °C, which allows us to estimate the position of the TS of
the present reaction as roughly 40% (1.16/2.70). This estimation
is in fair agreement with the result based on the D5 IE.
Probe Experiments. Although the results of KIEs and
substituent effects are fully consistent with the PL mechanism,
∆Hgas
R-H y z R- + H+
(3)
acidic reacts with benzaldehyde via the polar mechanism
whereas the reagents whose conjugate acids are less acidic go
through the ET process. It is thus implied that the rate of the
ET step depends more on the reactivity of a nucleophile than
the rate of the PL step as schematically shown in Figure 2 and
that the crossing point is in between allyllithium and lithium
pinacolone enolate for the reaction with benzaldehyde.
(14) House, H. O.; Weeko, P. D. J. Am. Chem. Soc. 1975, 97, 2770.
Ashby, E. C.; Wiesemann, T. L. J. Am. Chem. Soc. 1978, 100, 3101.
(15) Lias, S. G.; Bartmess, J. E.; Holmes, J. L.; Levin, R. D.; Mallard,
G. W. J. Phys. Chem., Ref. Data 1988, 17, Supple. 1.

9978 J. Am. Chem. Soc., Vol. 119, No. 42, 1997
Yamataka et al.
Table 3. Mechanism of the Reactions of Organolithium Reagents with Aromatic Carbonyl Compoundsa
Li regent
substrate
carbonyl-carbon KIEb
Hammett F value
dehalogentn
(yes)c
enone isomeriztn
no
rate determnng step
MeLi
PhLi
PhLi
AllylLi
Li enolate
benzophenone
benzophenone
benzaldehyde
benzophenone
benzaldehyde
1.000 ( 0.002
1.003 ( 0.001
0.998 ( 0.003
0.994 ( 0.003
1.039 ( 0.009d
0.23 ( 0.09
0.19 ( 0.08
0.13 ( 0.07
0.21 ( 0.09
1.16 ( 0.31
ET
ET
ET
ET
PL
no
no
b
a In THF at 0 °C. Data for MeLi, PhLi, and AllylLi were taken from ref 5. 12k/14k. c Complex mixture due to radical reactions. d Corrected to
12k/14k.
solution was prepared by mixing an equimolar amount of sodium
diisopropylamide and pinacolone in THF at -78 °C. The reactions of
the enolate with substituted benzaldehydes gave the corresponding
carbonyl adducts as the only product. The products were isolated by
TLC and characterized by 1H NMR (JEOL JNM-EX 400, CDCl3).
Material balance of the reaction was confirmed for the parent
benzaldehyde to be excellent (99.9 ( 0.9%).
Relative Reactivity Measurement. A pair of benzaldehydes
(normally the parent and a substituted one, 0.5 mmol each) and
1-methylnaphthalene (ca 0.1 mmol, internal standard) were placed in
a dry, serum-capped test tube and dissolved in 3.0 mL of dry THF.
The solution was divided to two parts; thus, each test tube contained
two kinds of benzaldehydes, 0.25 mmol each. To one test tube out of
the two was added 1.0 mL of a lithium pinacolone enolate solution
(0.33 M in THF, 0.33 mmol) at 0 °C by means of hypodermic syringe.
The solution was then allowed to react for 20 s. The reaction mixture
was treated with aqueous NH4Cl, extracted with ether, dried over
MgSO4, and subjected to HPLC analysis (P-18, CH3CN:H2O ) 7:3,
detected at 250 nm). The relative intensity of each reactant to that of
the internal standard was compared to the corresponding relative
intensity from the solution to which the lithium enolate solution was
not added. The fractions of reaction (f) were calculated for both
reactants, and the reactivity ratio was computed according to the
following equation
Figure 2. Schematic picture of the dependence of the rates of the PL
and ET processes on the nucleophile reactivity.
The EPR study by Ashby indicated that the ET can occur
between lithium pinacolone enolate and benzophenone in THF.
Since the reduction potential of benzaldehyde is similar to that
of benzophenone, one might expect that ET is also involved
during the reaction of the enolate with benzaldehyde. However,
the rate of ET to benzophenone was reported to be very slow,
taking about 20 h at 25 °C for the EPR active species to be the
maximum concentration.3 In contrast, the reaction of the enolate
with benzaldehyde is completed in 20 s. Thus, the polar
addition process is much faster than the potential ET process
for the reaction of the enolate with benzaldehyde.
