Isotope effects and the mechanism of chlorotrimethylsilane-mediated addition of cuprates to enones

Article abstract of DOI:10.1021/ja993373c

Kinetic isotope effects were determined for the chlorotrimethylsilane- mediated reactions of cyclohexenone with lithium dibutylcuprate in tetrahydrofuran and with lithium butyl(tert-butylethynyl)cuprate in ether. For the reaction in tetrahydrofuran, the observation of a significant carbonyl oxygen isotope effect (¹⁶k/¹⁷k = 1.018-1.019) and small olefinic carbon isotope effects (¹²k/¹³k = 1.003-1.008) is consistent with rate- limiting silylation of an intermediate π-complex. Theoretically predicted isotope effects for model reactions support this conclusion. Rate-limiting silylation is also supported by relative reactivity studies of chlorotrimethylsilane versus chlorodimethylphenylsilane. The absence of a significant butyl-group carbon isotope effect on product formation indicates that the cuprate butyl groups are nonequivalent in the intermediate leading to the product-determining step. In diethyl ether the isotope effects revert to values similar to those found previously in reactions of cyclohexenone with lithium dibutylcuprate in the absence of chlorotrimethylsilane, consistent with no change in the overall mechanism in this solvent. A mechanistic hypothesis for the differing effects of TMSCl with changes in solvent and substrate is presented.

Full text of DOI:10.1021/ja993373c

3288
J. Am. Chem. Soc. 2000, 122, 3288-3295
Isotope Effects and the Mechanism of Chlorotrimethylsilane-Mediated
Addition of Cuprates to Enones
Doug E. Frantz*,† and Daniel A. Singleton*,‡
Contribution from the Department of Chemistry, Texas A&M UniVersity, College Station, Texas 77843,
and Laboratorium fu¨r Organische Chemie, ETH-Zu¨rich, CH-8092 Zu¨rich, Switzerland
ReceiVed September 16, 1999. ReVised Manuscript ReceiVed February 12, 2000
Abstract: Kinetic isotope effects were determined for the chlorotrimethylsilane-mediated reactions of
cyclohexenone with lithium dibutylcuprate in tetrahydrofuran and with lithium butyl(tert-butylethynyl)cuprate
in ether. For the reaction in tetrahydrofuran, the observation of a significant carbonyl oxygen isotope effect
(16k/17k ) 1.018-1.019) and small olefinic carbon isotope effects (12k/13k ) 1.003-1.008) is consistent with
rate-limiting silylation of an intermediate π-complex. Theoretically predicted isotope effects for model reactions
support this conclusion. Rate-limiting silylation is also supported by relative reactivity studies of chlorotri-
methylsilane versus chlorodimethylphenylsilane. The absence of a significant butyl-group carbon isotope effect
on product formation indicates that the cuprate butyl groups are nonequivalent in the intermediate leading to
the product-determining step. In diethyl ether the isotope effects revert to values similar to those found previously
in reactions of cyclohexenone with lithium dibutylcuprate in the absence of chlorotrimethylsilane, consistent
with no change in the overall mechanism in this solvent. A mechanistic hypothesis for the differing effects of
TMSCl with changes in solvent and substrate is presented.
Introduction
Scheme 1
Organocuprate conjugate addition reactions are an important
tool in synthetic organic chemistry.3 The success of these
reactions and the stereochemistry of the products are often
dependent on the detailed reaction conditions, including choice
of solvent and the presence of additives. Due to a hazy
mechanistic understanding of cuprate conjugate addition reac-
tions in general, the nature of the effects of solvents and
additives is not well understood.
The important but mechanistically perplexing effect of
additives is exemplified by cuprate conjugate additions in the
presence of chlorotrimethylsilane (TMSCl). TMSCl has a
remarkable effect on the rate and stereochemical outcome of
many cuprate reactions and can allow reactions to proceed which
would otherwise fail.2-5 However, several conflicting theories
have arisen as to the origin of these effects (Scheme 1). Corey
and Boaz suggested that TMSCl accelerates the conjugate
addition of cuprates to R,â-enones by the silylation of a d,π*-
complex to produce the silyl enol ether of a “CuIII” â-adduct
(e.g., 2 f 7).3 Kuwajima has contended that TMSCl acts as a
Lewis acid to activate the enone by complexation with the enone
oxygen (1 f 5).6,7 Lipshutz has presented NMR evidence that
the chloride of TMSCl acts as a Lewis base in an association
with a lithium cation of cuprate dimers, and this is proposed to
result in a “push-pull” effect on the conjugate addition (e.g., 1
f 6).8 More recently, Snyder and Bertz have suggested that
the chloride of TMSCl coordinates with copper to stabilize the
(5) (a) Linderman, R. J.; Godfrey, A. Tetrahedron Lett. 1986, 27, 4553.
(b) Johnson, C. R.; Marren, T. J. Tetrahedron Lett. 1987, 28, 27. (c)
Lindstedt, E.-L.; Nilsson, M.; Olsson, T. J. Organomet. Chem. 1987, 334,
255. (d) Smith, A. B., III; Dunlap, N. K.; Sulikowski, G. A. Tetrahedron
Lett. 1988, 29, 439. (e) Matsuzawa, S.; Horiguchi, Y.; Nakamura, E.;
Kuwajima, I. Tetrahedron 1989, 45, 349. (f) Corey, E. J.; Hannon, F. J.;
Boaz, N. W. Tetrahedron 1989, 45, 545. (g) Alexakis, A.; Sedrani, P.;
Mangeney, P. Tetrahedron Lett. 1990, 31, 345. (h) Sakata, H.; Aoki, Y.;
Kuwajima, I. Tetrahedron Lett. 1990, 31, 1161. (i) Bertz, S. H.; Smith, R.
A. J. Tetrahedron 1990, 46, 4091.
(6) Horiguchi, Y.; Komatsu, M.; Kuwajima, I. Tetrahedron Lett. 1989,
30, 7087.
(7) This proposal has been criticized based on negligible Lewis acidity
of TMSCl itself. See: Lipshutz, B. H.; Aue, D. H.; James, B. Tetrahedron
Lett. 1996, 37, 8471. However, this does not exclude the discreet silylation
of enones apparently depicted in ref 6.
† ETH-Zu¨rich.
‡ Texas A&M University.
(1) Recent reviews: (a) Perlmutter, P. Conjugate Addition Reactions in
Organic Synthesis; Pergamon Press: Oxford, 1992. (b) Lipshutz, B. H.;
Sengupta, S. Org. React. 1992, 41, 135-631. (c) Kozlowski, J. A. In
ComprehensiVe Organic Chemistry; Trost, B. M., Fleming, I., Eds.;
Pergamon Press: Oxford, 1991; Vol. 4, Chapter 14. (d) Rossiter, B. E.;
Swingle, N. M. Chem. ReV. 1992, 92, 77.
(2) (a) Nakamura, E.; Kuwajima, I. J. Am. Chem. Soc. 1984, 106, 3368.
(b) Alexakis, A.; Berlan, J.; Besace, Y. Tetrahedron Lett. 1986, 27, 1047.
