Isotope effects and the distinction between synchronous, asynchronous, and stepwise Diels-Alder reactions

Article abstract of DOI:10.1016/S0040-4020(01)00354-4

A variety of symmetrical or nearly symmetrical Diels-Alder reactions are studied by a combination of experimental isotope effects, theoretical calculations, and rate observations. Becke3LYP calculations predicted highly asynchronous transition structures for Diels-Alder reactions of bis(boryl)acetylenes, dialkyl acetylenedicarboxylates, triazolinediones, and dialkyl maleates. Rate observations and kinetic isotope effects are consistent with these predictions, though the experimental support for the calculated structures is notably ambiguous in some cases. A concerted mechanism is supported for the retro-Diels-Alder reaction of norbornene. Overall, Diels-Alder reactions appear to be only very weakly biased toward synchronous transition states.

Full text of DOI:10.1016/S0040-4020(01)00354-4

TETRAHEDRON
Pergamon
Tetrahedron 57 32001) 5149±5160
Isotope effects and the distinction between synchronous,
asynchronous, and stepwise Diels±Alder reactions
Daniel A. Singleton,p Brian E. Schulmeier, Chao Hang, Allen A. Thomas, Shun-Wang Leung
and Steven R. Merrigan
Department of Chemistry, Texas A & M University, P.O. Box 30012, College Station, TX 77842-3012, USA
Dedicated to Professor Barry M. Trost and his inspiring enthusiasm for chemistry on the occasion of his 60th birthday
Received 5 February 2001; revised 28 February 2001; accepted 1 March 2001
AbstractÐA variety of symmetrical or nearly symmetrical Diels±Alder reactions are studied by a combination of experimental isotope
effects, theoretical calculations, and rate observations. Becke3LYPcalculations predicted highly asynchronous transition structures for
Diels±Alder reactions of bis3boryl)acetylenes, dialkyl acetylenedicarboxylates, triazolinediones, and dialkyl maleates. Rate observations
and kinetic isotope effects are consistent with these predictions, though the experimental support for the calculated structures is notably
ambiguous in some cases. A concerted mechanism is supported for the retro-Diels±Alder reaction of norbornene. Overall, Diels±Alder
reactions appear to be only very weakly biased toward synchronous transition states. q 2001 Elsevier Science Ltd. All rights reserved.
1. Introduction
It is widely accepted that most Diels±Alder reactions occur
by concerted mechanisms, and the common understanding
of the facility of Diels±Alder reactions centers on the idea
of an aromatic transition state with synchronous formation
of the two new carbon±carbon bonds. In the past there has
been some controversy over these ideas, with highly
asynchronous `biradicaloid' transition states or stepwise
mechanisms being favored at times by some groups.1
However, the view that Diels±Alder reactions intrinsically
prefer synchronous transition states has prevailed in later
studies. Reactions of unsymmetrical dienes or dienophiles
necessarily involve unsymmetrical transition states, but
Diels±Alder reactions of symmetrical addends are thought
to usually involve synchronous or nearly synchronous C±C
bond formation. The evidence for this proposition has
come mainly from ab initio calculations,2±4 substituent
effects on the rate,5 and deuterium kinetic isotope effects
3KIEs).6
calculations predict a synchronous concerted transition
state for this reaction,9 and Zewail and Houk have
Recently the argument over the nature of Diels±Alder
reactions was brie¯y rekindled by Zewail and coworkers'
femtosecond time-resolved observation of an intermediate
in the retro-Diels±Alder reaction of norbornene.7 Notably
this intermediate's short lifetime was thought to allow
found that the femtosecond norbornene observations are
consistent with reactions on the excited-state energy
surface.10
We have been intrigued by calculations predicting highly
asynchronous transition states for some Diels±Alder
reactions,11±13 and we present here our studies of several
symmetrical or nearly symmetrical Diels±Alder reactions.
A key tool in these investigations is the comparison of
a
stereospeci®c stepwise mechanism.8 Nevertheless,
Keywords: Diels±Alder reaction; isotope effect; asynchronous transition
state.