We have studied the reaction of the enolate with benzophe-
none by the same criteria as for benzaldehyde. However, it
was unsuccessful because the reaction with benzophenone did
not go to completion due apparently to the occurrence of the
backward reaction under the reaction conditions. In order to
check whether the backward reaction indeed occurred for the
reaction with benzophenone, the isolated addition product of
benzophenone (2,2-dimethyl-5-hydroxy-5-phenyl-3-pentanone)
was mixed with n-BuLi to generate lithium alkoxide. On
standing for several minutes the solution generated benzophe-
none indicating the presence of the backward reaction, which
deteriorates the competition experiment. Apparently methods
other than the competition experiment is needed to investigate
the reaction of the lithium enolate with benzophenone.
kA/kB ) log(1 - fA)/log(1 - fB)
Carbon-13 KIE Measurement. The isotope effects were deter-
mined for C6D5CHO vs C6H5CHO and C6H513CHO vs C6D5CHO in
order to avoid the interference of 13C natural abundance in intensity
measurement. Isotopically labeled and unlabeled benzaldehyde (2.0
mmol each) and 1-methylnaphthalene (0.14 mmol) were dissolved dry
THF (7.6 mL), and the solution was divided into eight parts. To each
solution was added at 0 °C a preset amount of lithium pinacolone
enolate (0.33 M in THF) in such a manner that the fraction of reaction
varies from 20% to 80%. After 20 s, the reaction solution was
hydrolyzed with NH4Cl, and the organic layer was subjected to HPLC
in order to determine the fraction of reaction. The organic layer was
then transferred to a sample inlet tube, and the volatile materials were
removed under reduced pressure. The isotopic ratio of the remaining
recovered benzaldehyde was analyzed as follows.
The relative abundance of carbon-13 and -12 was measured on an
Extrel FTMS 2001 Fourier transform mass spectrometer equipped with
a 3.0 T superconducting magnet. To reduce experimental errors, all
operating parameters to affect the relative peak heights were kept
constant as far as possible. All experiments were conducted with the
EI mode, pulse width of 3 ms, electron energy of 20 eV, and filament
current of 2 µA. Delay time for excitation was 20 µs. The trapping
voltage of the analyzing cell was set at 2 V. The pressure of a neutral
sample was carefully controlled at 1.3 ( 0.1 × 10-7 Torr using a
variable leak valve (Anelva, High Vacuum Leak Valve 951), because
the pressure of the neutral, the number of ions in the analyzing cell,
was the most important factor for determining the relative peak heights.
The pressures were measured by means of a Bayard-Alpert type
ionization gauge (Anelva, MIG-430). The temperature of the sample
inlet system was 25 °C, and the ICR chamber was kept at 50 ( 1 °C.
The data points of 64 K were used in the acquisition. The mass range
for excite and detect events was from 90 to 130. The number of time-
domain transients summed prior to the Fourier transform was about
Experimental Section
Materials. THF was dried over sodium/benzophenone and distilled
immediately before use. All substituted benzaldehydes were com-
mercially available and purified either by distillation or recrystallization.
Benzaldehyde-carbonyl-13C was synthesized by the pyridinium dichro-
mate oxidation of benzyl alcohol-7-13C, which was prepared by
carbonylation of phenylmagnesium bromide with 13CO2 gas (99%,
Aldrich) and the borane reduction of the resultant benzoic acid-7-13C.
Benzaldehyde-d5 was prepared according to the same sequence of
reactions starting with bromobenzene-d5.
Reactions. All reactions were carried out at 0 °C under dry nitrogen
(16) Schriver, D. F. Manipulation of Air SensitiVe Compounds; McGraw-
Hill: New York, 1969; Chapter 7.
using the Schlenk tube technique.16 A lithium pinacolone enolate

Addition of Lithium Pinacolone Enolate to Benzaldehyde
J. Am. Chem. Soc., Vol. 119, No. 42, 1997 9979
1000, and a three term Blackman-Harris apodization and one zero fill
were applied to the time-domain data. The transient was finally
transformed into a mass spectrum. The relative abundance of carbon-
13 or carbon-12 to the deuterated benzaldehyde was calculated using
total peak heights of [M - 1]+, M+, and [M + 1]+. Under these
conditions, the relative abundance values were reproduced to about
(0.2% on repeated runs. The ICR chamber was baked out at 200 °C
for an overnight in order to avoid contamination before a different
sample was introduced into the analyzing cell.
Probe Experiments. Enone-isomerization and dehalogenation probe
experiments were done in a similar manner as reported previously.6c,7
Z-2,2,6,6-Tetramethylhepta-4-ene-3-one (0.5 mmol) and 1-methylnaph-
thalene (0.25 mmol) were dissolved in dry THF (1.5 mL), and the
solution was divided into three parts. To each of the divided solution
was added lithium pinacolone enolate (0.33 M, 0.5 mL) at 0 °C. The
solution was allowed to react for a preset time (1 min, 1 h, and 2 h),
and the degree of isomerization was analyzed by GLC (PEG 2m, 180
°C). Dehalogenation experiment was carried out similarly using
o-iodobenzophenone as a probe substrate instead of enone. After a
preset reaction time (1 min, 10 min, and 1 h), no dehalogenated product
was detected, and the starting material was recovered.
Acknowledgment. This investigation was supported by the
Grant in Aid for Scientific Research from the Ministry of
Education, Science, Sports and Culture, Japan as well as Ciba-
Geigy Foundation Japan for te promotion of Science.
JA971262C