(c) Horiguchi, Y.; Matsuzawa, S.; Nakamura, E.; Kuwajima, I. Tetrahedron
Lett. 1986, 27, 4025. (d) Nakamura, E.; Matsuzawa, S.; Horiguchi, Y.;
Kuwajima, I. Tetrahedron Lett. 1986, 27, 4029.
(8) Lipshutz, B H.; Dimock, S. H.; James, B. J. Am. Chem. Soc. 1993,
115, 9283.
(3) Corey, E. J.; Boaz, N. W. Tetrahedron Lett. 1985, 26, 6015.
(4) Corey, E. J.; Boaz, N. W. Tetrahedron Lett. 1985, 26, 6019.
(9) Bertz, H. B.; Miao, G.; Rossiter, B. E.; Snyder, J. P. J. Am. Chem.
Soc. 1995, 117, 11023.
10.1021/ja993373c CCC: $19.00 © 2000 American Chemical Society
Published on Web 03/25/2000

TMSCl-Mediated Addition of Cuprates to Enones
J. Am. Chem. Soc., Vol. 122, No. 14, 2000 3289
formation of the formal CuIII intermediate (7, X ) Cl-TMS)9
(with a CuI-like electronic distribution).10
Bu2CuLi (with excess cyclohexenone present).12 Because sig-
nificant reaction of butylcopper could complicate the isotope
effect analysis, we did not study the reactions of Bu2CuLi
generated from CuI in detail. The 13C KIEs for this reaction
The problem in distinguishing these diverse proposals is that
there is no direct evidence regarding the mechanistic effect of
TMSCl. Thus, it has not been resolved whether the changes in
rate and product stereochemistry arise from a fundamental
change in the rate-limiting step as in the Corey and Kuwajima
proposals or are the result of TMSCl coordination with either
the cuprate reagent itself or intermediates along the reaction
pathway. The difficulty has been compounded by large changes
in the effect of TMSCl with solvent5e,i and by some uncertainty
as to whether a silyl enol ether (e.g., 8) or an enolate (e.g., 4)
is the initial product in these reactions (vide infra).
We have recently demonstrated the utility of kinetic isotope
effects (KIEs) in the mechanistic study of cuprate conjugate
additions.11 The observation of significant 13C KIEs for both
the â carbon of the enone and the first carbon of the entering
alkyl group was indicative of rate-limiting reductive elimination,
and theoretically predicted KIEs for various mechanistic pos-
sibilities were found to support this interpretation. KIEs do not
provide direct information as to overall reaction pathways or
cuprate reagent structure, but they are uniquely illuminative as
to the major bonding changes occurring in the rate-limiting step.
We report here the results of a study of the 13C and 17O KIEs
for the TMSCl-mediated addition of organocuprates to cyclo-
hexenone in both THF and ether. The results provide a clear
indication of how TMSCl changes the cuprate conjugate addition
mechanism in these cases and how the effect of TMSCl changes
with solvent.
were determined by a recently reported methodology for the
combinatorial high-precision determination of small KIEs at
natural abundance.13 Reactions using natural abundance cyclo-
hexenone on a 0.3 mol scale were taken to 87.6 ( 0.9, 86.2 (
1.2, and 78.5 ( 1.3% conversion (based on unreacted cyclo-
hexenone) by the rapid addition of Bu2CuLi-LiBr-SMe2 to
vigorously stirred THF solutions of TMSCl/cyclohexenone (as
in the exploratory reactions, except using TMSCl/Bu2CuLi-
LiBr-SMe2/cyclohexenone in a 1:1:∼1.1 ratio) at -85 °C. (See
Method A in the Experimental Section.) The temperature of
the reaction mixtures rapidly rose to ca. -65 °C, and the
reactions were quenched after 5 min. The unreacted cyclohex-
enone was recovered by an extractive workup followed by
column chromatography and analyzed by 13C NMR8 compared
to a standard sample of original cyclohexenone from the same
bottle. The changes in 13C isotopic composition were calculated
using C6 as an “internal standard” assuming that its isotopic
composition does not change during the reaction.14 From the
changes in 13C isotopic composition the KIEs and errors were
calculated using the previously reported method.13
Results
Reactions in THF. In exploratory reactions, Bu2CuLi-
LiBr-SMe2 (generated from n-butyllithium and CuBr‚SMe2)
in THF was added rapidly to 1.1 equiv of TMSCl and 1.25
equiv of cyclohexenone dissolved in THF. In reactions with
initial temperatures of -78 and -90 °C quenched after 5 s by
rapid mixing with aqueous NH4Cl, there was approximately
quantitative formation (by GC analysis, based on reaction of
one butyl group) of 3-butyl-1-(trimethylsilyloxy)cyclohexene
(9). A parallel reaction without TMSCl required ∼30 min at
-90 °C for complete reaction. In the presence of TMSCl at
-78 °C the formation of 1-2% of 3-butylcyclohexanone (10)
was observed and at -90 °C no 3-butylcyclohexanone (<1%)
was observed. In no case could 1-butyl-2-cyclohexen-1-ol be
detected (<3%). Control reactions employing <1 equiv of
cyclohexenone did not return any detectable unreacted cyclo-
hexenone. This indicates that there was no significant formation
of an enolate of cyclohexenone, which would interfere with the
isotope effect determination below. As a control reaction for
the possibility that the reactions were occurring on mixing with
the warmer aqueous quench, reactions quenched after 5 s by
addition of a large excess of cold (-78 °C) methanol were also
complete. (As a control on the quenching ability of methanol,
no reaction occurred when the methanol was added first.)
However, extensive hydrolysis of the silyl enol ether occurs in
the methanol-quenched reactions (verified by the addition of
purified 9 to the reaction mixture). In reactions in which the
Bu2CuLi was generated from CuI we were unable to cleanly
stop the reaction at the utilization of one butyl group per
The 17O KIE was determined by a novel method15 that is a
variation on the above procedure. To both recovered and
standard samples of cyclohexenone was added ∼1 mole equiv
of diethyl ether (taken from the same bottle for each sample).
The 13C NMR integrations of C6 versus the methylene carbon
of diethyl ether were then used to precisely determined the
relatiVe ratios of diethyl ether in the two samples. The relative
integration of the 17O peaks could then be used to calculate the
change in 17O isotopic composition in the recovered material
versus the standard, which could be used to calculate the 17O
KIEs. This is the first determination of 17O KIEs by NMR and
this simple methodology should find general utility.
The 13C KIEs for the incoming butyl group were determined
by analysis of the product from reactions taken to low
conversion. Reactions of natural-abundance Bu2CuLi-LiBr-
SMe2 in THF and 1 equiv of TMSCl at -85 °C were taken to
∼10% conversion by the addition of 0.1 equiv of cyclohexenone.