2
experimental and theoretically predicted H and 13C KIEs.
p
Corresponding author. Tel.: 11-409-845-9166; fax: 11-409-845-4719;
e-mail: singleton@chemvx.tamu.edu
In several studies we have recently shown that high-level
0040±4020/01/$ - see front matter q 2001 Elsevier Science Ltd. All rights reserved.
PII: S0040-4020301)00354-4

5150
D. A. Singleton et al. / Tetrahedron 57 -2001) 5149±5160
calculations can lead to excellent predictions of heavy-atom
KIEs and reasonably good predictions of deuterium KIEs
for reactions not involving hydrogen transfer, provided that
the calculated transition structure more or less accurately
re¯ects reality.14 In this way KIEs can be used to provide
an experimental basis for calculated transition state
geometries. In the ®rst application of this tool with the
prototypical `nearly symmetrical' Diels±Alder reaction of
isoprene with maleic anhydride, a near-perfect prediction of
the complete set of experimental KIEs was possible from
the Becke3LYP-calculated transition structure.15 This
Diels±Alder adducts with acetic acid 3Eqs. 31)±33)).
ꢀ1†
ꢀ2†
observation provided strong support for
a
nearly
synchronous transition state, as well as supporting the
applicability of Becke3LYPcalculations to other Diels±
Alder reactions. In this paper, we experimentally and
calculationally probe a variety of reactions to distinguish
synchronous versus asynchronous versus stepwise
pathways. We ®nd that none of the mechanistic probes
used was consistently convincing by itself. Overall,
however, the expected preference for synchroneity
with symmetrical addends appears readily countered by a
number of factors. The results have implications toward our
general understanding of Diels±Alder reactions.
ꢀ3†
Surprisingly, however, 1 was less reactive than a mono-
activated acetylenic borane. For example, a competition
between 1 and a mono-BBN-activated dienophile16 for
reaction with butadiene had krel 3bis/mono)ˆ0.16 at
1308C. For symmetrical or nearly synchronous cyclo-
addition transition states, it might be expected that a second
activating group would provide similar activation to the
®rst.1c,5 While lesser activation by the second activating
group has usually been observed,18 the observation that a
single boryl group is highly activating while a second boryl
group is deactivating is highly unusual.19 This suggested
that Diels±Alder reactions of 1 do not proceed by a
symmetrical transition state.
2. Results and discussion
2.1. Bisꢀ9-borabicyclo[3.3.1]nonyl)acetylene
Our interest in unsymmetrical Diels±Alder reactions
involving symmetrical dienophiles was initiated by an
unusual observation with a `doubly activated' dienophile.
We have previously found that mono-dialkylboryl
acetylenes were moderately reactive but successful dieno-
philes,16 and bis39-borabicyclo[3.3.1]nonyl)acetylene 31)
was of interest owing to its expected higher reactivity.
The neat reaction of bis3tributylstannyl)acetylene with two
equivalents of 9-bromo-9-BBN was found to afford 1
rapidly at 2788C. Although 1 could not be isolated, support
Density function theory calculations support this idea
3Fig. 1). Becke3LYPcalculations with a 6-31G basis set
p
1
for its formation comes from the H NMR observation of
predict a highly unsymmetrical transition structure 33) for
the reaction of 1,3-butadiene with the parent bis3boryl)-
acetylene. The symmetrical structure 4, found by forcing
Cs symmetry, is a second-order saddle point exhibiting
two imaginary frequencies, and 4 is predicted to be
6.5 kcal/mol higher in energy than 3. In concert with the
experimental observation, the barrier for reaction of
bis3boryl)acetylene is predicted to be higher than for
almost quantitative formation of Bu3SnBr, the observation
of BBN-group peaks similar to other BBN-alkynes,16,17 and
the subsequent reactivity of the solution. This convenient
generation of 1 from commercially available materials
provides an effective dienophileÐreaction with simple
acyclic dienes occurs at 1108C and good yields of 1,4-cyclo-
hexadienes are obtained after deboronation of the presumed
Figure 1. The calculated unsymmetrical transition structure 33) and symmetrical second-order saddle point 34) for the Diels±Alder reaction of 1,3-butadiene
with bis3boryl)acetylene. Relative energies versus starting materials 3including zpe) are given in kcal/mol.