(See Method B in the Experimental Section.) The quantitatively
formed 9 was completely hydrolyzed using either aqueous NH4-
Cl or TBAF, and the resulting 10 was isolated by an extractive
workup followed by column chromatography. An NMR-
standard sample of 10 was prepared by the addition of BuLi
(from the same bottle as used to form the Bu2CuLi-LiBr-
SMe2) to excess 3-ethoxy-2-cylohexen-1-one followed by
hydrolysis (1N HCl) and hydrogenation (H2/Pd/C). The two
(12) Butylcopper is known to add to cyclohexenone at -78 °C in the
presense of TMSCl. See ref 5b.
(13) Singleton, D. A.; Thomas, A. A. J. Am. Chem. Soc. 1995, 117, 9357.
(14) The KIEs determined here are relatiVe, being determined based on
relative integrations versus C6 of cyclohexenone and Cc of butylcyclohex-
anone. (See ref 13.) These carbons were chosen as internal standards because
their 13C peaks were most cleanly separated from other peaks in the 13C
NMR and potential impurities. Small deviations from KIEs of 1.000 for
these carbons will not affect the conclusions.
(10) Snyder, J. P. J. Am. Chem. Soc. 1995, 117, 11025.
(11) Frantz, D. E.; Singleton, D. A.; Snyder, J. P. J. Am. Chem. Soc.
1997, 119, 3383. A recent paper has concluded that the rate-limiting step
is formation of the Cu(III) intermediate (see: Canisius, J.; Gerold, A.;
Krause, N. Angew. Chem., Int. Ed. Engl. 1999, 38, 1644); however, this
appears inconsistent with the observed isotope effects.
(15) This method was developed by S. R. Merrigan in this laboratory.

3290 J. Am. Chem. Soc., Vol. 122, No. 14, 2000
Frantz and Singleton
Table 1. 13C and 17O KIEs (12k/13k or 16k/17k) for the Addition of
Bu2CuLi-LiBr-SMe2 to Cyclohexenone/TMSCl in THF (-85 to
-65 °C)
Table 2. 13C and 17O KIEs (12k/13k or 16k/17k) for the Addition of
11 to Cyclohexenone/TMSCl in Ether (-82 to -66 °C)
experiment
KIEs based on
experiment
starting materiala-c
7
8
KIEs based on
starting materiala-c
1
2
3
C1
C2
C3
O
1.002(4)
1.001(4)
1.014(6)
e
1.003(3)
1.005(4)
1.018(4)
1.003(10)
C1
C2
C3
O
1.005(2)
1.003(2)
1.007(2)
e
1.001(2)
1.007(1)
1.004(2)
1.018(6)
1.000(2)
1.006(2)
1.008(2)
1.019(7)
experiment
KIEs determined
experiment
5
from productb-d
9
10
KIEs determined
from productb-d
4
6
Ca
1.009(4)
1.010(3)
Ca
0.996(6)
0.999(4)
1.002(5)
a Experiments 7-8 are reactions carried to 80.6(5) and 84.9(7)%
completion, respectively, to determine the KIEs for cyclohexenone.b The
numbers in parentheses are standard deviations in the last digit. c See
ref 14. d Experiments 9-10 are reactions carried to ≈10% completion
to determine the KIEs for the incoming butyl group. e Not determined.
a Experiments 1-3 are reactions carried to 87.6(9), 86.2(1.2), and
78.5(1.3)% completion, respectively, to determine the KIEs for cyclo-
hexenone. b The numbers in parentheses are standard deviations in the
last digit. c See ref 14. d Experiments 4-6 are reactions carried to ≈10%
completion to determine the KIEs for the incoming butyl group. e Not
determined.
°C. The temperature of the reaction mixtures rapidly rose to
ca. -66 °C, and the reactions were quenched after 10 min. The
unreacted cyclohexenone was recovered and analyzed by 13C
and 17O NMR as described for the THF reactions. The 13C KIEs
for the incoming butyl group were determined from reactions
of cyclohexenone with 11 and 1 equiv of TMSCl taken to ∼10%
conversion by using limiting cyclohexenone, and the product
10 was analyzed as before.
The 13C and 17O KIEs observed in ether are summarized in
Table 2. Most notable of the observations are significant 13C
KIEs at both the â carbon of the enone and first carbon of the
butyl group and the absence of a significant 17O KIE.
samples of 10 were compared by 13C NMR, and the 13C KIEs
were calculated directly from the change in integrations relative
to Cc as internal standard.14 The precision of KIEs determined
in this manner ((0.3-0.6%) is limited by the reproducibility
of 13C NMR integrations but is sufficient for the purpose at
hand.
The 13C and 17O KIEs observed in THF are summarized in
Table 1. The most notable observations are (1) a substantial
17O KIE, (2) very small but significant normal 13C KIEs at both
the R and â carbons of the enones, and (3) the complete absence
of a 13C KIE for the first carbon of the butyl group.
Theoretical Calculations and Predicted Isotope Effects.
To aid in the interpretation of the experimental isotope effects,
the model reactions in Figure 1 were examined in Becke3LYP
density-functional theory calculations employing Ahlrichs’
“SVP” all-electron basis set for copper16 and a 6-31G* basis
set for the lighter elements.17 Structure 14 has been previously
reported.18 Equilibrium or kinetic isotope effects were calculated
by the method of Bigeleisen and Mayer19 using the program
QUIVER,20 with frequencies scaled by 0.9614.21 Tunneling
corrections were applied to the kinetic isotope effects using a
one-dimensional infinite parabolic barrier model.22 The accuracy
of these isotope effect calculations for simpler reactions has been
strongly supported by recent studies.23 The predicted isotope
effects are summarized in Figure 1, and the calculated structures
and energies are summarized in the Supporting Information.
On the basis of the predicted equilibrium isotope effects for
formation of the π complex in Figure 1c,d, we considered in
Reactions in Ether. We were unable to cleanly stop the
reactions of Bu2CuLi-LiBr-SMe2/TMSCl with cyclohexenone
at the utilization of one butyl group with excess cyclohexenone
present. For this reason the reactions in ether employed the
mixed cuprate 11 (prepared from the sequential reaction of
CuBr‚SMe2 with lithium tert-butylacetylide then n-BuLi) in
which the tert-butylethynyl group is a “non-transferable ligand”.
Reactions of 11 with 1-1.1 equiv of TMSCl and 1.25 equiv of
cyclohexenone quenched after 5 s at -100 °C were >90%
complete, and afforded 3-butylcyclohexanone nearly quantita-
tively with 1-3% 3-butyl-1-(trimethylsilyloxy)cyclohexene as
the only detectable cyclohexenone-derived byproduct. A parallel
reaction without TMSCl was also >90% complete in 5 s. In
(16) Scha¨fer, A.; Horn, H.; Ahlrichs, R. J. Chem. Phys. 1992, 97, 2571.
(17) For a discussion of theoretical methodology for organocuprate
structures, see: Nakamura, E.; Mori, S.; Nakamura, M.; Morokuma, K. J.
Am. Chem. Soc. 1997, 119, 4887.
(18) (a) Nakamura, E.; Mori, S.; Morokuma, K. J. Am. Chem. Soc. 1997,
119, 4900.
(19) (a) Bigeleisen, J.; Mayer, M. G. J. Chem. Phys. 1947, 15, 261. (b)
Wolfsberg, M. Acc. Chem. Res. 1972, 5, 225.