D. A. Singleton et al. / Tetrahedron 57 -2001) 5149±5160
5151
reaction of the mono-activated HCCBH2 3transition
structure not shown) by 3.4 kcal/mol.20
synchronous transition states.1c However, the pattern here
is typical of other Diels±Alder reactions,18 and no clear
interpretation of the synchroneity of the transition state is
possible from these data alone.23
Structure 3 is a [4 atom13 atom] transition structure as
previously predicted for Diels±Alder reactions of both
vinylboranes and acetylenic boranes.16,21 The intimate
involvement of the boron atom in the transition state for
boron-activated Diels±Alder reactions has been explained
on an FMO basisÐthe LUMO of a vinyl- or alkynylborane
has a large coef®cient on the boron atom for interaction with
the diene HOMO. The HOMO±LUMO overlap is also
geometrically highly favorable as the distance between C1
and C4 on a 1,3-diene is comparable to the distance between
the boron atom and terminal alkynyl carbon on an acetylenic
borane. However, a [413] cycloaddition cannot be
consummated, as it does not lead to a stable product.
After the transition state, IRC analysis has the C4±C5
bond distance decreasing rapidly. This transforms the
[413] transition structure into the stable [4 atom12 atom]
cycloadduct.
We next probed the synchroneity of these reactions with
kinetic isotope effects. Because a symmetrical Diels±
Alder reaction necessarily exhibits symmetrical isotope
effects due to averaging, a `moderately unsymmetrical'
Diels±Alder reaction 3see Ref. 6 for a discussion of
this strategy) was required. The reaction of diethyl or
dimethyl acetylenedicarboxylate 36, RˆMe or Et) with
isoprene 35) was chosen for study 3Eq. 34)).6 Isotope effects
for the uncatalyzed reaction of isoprene with 6 were
determined combinatorially at natural abundance by
NMR methodology.24 Reactions of isoprene on a 0.5-mol
scale in toluene at 508C were taken to 84^2% and 83^2%
conversion using limiting diethyl acetylenedicarboxylate
and 75^3% conversion using dimethyl acetylenedi-
carboxylate. The unreacted isoprene, recovered by
successive crude and fractional distillations, was
analyzed by 13C and 2H NMR along with a standard
sample of the isoprene not subjected to the reaction
conditions. The changes in 13C and 2H composition
were then calculated using the methyl group as an
internal standard with the assumption that its isotopic
composition does not change during the reaction. From the
changes in isotopic composition the isotope effects were
calculated as previously described.24
The alternative structure 4 could be described as a
[4 atom14 atom] transition structure as C1 and C4 of
the diene are nearly as closely associated with the boron
atoms as with the alkynyl carbons to which they would
ultimately bond. Such a [414] structure might be
expected from FMO theory because the LUMO of a
planar bis3boryl)acetylene has larger coef®cients on the
boron atoms than on carbon.22 However, the geometry
for HOMO±LUMO overlap would not be favorableÐ
the p orbitals for the two boron atoms cannot overlap
well with the p orbitals of C1 and C4 at the same time.
Instead, the second boryl group twists 908 from the ®rst
to have its empty p orbital conjugated with the second
p bond of the alkyne. This rationalizes the lower reac-
tivity of 1 compared to a monoborylalkyne.
ꢀ4†
2.2. Dialkyl acetylenedicarboxylates
The Diels±Alder reactions of 1 have the appearance of
being a `special case,' and carbonyl-activated dienophiles
would not be expected to behave similarly. It was therefore
surprising to note the recent calculational prediction by
Morokuma and coworkers of a highly unsymmetrical
transition structure for the reaction of acetylenedicarboxylic
acid with cyclopentadiene.13 We set out to assay the
synchroneity of Diels±Alder reactions of carbonyl-
activated alkynes in detail.