(20) Saunders, M.; Laidig, K. E.; Wolfsberg, M. J. Am. Chem. Soc. 1989,
111, 8989.
contrast, the reaction of 11 with cyclohexenone/TMSCl in THF
at -78 °C afforded a 98:2 ratio of 9:10. This suggests that the
divergent results with Bu2CuLi-LiBr-SMe2 in THF and 11
in ether are due to the change in solvent and not due to the
change in cuprate structure.
To determine the KIEs for the reactions of 11 in ether, reac-
tions of cyclohexenone on a 0.3 mol scale were taken to 80.6
( 1.4 and 84.9 ( 0.8% completion by the rapid addition of 11
to vigorously stirred solutions of cyclohexenone and TMSCl
(using TMSCl/11/cyclohexenone in a 1:1:1.1-1.2 ratio) at -82
(21) Scott, A. P.; Radom, L. J. Phys. Chem. 1996, 100, 16502.
(22) Bell, R. P. The Tunnel Effect in Chemistry; Chapman & Hall:
London, 1980; pp 60-63.
(23) Beno, B. R.; Houk, K. N.; Singleton, D. A. J. Am. Chem. Soc. 1996,
118, 9984. Singleton, D. A.; Merrigan, S. R.; Liu, J.; Houk, K. N. J. Am.
Chem. Soc. 1997, 119, 3385.

TMSCl-Mediated Addition of Cuprates to Enones
J. Am. Chem. Soc., Vol. 122, No. 14, 2000 3291
Effect of TMSCl on Stereochemistry. The stereochemistry
of cuprate additions to 5-methyl-2-cyclohexenone (17) was
studied under the conditions of the isotope effect studies and
also in the absence of TMSCl. After treatment of the crude
product mixtures with Bu4NF (when using TMSCl), the ratio
of trans- and cis-5-methyl-3-butylcyclohexanones (18 and 19)
was analyzed by integration of the 13C signals for the C2 and
C6 carbons (with delays >5T1). For additions of Bu2CuLi-
LiBr-SMe2 in THF the ratio of trans and cis products in the
absence of TMSCl was 99:1 (in rough agreement with previous
observations25) but the selectivity decreased to 90:10 with
TMSCl. For the addition of 11 in ether the cis isomer could
not be detected (>150:1) in either the presence or absence of
TMSCl.
Figure 1. Theoretically predicted equilibrium or kinetic 13C isotope
12
13
effects (k C/k C) for model reactions at -78 °C. The prediction in
part d is based on the most stable chair form 15. Numbers in brackets
in parts c and d are predicted values for the observed KIEs for rate-
limiting silylation of the π complexes, based on a simulation of eq 1
(see text).
detail the observed kinetic isotope effect to be expected if the
mechanism involved rate-limiting silylation of the π complex
as in eq 1 . The overall KIE for rate-limiting silylation of a π
K
k
1 + R2CuLi y
\
z 2
8 6 or 7
(1)
TMSCl
rate-limiting step
Discussion
We analyze here our results within the mechanistic framework
of possibilities shown in Scheme 1. The commonly proposed
general reaction scheme involving an initial π complex followed
by a formal Cu(III) intermediate appears consistent with
experimental observations.3,26 However, there are many unestab-
lished details within this framework, particularly with regard
to the cuprate cluster structure, and other more complex
mechanisms cannot be excluded. A specific alternative mech-
anism worth noting is the 1,2-addition of a Cu-R bond across
the CdC bond of enones to afford R-cuprio ketones.27 This
mechanism was first proposed in the 1970s and a variation of
it has recently been revived based on theoretical calculations
on the addition of the cluster (Me2CuLi)2 to acrolein.18 This
mechanism does not account for the observed KIEs11,28 nor does
it provide an explanation for the effect of TMSCl on cuprate-
mediated cis-trans isomerization of enones (vide infra). Con-
sidering the critical influence of solvent on the course of these
complex would be a complex function of the equilibrium isotope
effect for formation of the π complex, the kinetic isotope effect
for the silylation step, and the relative amounts of π complex
and free enone present in the reaction mixture. The latter should
change as the reaction proceeds, further convoluting a calcula-
tion of the KIE observed in recovered starting material from
reactions taken to high conversion. To allow for these complica-
tions, the reaction of eq 1 was simulated in an Excel spreadsheet
(see the Supporting Information). Isotope effects (12k/13k) in the
silylation step of 1.000 were assumed for the R and â enone
carbons. In the limiting case where free enone is absent, the
observed R and â enone KIEs would then be 1.000. In the
opposite limiting case of minimal π-complex formation, the
observed KIEs would equal the equilibrium isotope effects for
formation of the π complex. Using the actual experimental
concentrations and a K of 110,24 the observed KIEs would be
slightly over half of the equilibrium isotope effects (Figure 1c,d).
For the complex of cyclohexenone with Me2CuLi, the two
chair forms 15 and 16 were found. Of these, 15 was predicted
to be more stable than 16 by 1.4 kcal/mol (including ZPE).
(Structures 15 and 16 are geometrically based on the calculated
structures; see the Supporting Information for the detailed
geometries. Structures 14, 15, and 16 are simplistic models for
the complex dimeric solvent-coordinated species actually in
solution.)
(25) House, H. O.; Fischer, W. F., Jr. J. Org. Chem. 1968, 33, 949.
(26) (a) Hallnemo, G.; Olsson, T.; Ullenius, C. J. Organomet. Chem.
1985, 282, 133. (b) Lipshutz, B. H.; Dimock, S. H.; James, B. J. Am. Chem.
Soc. 1993, 115, 9283. (c) Bertz, S. H.; Smith, R. A. J. J. Am. Chem. Soc.
1989, 111, 8276. (d) Alexakis, A.; Sedrani, R.; Mangeney, P. Tetrahedron
Lett. 1990, 31, 345. (e) Krause, N.; Wagner, R.; Gerold, A. J. Am. Chem.
Soc. 1994, 116, 381. (f) Vellekoop, A. S.; Smith, R. A. J. J. Am. Chem.
Soc. 1994, 116, 2902.
(27) (a) Berlan, J.; Battioni, J.; Koosha, K. J. Organomet. Chem. 1978,
152, 359-365. (b) Berlan, J.; Battioni, J.; Koosha, K. Bull. Soc. Chim. Fr.
1979, 183-190.
(24) Krauss, S. R.; Smith, S. G. J. Am. Chem. Soc. 1981, 103, 141-148.
The observed equilibrium in this paper probably reflects a combina-
tion of π complex formation and lithium coordination to the enone oxygen.
See: Vellekoop, A. S.; Smith, R. A. J. J. Am. Chem. Soc. 1994, 116, 2902.
(28) Allowance for the equilibrium presence of the enone π complex as
done here may decrease the discrepancy between experimental and
calculated isotope effects in these reactions. See: Mori, S.; Nakamura, E.
Chem. Eur. J. 1999, 5, 1534-1543.

3292 J. Am. Chem. Soc., Vol. 122, No. 14, 2000
Frantz and Singleton
reactions,29 further calculations involving solvated reactants may
resolve these problems.
methylation of acrolein, are all slightly inverse (Figure 1a,b).
This supports the qualitative argument that rate-limiting sily-
lation of the starting enone cannot account for the observed
KIEs.
A major point of confusion in TMSCl-mediated reactions has
been the identity of the initial product. Many workers have used
workup conditions which would hydrolyze silyl enol ethers such
as 9, and most others have not attempted to distinguish between
direct formation of a silyl enol ether and silylation of an initially
formed enolate. Nonetheless, it has often been assumed that
silyl enol ethers or related intermediates are the initial products
in these reactions. It was therefore surprising when it was
reported that the reaction of Bu2CuLi‚LiI with cyclohexenone
in THF in the presence of 6 equiv of TMSCl afforded 10 as the
major product (after quenching of a presumed initial enolate)
with only 27% of silyl enol ether 9 found at the shortest reaction
time.9 Recent workers have found that TMSI-mediated reactions
of Bu2CuLi‚LiI with cyclohexenone afford predominately 9,30
and a footnote in that paper reports a private communication
from one of the previous workers stating that the use of only
1-2 equiv of TMSCl affords “mainly TMS enol ether”.
Apparently the use of 6 equiv of TMSCl and aqueous quenching
resulted in the hydrolysis of some of the initially formed 9. In
our experiments using only 1 equiv of TMSCl and cuprate
derived from CuBr‚SMe2, the silyl enol ether 9 is virtually the
exclusive product present at very short reaction times in THF.
However, the observation of silyl enol ether at short reaction
times does not in itself conclusively establish whether silylation
is an integral part of the reaction mechanism or occurs “after
the fact”, i.e., after the rate-limiting step by silylation of an
enolate. In this regard the 17O KIEs observed in these reactions
are telling. The substantial 17O KIE observed in THF is
indicative of the carbonyl oxygen undergoing a major bonding
change in the rate-limiting step. The 17O KIE in THF is much
larger than would be expected from a 1,4-addition to the
enone: very small 18O KIEs are observed in carbonyl addition
reactions.31 In contrast, the 17O KIE in ether, where there is no
significant formation of silylated product (consistent with
previous observations9), is within experimental error of unity.
Taken together, these results strongly implicate rate-limiting
silylation of the carbonyl oxygen in THF.
The involvement of silylation in the rate-limiting step in THF
places significant constraints on the mechanism. If TMSCl were
simply acting as a Lewis acid as proposed by Kuwajima, then
the coordination of TMSCl with cyclohexenone to form 5 would
have to be rate limiting and (surprisingly!) irreversible under
the reaction conditions. The cyclohexenone 13C isotope effects
would reflect this silylation step, and would be expected to be
approximately unity at the olefinic carbons. The small but
significant positive deviation from unity of the 13C KIEs at both
the R and â carbons argues against this mechanism. This
conclusion is supported by the theoretically predicted isotope
effects for model reactions. The equilibrium 13C KIEs for the
formation of 12 were calculated as a rough model for the
secondary KIEs expected if silylation of the starting enone is
rate limiting, since the carbons are not being transferred in this
step.32 The predicted equilibrium isotope effects for the forma-
tion of 12, as well as the predicted kinetic isotope effects for
An important basis for the Corey proposal of silylation of a
d,π*-complex was the observation that acyclic enones underwent
cis-trans isomerization in the absence of TMSCl but not in
the presence of TMSCl in THF.3 It was proposed that the cis-
trans isomerization without TMSCl is the result of reversible
formation of a “CuIII” â-adduct before the rate-limiting step,
and this proposition has recently been strongly supported in KIE
studies.11 The lack of isomerization with TMSCl would then
be consistent with a switch to an earlier rate-limiting step, but
does not by itself distinguish between rate-limiting silylation
of a π complex, rate-limiting formation of the Cu â-adduct with
rapid silylation, or rate-limiting formation of the π complex.
The last of these has often been assumed in stereochemical
arguments6,5g (misleadingly! vide infra), but the latter two
possibilities are both excluded here by the requirement for a
rate-limiting silylation step.
The remaining possibility of rate-limiting silylation of a π
complex is supported by the theoretically predicted isotope
effects. From Figure 1c,d, small normal kinetic 13C isotope
effects are predicted at both the R and â enone carbons. These
predictions involve ad hoc assumptions regarding the kinetic
isotope effects for the silylation step and the equilibrium constant
for formation of the π complex, as well as a simplistic modeling
of a likely more complex enone-π complex structure. Nonethe-
less the striking agreement between the predicted and experi-
mental KIEs demonstrates that the observed R and â enone 13
C
KIEs are consistent with rate-limiting silylation of a π complex.
Rate-limiting silylation of the π complex also provides an
explanation for the enigmatic stereochemical results of Kuwa-
jima which had led to the proposal of an initial Lewis acid
complex.6 The key observation was that while the cis π complex
20 was expected to be most stable in additions to 6-tert-butyl-
2-methylcyclohexenone, the trans product was predominant in
the presence of TMSCl. However, the stereoelectronically
enforced approach of TMSCl in the plane of the carbonyl group
(toward the lone pairs of the carbonyl oxygen) would be highly
sterically hindered, blocking the formation of the cis product.
The trans product can be formed by the less disadvantageous
silylation of conformation 21 that places the tert-butyl group
in a pseudoaxial position.33
The stereochemistry of the TMSCl-mediated addition of
Bu2CuLi-LiBr-SMe2 to 5-methylcyclohexenone in THF may
also be rationalized by rate-limiting silylation of a π complex.
Based on the theoretical predictions for 15 and 16, the trans π
complex 22 would be expected to be preferred over the cis
(29) (a) Kingsbury, C. L.; Smith, R. A. J. J. Org. Chem. 1997, 62, 7637.
(b) Kingsbury, C. L.; Smith, R. A. J. J. Org. Chem. 1997, 62, 4629.
(30) Eriksson, M.; Johansson, A.; Nilsson, M.; Olsson, T. J. Am. Chem.
Soc. 1996, 118, 10904. See especially footnote 20. In addition, a recent
paper by one of the original authors in ref 9 also states that with the use of
a large excess of TMSCl, hydrolysis of the initially formed silyl enol ether
occurs upon workup. See: Bertz, S. H.; Chopra, A.; Eriksson, M.; Ogle,
C. A.; Seagle, P. Chem. Eur. J. 1999, 5, 2680.
(32) (a) It is commonly assumed that equilibrium secondary hydrogen
isotope effects are an upper bound for the kinetic isotope effect, though
exceptions have been noted. See: Glad, S. S.; Jensen, F. J. Org. Chem.
1997, 62, 253. (b) Modeling of the secondary KIEs using equilibrium isotope
effects is bolstered by the expectation that the endothermic silylation step
should have a late transition state.
(33) Molecular mechanics calculations (MM2) predict that having the
tert-butyl group axial in 6-tert-butylcyclohexenone is disfavored by only
0.8 kcal/mol.