The results are summarized in Table 1. The 13C KIE at C1 is
larger than that observed at C4, and the inverse H KIEs
observed for H1in and H1out are more pronounced than those
for H4in and H4out. However, the differences in 1-position
versus 4-position isotope effects are similar to those
2
Table 1. Experimental isotope effects 3k12C/k13C or kH/kD, 508C) for the
Diels±Alder reaction of isoprene with 6
Experimenta
As with 1, rate observations provided a starting point.
Dimethyl acetylenedicarboxylate is notably less reactive
than dimethyl fumarate, the former requiring prolonged
heating with dienes that react at room temperature with
the latter. To study the effect of having one activating
substituent versus two in carbonyl-activated acetylenes,
competition reactions were carried out with mono- and
bis-activated dienophiles in their Diels±Alder reactions
with butadiene. For dimethyl acetylenedicarboxylate/
methyl propiolate, krel 3bis/mono)ˆ13 31008C), while for
3-hexyn-2,5-dione/3-butyn-2-one, krel 3bis/mono)ˆ20
3258C). The second carbonyl in each case certainly has
much less activating effect than the ®rst one did. This
does not ®t with expectations at the extreme of purely
1
2
3
C1
C2
C3
C4
H1in
H1out
H3
H4in
H4out
1.023 33)
1.000 32)
1.000 31)
1.016 33)
0.936 310)
0.955 34)
0.988 37)
0.967 36)
0.980 36)
1.021 32)
1.000 33)
0.999 32)
1.016 32)
0.935 311)
0.963 33)
0.992 35)
0.968 39)
0.983 36)
1.020 33)
1.001 32)
0.999 32)
1.014 32)
a
Experiments 1, 2 and 3 were taken to 84^2%, 83^2%, and 75^3%
conversion, respectively. Experiments 1 and 2 used diethyl acetylene-
dicarboxylate while experiment 3 used dimethyl acetylenedicarboxylate.
Deuterium isotope effects were not measured for experiment 3.

5152
D. A. Singleton et al. / Tetrahedron 57 -2001) 5149±5160
Figure 2. Transition structures for the Diels±Alder reaction of acetylenedicarboxylic acid with isoprene. Relative energies versus 10 3Becke3LYP/6-
31Gp1zpe) are given in kcal/mol and estimated relative entropies 3258C) are in e.u.
observed in the reaction of isoprene with maleic
anhydride,24 a reaction concluded to involve a nearly
synchronous transition state.15 The isotope effects are
indicative of greater bond formation to C1 than C4 at the
transition state, but they also qualitatively suggest relatively
modest asynchroneity in the formation of the two new C±C
bonds.
The predominance of structures such as 8 with greater bond
formation to C1 of isoprene was a key assumption in a
previous study.6 However, if 9 contributes signi®cantly
the observed isotope effects will not re¯ect the asynchro-
neity of the reaction.
In order to quantitatively interpret the isotope effects,
theoretical calculations were carried out on the model
reaction of acetylenedicarboxylic acid with isoprene. The
four transition structures 10±13 were found for this reaction
in Becke3LYPcalculations with a 6-31G p basis set 3Fig. 2).
Each is highly unsymmetrical, and the transition structures
arise from two pairs of regioisomerically unsymmetrical
structures where the carbonyl adjacent to the longer of the
two forming C±C bonds may be directed either endo or exo.
The structures 10 and 11 with more advanced bonding to C1
of isoprene are predicted to be enthalpically favored, but
entropy estimates at 258C based on the unscaled vibrational
frequencies favor the regioisomerically opposite transition
structures 12 and 13.
This interpretation must be tempered by a more careful
consideration of the isotope effects expected if a highly
asynchronous transition state is preferred for these
reactions. In such a case, two transition states may
contribute to the observed isotope effectsÐone in which
bond formation to C1 is advanced over that to C4 3e.g. 8)
and one in which bond formation to C4 is advanced 3e.g. 9).