(31) Marlier, J. F. J. Am. Chem. Soc. 1993, 115, 5953. O’Leary, M. H.;
Marlier, J. F. J. Am. Chem. Soc. 1979, 101, 3300.

TMSCl-Mediated Addition of Cuprates to Enones
J. Am. Chem. Soc., Vol. 122, No. 14, 2000 3293
complex 23, but not overwhelmingly. If 22 and 23 are assumed
to be equally reactive in the rate-limiting silylation, the trans:
cis product ratio should reflect the energy difference between
22 and 23. The ∆∆G‡ of 0.9 kcal/mol for formation of 18 and
19 (based on a 91:9 ratio at -78 °C) is in very reasonable
agreement with the predicted difference of 1.4 kcal/mol for 15
and 16. It should be noted that the observation of rate-limiting
in the reaction of Bu2CuLi-LiBr-SMe2 with cyclohexenone
without TMSCl (1.020-1.0026 at the â enone carbon, 1.011-
1.016 at the first carbon of the butyl group).11 There is thus no
indication that TMSCl changes the rate-limiting step in ether,
and the results are consistent with rate-limiting reductive
elimination. This conclusion is supported by Corey’s observation
that TMSCl did not inhibit cis-trans isomerization of an acyclic
enone in ether,5f in contrast to observations in THF. (Notably,
the use of trimethylsilyl triflate in ether instead of TMSCl did
inhibit cis-trans isomerization.) Unlike the THF reaction,
TMSCl did not affect the stereochemistry of addition of 11 to
5-methylcyclohexenone. There is no clear indication that TMSCl
affects the additions of 11 to enones in ether at all.
However, TMSCl does have a large impact on some cuprate
conjugate additions in ether, particularly those involving addi-
tions to R,â-unsaturated esters, amides, and nitriles.2b,5g It is
notable that these are relatively unreactive substrates for un-
catalyzed cuprate conjugate additions. The large changes ob-
served in the reactivity of unsaturated amides2b and the stereo-
chemistry of additions to unsaturated esters5g would indicate
that TMSCl is not merely improving yields by trapping enolates
but is mediating a mechanistic change akin to its effect on
addition to cyclohexenone in THF.
A consistent mechanistic hypothesis for the differing effects
of TMSCl with changes in solvent and substrate starts with the
observation that the coordination of lithium to the carbonyl
oxygen is a critical factor in cuprate conjugate additions without
additives. No reaction occurs in the presence of excess 12-
crown-4.34 In ether the lithium coordination is sufficient for the
reaction to proceed rapidly with cyclohexenone, and the TMSCl-
accelerated process cannot compete. (However, as mentioned
above, the more powerful electrophile trimethylsilyl triflate can.)
In the more basic THF the lithium coordination is attenuated28
and acceleration by irreversible silylation of an enone-cuprate
π complex dominates. The irreversible nature of the silylation
brings about the change in the rate-limiting step from carbon-
carbon bond formation in the absence of TMSCl. (It is important
to remember that the cuprate reagent itself has changed in going
from ether to THF, but we would emphasize the effect of lithium
coordination.) With unsaturated esters and amides the lithium-
carbonyl coordination is not sufficient, even in ether, to activate
the conjugate addition (though more powerful Lewis acids can35)
and the TMSCl-mediated process takes over.
silylation of the π complex does not exclude an important role
for TMSCl-lithium coordination as proposed by Lipshutz or
the stabilization of the formal CuIII intermediate by a TMSCl-
copper coordination proposed by Snyder and Bertz. The latter,
however, would occur after the rate-limiting step and could not
be responsible for the rate and stereoselectivity effects of TMSCl
in THF.
An intriguing feature of the isotope effects observed in THF
is the complete absence of a 13C KIE for the first carbon of the
butyl group. Obviously, the butyl group is not being transferred
in the rate-determining step, but a subtle yet important inference
may be made about the product-determining step. If the
intermediate before reductive elimination were, for example,
24, in which the two butyl groups are equivalent (except for
isotopic substitution), then the intramolecular competition
between these butyl groups would result in an isotope effect
on which product is formed. It can be deduced that 24 is not
the penultimate intermediate! In more general terms it can be
concluded that in the intermediate preceding product formation
the butyl groups are neither equivalent nor do they rapidly
interconvert relative to the rate of product formation. This
observation may be taken as being consistent with the proposal
of Snyder that a formal CuIII intermediate would require
stabilization by solvation in a square-planar complex,10 with
the proviso that the butyl groups are diastereotopic (i.e. cis) as
in 25. A more complex cuprate cluster in which the butyl groups
are nonequivalent could also account for this result.
Conclusions
The central cause of the effect of TMSCl on the addition of
Bu2CuLi to cyclohexenone in THF is a change in the rate-
limiting step. The KIEs observed are qualitatively consistent
with Corey’s proposal that TMSCl traps an intermediate π
complex, with the important specification that silylation is the
rate-limiting step. This interpretation is supported by theoreti-
cally predicted isotope effects and is consistent with a variety
of observations. The results have no bearing on the “push-
pull” proposal of Lipshutz (e.g., 1 f 6) but do not support any
importance for TMSCl coordination of the initial enone or
TMSCl coordination of a formal CuIII intermediate. There is
still considerable uncertainty as to the cuprate cluster structure
at every stage of the reaction pathway, but a consistent
qualitative mechanistic picture is emerging.
It is well-known that TMSCl has less effect on additions to
enones in ether than in THF,5e,i but it was important to establish
the mechanistic basis for this difference. The KIE results in ether
are very different from those in THF. The KIEs for the reaction
of 11 with cyclohexenone in the presence of TMSCl (1.014-
1.018 at the â enone carbon, 1.009-1.010 at the first carbon of
the butyl group) are very similar to those observed previously
The outcome of the black box of organocuprate reactions can
depend on cuprate structure, substrate, solvent, additives, and
(34) Ouannes, C.; Dressaire, G.; Langlois, Y. Tetrahedron Lett. 1977,
10, 815.
(35) Yamamoto, Y.; Yamamoto, S.; Yatagai, H.; Ishiyama, Y.; Maruya-
ma, K. J. Org. Chem. 1982, 47, 119.

3294 J. Am. Chem. Soc., Vol. 122, No. 14, 2000
Frantz and Singleton
Table 3. Average Integrations for Cyclohexenonea,b
the detailed reaction conditions, and the mechanistic understand-
ing of these effects should greatly aid progress in the area. We
anticipate that kinetic isotope effects will continue to be an
effective tool toward that end.
experiment
C1
C2
C3
THF
1
standard
R/R0
991.6(2.1)
980.9(2.6)
1.011(0.003)
1096.1(2.9)
1089.7(3.5)
1.006(0.004)
1103.5(2.7)
1087.6(4.4)
1.015(0.005)
Experimental Section
2
990.2(2.8)
987.4(2.0)
1.003(0.003)
1095.9(1.9)
1081.5(2.1)
1.013(0.003)
1104.6(3.7)
1095.8(3.5)
1.008(0.005)
All reactions were carried out in dried glassware and freshly purified
solvent under a positive pressure of nitrogen using standard airless
techniques. Chlorotrimethylsilane was distilled from tributylamine
immediately before use.
standard
R/R0
3
985.3(1.9)
985.2(1.7)
1.000(0.003)
1096.0(2.4)
1085.4(1.7)
1.010(0.003)
1098.4(2.4)
1085.1(1.2)
1.012(0.003)
standard
R/R0
Reactions in THF. Method A. To a mixture of 55.50 g (0.27 mol)
of CuBr‚SMe2 and 1.5 L of THF at -78 °C was added dropwise
(maintaining a temperature below -70 °C) 211 mL (0.54 mol) of a
freshly titrated 2.56 M solution of n-butyllithium in hexanes, and the
mixture was stirred for 30 min. This mixture was then added, in less
than 10 s via a glass-tubing connection, to a vigorously mechanically
stirred separate flask containing 28.84 g (0.30 mol) of 2-cyclohexenone,
2.0 L of THF, 29.33 g (0.27 mol) of chlorotrimethylsilane, and 5.11 g
of dodecane (internal standard) at -85 °C. The temperature of the
resulting mixture rapidly increased to -65 °C. After 5 min the reaction
was quenched by the rapid addition of 1 L of saturated aqueous
NH4Cl. The percent conversion based on GC analysis of an aliquot
was 87.6 ( 0.9%. The reaction mixture was allowed to come to room
temperature and was extracted in 1-L portions with three 500-mL
portions of diethyl ether (adding sufficient water initially to make all
salts dissolve). The ether extracts were combined, washed with saturated
NH4Cl, brine, and water, dried (MgSO4), and concentrated under
vacuum. The residue was then chromatographed on a 4 cm × 30 cm
silica gel column using 10% ethyl acetate/petroleum ether (30-60 °C)
as eluent to afford 0.98 g of 2-cyclohexenone in 99% purity by GC.
ether
7
standard
R/R0
980.2(4.2)
977.3(4.2)
1.003(0.006)
1097.3(5.6)
1095.8(3.2)
1.001(0.006)
1127.1(8.7)
1101.8(4.6)
1.023(0.009)
8
954.0(3.0)
948.8(4.7)
1.005(0.006)
1085.1(5.4)
1073.9(5.5)
1.010(0.007)
1107.9(3.9)
1072.2(5.7)
1.033(0.007)
standard
R/R0
a See Tables 1 and 2 for the numbering scheme for experiments and
carbons. b Integrations are relative to C6 which was set at 1000. The
integrations are averaged from five spectra for each sample. Standard
deviations are shown in parentheses.
A second reaction performed by an analogous procedure proceeded
to 84.9 ( 0.7% conversion.
Method B. In a typical procedure, 21.73 g (0.20 mol) of chloro-
trimethylsilane was added to a solution of 11 at -78 °C prepared as
described in Method A. After the mixture was stirred for 5 min, 1.92
g (20 mmol) of 2-cyclohexenone in 50 mL of ether was added dropwise.
2-Cyclohexenone could not be detected by GC analysis of an aliquot
taken after 10 min, and the reaction was quenched by the addition of
1 L of saturated aqueous NH4Cl. An extractive workup as in Method
A yielded a residue that was chromatographed on a 4 cm × 30 cm
silica gel column using 7% ethyl acetate/petroleum ether as eluent to
afford 3-butylcyclohexanone (>99% purity by GC).
NMR Standard 3-Butylcyclohexanone. To a mixture of 60 mL of
ether and 4.91 g (35 mmol) of 3-ethoxy-2-cyclohexen-1-one at 0 °C
was added dropwise 30 mmol of n-butyllithium in hexanes (taken from
the same bottle used to generate the cuprate in each case). After being
stirred for 30 min at 0 °C, the reaction mixture was quenched with
excess 1 N HCl, extracted with three 20-mL portions of ether, dried
(MgSO4), and concentrated under vacuum. The residue was dissolved
in 100 mL of absolute ethanol and hydrogenated using 0.1 g of 5%
Pd/C in a Parr apparatus at 42 psi of H2. Filtration and concentration
of the reaction mixture followed by silica gel chromatography using
7% ethyl acetate/petroleum ether as eluent afforded 3-butylcyclohex-
anone (>99% purity by GC).
Two reactions carried out by analogous procedures proceeded to
86.2 ( 1.2% and 78.5 ( 1.3% conversion.
Method B. In a typical procedure, 156 mL (0.40 mol) of a 2.56 M
solution of n-butyllithium in hexanes was added dropwise (maintaining
a temperature below -70 °C) to a mixture of 41.1 g (0.20 mol) of
CuBr‚SMe2 and 1.5 L of THF at -78 °C, and the mixture was stirred
for 30 min. At this point, the reaction was cooled to -85 °C and 21.73
g of chlorotrimethylsilane was added. The mixture was stirred for 1
min and 1.92 g (20 mmol) of 2-cyclohexenone in 50 mL of THF was
added dropwise over 30 min. 2-Cyclohexenone could not be detected
by GC analysis of an aliquot taken after 10 min. The reaction was
then quenched by the addition of 1 L of saturated aqueous NH4Cl, and
the solution was allowed to warm to room temperature and then stirred
until all of the 3-butyl-1-(trimethylsilyloxy)cyclohexene was hydrolyzed
to 3-butylcyclohexanone. An extractive workup as in Method A yielded
a residue that was chromatographed on a 4 cm × 30 cm silica gel
column using 7% ethyl acetate/petroleum ether as eluent to afford 1.7
g of 3-butylcyclohexanone (>99% purity by GC).
Reactions in Ether. Method A. A mixture of 200 mL of ether, 76
mL (0.20 mol) of a 2.62 M solution of n-butyllithium in hexanes, and
17.25 g (0.21 mol) of tert-butylacetylene (added last, dropwise) was
stirred at -78 °C for 30 min and 0 °C for 30 min. The resulting solution
was then transferred to a mixture of 41.12 g (0.20 mol) of CuBr‚SMe2
and 1.2 L of Et2O at 0 °C and stirred until all of the CuBr‚SMe2
dissolved. At this point the solution was cooled to -78 °C and 72.2
mL (0.19 mol) of a 2.62 M solution of n-butyllithium in hexanes was
added dropwise, and the resulting light brown solution of 11 was stirred
at -78 °C for 1 h. The solution of 11 was then added, in less than 10
s via a glass-tubing connection, to a vigorously mechanically stirred
separate flask containing 23.07 g (0.24 mol) of 2-cyclohexenone, 2.0
L of Et2O, 21.73 g (0.20 mol) of chlorotrimethylsilane, and 6.13 g of
dodecane (internal standard) at -82 °C. The temperature of the resulting
mixture rapidly increased to -66 °C. After 10 min the reaction was
quenched by the rapid addition of 1 L of saturated aqueous NH4Cl.
The percent conversion based on GC analysis of an aliquot was 80.6
( 1.4%. An extractive workup and chromatography as in Method A
of the THF reactions, using 90:10 v/v saturated NH4Cl:NH4OH in place
of the saturated NH4Cl in the extractions, afforded 1.54 g of recovered
cyclohexenone in 99% purity by GC analysis.