KIEs were predicted for each of the transition structures 10±
13 by the method of Bigeleisen and Mayer25 from the scaled
theoretical vibrational frequencies,26 and tunneling
corrections were applied using the one-dimensional in®nite

D. A. Singleton et al. / Tetrahedron 57 -2001) 5149±5160
5153
Table 2. Theoretically predicted isotope effects 3k12C/k13C or kH/kD, 508C)
for the Diels±Alder reaction of isoprene with acetylenedicarboxylic acid
pathways also highlights the dif®culty of interpreting rate
ratios for mono- versus bis-activated dienophiles. The
calculations predict a 1.0 kcal/mol lower barrier for the
reaction of 1,3-butadiene with acetylenedicarboxylic acid
than with propiolic acid, in good agreement with the relative
rate observations above with dimethyl acetylenedicarboxy-
late/methyl propiolate. However, this cannot be taken as
favoring an asynchronous over a synchronous transition
state, as the barrier for formation of the synchronous
structure 3only 0.1 kcal different) ®ts just as well with
experiment.
10
11
12
13
Weighted
average for
10±13a
C1
C2
C3
C4
H1in
H1out
H3
H4in
H4out
1.030
1.003
1.001
1.004
0.901
0.900
0.965
0.965
1.003
1.029
1.003
1.000
1.006
0.908
0.906
0.964
0.965
1.000
1.015
1.004
1.003
1.023
0.953
0.991
0.987
0.938
0.940
1.008
1.001
1.001
1.025
0.970
1.001
0.993
0.936
0.920
1.022
1.003
1.001
1.013
0.928
0.944
0.975
0.953
0.972
2.3. Triazolinediones
1,2,4-Triazoline-3,5-diones 3TADs) are highly reactive
dienophiles.28 In a study of the Diels±Alder reaction of
4-phenyl-1,2,4-triazoline-3,5-dione 3PTAD) with isomeric
2,4-hexadienes, Jensen and Foote observed a lack of stereo-
speci®city with the Z,Z isomer. In addition, when the
reactions were carried out in methanol with either the Z,Z
or Z,E isomers, products indicative of intermediate trapping
were observed. No trapping was observed for the E,E
isomer. It was concluded that the Z,Z and Z,E isomers
react by stepwise mechanisms involving aziridinium
imides, while it could not be decided whether dienes that
more readily obtain an s-cis conformation involve stepwise
or concerted processes. Recent calculations by Houk and
Foote have predicted that the concerted mechanism should
be favored when the s-cis conformation is not sterically
precluded.12 However, the best transition structure found
in Becke3LYP/6-31Gp calculations for the reaction of the
parent TAD with 1,3-butadiene was 14. This highly unsym-
metrical and asynchronous transition structure was
predicted to be favored by only 1.1 kcal/mol over the
second-order saddle point 15 obtained by forcing Cs
symmetry. The asynchroneity of 14 was analogous with
asynchroneity previously predicted for the Diels±Alder
reaction of butadiene with cis-diazene.11
parabolic barrier model.27 The results are shown in Table 2.
The asynchroneity of C±C bond formation in 10±13 is
re¯ected in predicted isotope effects for the individual
structuresÐeven the relatively synchronous structure 12
is predicted to exhibit a substantial difference in KIEs at
C1 and C4. Because 10±13 are comparable in energy,
each should contribute to the observed KIEs.
Unsurprisingly, then, none of the sets of KIEs for the
individual structures compare well with experiment. Table
2 shows that a much better prediction of the experimental
KIEs comes from a Boltzmann-weighted average of the
KIEs predicted for 10±13 based on the energies and
entropies of Fig. 2. In this case the predicted carbon KIEs
are in good agreement with experiment, while the weighted
KIEs underpredict the observed deuterium KIEs by 1±2%.
The tendency to underpredict experimental deuterium KIEs
has been previously observed with the Claisen rearrange-
ment, one possible explanation being an underprediction of
tunneling by the one-dimensional tunneling correction
employed.14d
In considering the weighted predicted KIEs it should be
noted that the predictions would change substantially with
small changes in the relative energies or entropies of 10±13,
and these relative energies and entropies are not reliable to
such a degree. A change in the weighting could make the
prediction better 3for example, a small increase in contribu-
tion from 13 would bring the predicted 13C KIEs into essen-
tially perfect alignment with experiment) or worse. The
possibility of fortuity in the agreement of predicted and
experimental KIEs should thus be recognized. The general
possibility for multiple contributing transition states limits
the certainty of conclusions based on the isotope effects for
Diels±Alder reactions. Nonetheless, the experimental
isotope effects are certainly consistent with the highly asyn-
chronous theoretical structures.