Additions to 5-Methyl-2-cyclohexenone (17). The reactions of 1736
were carried out by adding 220 mg (2.0 mmol) of 17 dropwise to either
2.0 mmol of Bu2CuLi-LiBr-SMe2 (prepared as described above) in
20 mL of THF or 2.0 mmol of 11 (prepared as described above) in 20
mL of ether, with or without 272 mg (2.5 mmol) of TMSCl (added 1
min before the 17). All of the reactions were stirred for 30 min at -78
°C and quenched by the addition of 10 mL of saturated NH4Cl (reactions
in THF) or 10 mL of 90:10 v/v saturated NH4Cl:NH4OH (reactions in
ether). After an extractive workup and treatment of the product mixtures
with Bu4NF (when using TMSCl), the crude product was analyzed by
1H and 13C NMR.
NMR Measurements. NMR samples were generally prepared using
1.00 g of material in a 10 mm NMR tube with a sample height of 5.0
cm, and in all cases recovered and standard material samples were
prepared identically. A T1 determination by the inversion-recovery
method was carried out for each NMR sample, and the T1 for each
carbon remained constant within experimental error from sample to
sample.
The 13C spectra were obtained at 100.577 MHz on a Varian XL400
broadband NMR spectrometer, with inverse gated decoupling, calibrated
(36) Blanchard, J. P.; Goering, H. L. J. Am. Chem. Soc. 1951, 73, 5863.

TMSCl-Mediated Addition of Cuprates to Enones
J. Am. Chem. Soc., Vol. 122, No. 14, 2000 3295
Table 4. Average Integrations for 3-Butylcyclohexanonea,b
standard, and ∆IntSample and ∆IntStandard are the standard deviations
in the integrations for the sample and standard, respectively.
The KIEs for cyclohexenone were then calculated from eq 3, with
the standard deviations calculated from eqs 4, 5, and 6.13
experiment
Ca in sample
Ca in standard
THF
4
1067.9(5.9)
1064.5(3.1)
1034.6(4.6)
1063.6(3.1)
1063.6(3.1)
1036.2(1.8)
5
6
ln(1 - F)
ln[(1 - F)R/R0]
KIEcalcd
)
(3)
(4)
ether
9
1026.3(3.3)
1025.0(2.6)
1035.5(1.7)
1035.5(1.7)
10
-ln(R/R0)
(1 - F) ln2[(1 - F)R/R0]
∂KIE
∆F )
∆KIEF )
∂KIE
∆F
a See Tables 1 and 2 for numbering the scheme for experiments and
carbons. b Integrations are relative to Cc which was set at 1000. The
integrations are averaged from five spectra for each sample. Standard
deviations are shown in parentheses.
∂F
-ln(1 - F)
(R/R0) ln2[(1 - F)R/R0]
∆KIER )
∆(R/R0) )
∆(R/R0) (5)
∂(R/R0)
Table 5. Average Integrations for Spectra of Cyclohexenone/Ether
Mixtures
∆KIE ) KIE*(∆KIER/KIE)2 + (∆KIEF/KIE)2)1/2
(6)
experiment
C6 in 13C NMR
CdO in 17O NMR
THF
2
standard
The KIEs in Tables 1 and 2 for experiments 4-6, 9, and 10 were
calculated from the data in Table 4, without correction for percent
conversion, using eq 7, and the standard deviations (∆KIE) were
calculated using eq 8.
229.9(0.2)
454.9(0.4)
572.0(2.1)
1092.6(11.4)
3
233.6(1.0)
220.1(1.2)
479.4(2.8)
438.7(2.4)
standard
ether
8
∆IntStandard
IntSample
KIE )
(7)
(8)
390.7(7.0)
470.1(1.4)
915.9(3.4)
1095.0(1.4)
standard
2
2 1/2
∆IntSample
IntSample
∆IntStandard
IntStandard
∆KIE ) KIE ×
+
((
)
(
) )
45° pulses, 120-s delays between pulses, and at a controlled temperature
of 30 °C. To obtain sufficient digital resolution (5 points/ν1/2 is minimal),
a 220032 point FID was zero-filled to 512K points before Fourier
transformation. Integrations were determined numerically using a (3.0
Hz region for each peak. A zeroth-order baseline correction was
generally applied, but in no case was a first-order (tilt) correction
applied.
17O KIE Results. For the cyclohexenone/ether mixtures, the
integration of cyclohexenone C6 in the 13C spectra and the carbonyl
oxygen in the 17O spectra were set relative to integrations of 1000 for
the methylene carbon of ether and 1000 for the ether oxygen,
respectively. The averaged results from four 13C spectra and five 17O
spectra are shown in Table 5. The R/R0 for the cyclohexenone carbonyl
oxygen was then calculated from eq 9, and ∆R/R0 was calculated in a
standard fashion from the square root of the sum of the squares of the
standard deviations in the integrations used to calculate R/R0. The R/R0’s
for experiments 2, 3, and 8 were 1.036(0.011), 1.030(0.011), and 1.006-
(0.019), respectively. From these values, the KIEs in Tables 1 and 2
were calculated using eqs 3-6.
The 17O spectra were taken on samples of cyclohexenone containing
∼1 mole equiv of diethyl ether at 54.219 MHz using calibrated 90°
pulses. An acquisition time of 0.050 s was used and 4032 points were
collected. A linear prediction based on the first 256 points of the FID
was used to back-calculate the first two points to avoid a roll in the
baseline. Integral regions of ca. (5 times the peak width at half-height
were used, using identical regions for corresponding peaks from the
standard and the reisolated samples. 13C spectra were obtained for these
samples, as before except at ambient temperature, immediately before
or after the 17O spectra.
CStandard
CSample
OSample
OStandard
R/R0 )
×
(9)
Results from All Reactions. The integrations for the relevant peaks
in cyclohexenone or 3-butylcyclohexanone were set relative to integra-
tions of 1000 for C6 of cyclohexenone or 1000 for Cc of 3-butylcy-
clohexanone, and the averaged results from five spectra for each sample
are shown in Tables 3 and 4. The relative changes in isotope
composition (R/R0) at each position in cyclohexenone were calculated
as the ratio of average integrations in Table 3 relative to the standard.
The standard deviations (∆R/R0) were calculated from eq 2. In these
Acknowledgment. We thank NIH grant No. GM-45617 and
The Robert A. Welch Foundation for support and NSF grant
No. CHE-9528196 and the Texas A&M University Supercom-
puting Facility for computational resources. We also thank James
Snyder for helpful discussions and Eiichi Nakamura for sharing
preliminary results which led to the calculations on 15 and 16.
Supporting Information Available: Energies and full
geometries of all calculated structures and equations for
spreadsheet simulation of eq 1 (PDF). This material is available
free of charge via the Internet at http://pubs.acs.org.
2
2 1/2
∆IntSample
IntSample
∆IntStandard
IntStandard
∆R/R0 ) R/R0 ×
+
(2)
((
)
(
) )
equations IntSample is the average integration for each carbon in the
sample, IntStandard is the average integration for each carbon in the
JA993373C