Because TADs are also highly reactive in ene reactions, a
diene lacking allylic hydrogens, 2-t-butyl-1,3-butadiene
316), was chosen for the study of the isotope effects in
TAD Diels±Alder reactions. Reactions of 16 with PTAD
in CH2Cl2 at 2788C afforded the Diels±Alder adduct 17
quantitatively. The unreacted 16 from reactions taken to
71^3% and 70^3% conversion was recovered by vacuum
transfer followed by fractional distillation and analyzed
compared to a standard sample 3vide supra) using the
methyl groups as the internal standard for the NMR
integrations.
It is important to note that the calculational support for a
highly asynchronous transition state is very weak by itself.
For the reaction of 1,3-butadiene with acetylenedicar-
boxylic acid, BeckeLYP/6-31Gp calculations do predict
highly asynchronous transition structures 3not shown, but
very similar to 10 and 11). However, a synchronous
second-order saddle point obtained by enforcing Cs symme-
try is only 0.1 kcal/mol higher in energy. Such a small
energy difference is in no way reliable. This low difference
in predicted barriers for asynchronous versus synchronous

5154
D. A. Singleton et al. / Tetrahedron 57 -2001) 5149±5160
Table 3. Experimental and theoretically predicted isotope effects 3k12C/
k13C or kH/kD, 2788C) for the Diels±Alder reaction of 2-t-butyl-1,3-buta-
diene with PTAD
Experimentala
2
Predicted
Best-®t
averageb
1
18
19
15
C1
C2
C3
C4
1.023 34) 1.025 33) 1.043
1.007 34) 1.003 33) 1.009
1.001 33) 1.001 33) 1.000
1.022 33) 1.021 33) 1.007
1.006
1.025
1.005
1.001
1.023
0.862
0.931
0.938
0.855
0.928
1.028
1.002
1.002
1.028
0.805
0.906
0.952
0.805
0.906
1.002
1.004
1.0045
0.929
1.013
0.937
0.775
0.821
H1in 0.88 32)
H1out 0.95 32)
H3
H4in 0.86 31)
H4out 0.92 31)
0.89 32)
0.95 32)
0.98 31)
0.88 31)
0.94 32)
0.807
0.863
0.939
0.921
1.016
0.96 31)
a
Experiments 1 and 2 were taken to 71^3% and 70^3% conversion,
respectively.
Based on a 55:45 mixture of 18 and 19.
b
Instead of a mixture of highly unsymmetrical transition
structures, could a single nearly symmetrical transition
structure account for the observed isotope effects? To
address this question, we have calculated isotope effects
for the Houk and Foote second-order saddle point 15. The
predicted isotope effects for 15 are necessarily symmetrical
but their magnitudes show substantial differences with
experiment. The C1/C4 KIEs predicted for 15 are a bit
high at 1.028, and the predicted H1out/H4out KIEs at 0.906
are 2±4% low, but the most striking difference is observed
with H1in/H4in, in which the prediction for 15 is 6±9%
below the experimentally observed values. Although the
experimental 1-position and 4-position isotope effects in
the reaction of PTAD with 16 are very similar, we conclude
that they best support a mixture of highly asynchronous
transition states in this reaction.
2.4. Dialkyl maleates
An intriguing observation in Diels±Alder reactions is the
great reactivity difference between dialkyl fumarates and
dialkyl maleates. The former react at room temperature
with simple acyclic dienes while the latter require prolonged
heating at <1008C. This observation of greater reactivity in
trans-1,2-disubstituted alkenes over the corresponding cis
isomers is a normal observation in concerted cyclo-
additions.30 However, a factor in this trend when there is
an activating carbonyl group on the alkene is that the
carbonyl group can be sterically inhibited from the ideal
conformation for activation. For this reason the reactivity
difference between trans and cis isomers is greater for
carbonyl-activated dienophiles than for those where confor-
mational inhibition of activation is not applicable. For
example, cis- and trans-dicyanoethylene are nearly equally
reactive in Diels±Alder reactions.31 It was therefore of
interest whether dialkyl maleates adopt a synchronous
Diels±Alder transition state with both carboxylates
activating the cycloaddition equally or instead prefer to
activate the transition state with a single carboxylate,
allowing the other to rotate into a sterically unencumbered
position.
The isotope effects obtained in this way are summarized in
Table 3. The most striking observation in the reaction of 16
with PTAD is the similarity of the 1-position and 4-position
isotope effects, with the H1in and H1out KIEs differing very
little from those observed for H4in and H4out. Qualitatively
these results suggest very little asynchroneity in this
reaction.
Theoretical calculations suggest a very different interpreta-
tion of the isotope effects. The transition structures 18 and
19 were obtained in Becke3LYPcalculations with a 6-31G p
basis set.29 Each exhibits very asynchronous C±C bond
formation. Interestingly, 18 and 19 are very nearly equal
in energy 3within 0.05 kcal/mol at the Becke3LYP/6-31Gp
level). The isotope effects predicted for these structures are
given in Table 3. Each structure by itself would be predicted
to exhibit quite differing 1-position and 4-position isotope
effects, but a mixture of the two provides a very good
prediction of the experimental isotope effects. A best-®t
weighting of 55% of the predicted KIEs for 18 with 45%
of those for 19 gives 13C KIEs that are within experimental
error of those observed and 2H KIEs that are in very
reasonable agreement with the experimental values 3with
a continued tendency, as discussed above, to slightly under-
Theoretical study of the reaction of 1,3-butadiene with
maleic acid was complicated by the possibility for involve-
ment of a large number of structures, depending on whether
the transition state was endo or exo and the relative
2
predict the experimental H values).

D. A. Singleton et al. / Tetrahedron 57 -2001) 5149±5160
5155
Figure 3. The best calculated unsymmetrical transition structure 320) and symmetrical second-order saddle point 321) for the Diels±Alder reaction of 1,3-
butadiene with maleic acid. Relative energies versus starting materials 3including zpe) are given in kcal/mol.
orientation of the carbonyl and hydroxyl groups for each
acid group. These possibilities were systematically investi-
gated in RHF 6-31Gp/3-21G calculations and a total of 10
transition structures or second-order saddle points were
found. Of these the lower-energy unsymmetrical and Cs-
symmetric structures were reoptimized in Becke3LYP/6-
31Gp calculations.
tive interpretation is similarÐthe KIEs suggest relatively
modest asynchroneity in the formation of the two new C±C
bonds.
Once again, however, it is important to consider the possible
contribution of two regioisomeric highly asynchronous tran-
sition states to the observed KIEs. To do this, the two
transition structures 22 and 23, corresponding to 20 with
the addition of a methyl group in either of two possible
positions, were calculated for the reaction of isoprene with
maleic acid 3Fig. 4). Structure 22 with advanced bonding to
C1 would be expected to be favored, and it is, but by a
predicted amount of only 0.3 kcal/mol. It would thus be
expected that both isomers would contribute to the observed
isotope effects.
The highly unsymmetrical endo transition structure 20
3Fig. 3) is predicted to be lowest in energy for this reaction.
Carbon±carbon bond formation in this structure is highly
Ê
asynchronous, with a 0.55 A difference between the length
of the two forming bonds. Notably, only one of the carboxyl
groups is oriented to activate the cycloaddition, the other
being turned away from aligning its p orbitals with the
shorter of the two forming bonds. The best synchronous
structure 321), a second-order saddle point obtained by
forcing Cs symmetry, is predicted to be 1.6 kcal/mol higher
in energy.
The calculated isotope effects for 22 and 23 are shown in
Table 4, along with a weighted average predicted from a
Boltzmann distribution of the two using the calculated
energy difference. The isotope effects predicted for 22 and
23 individually do not resemble the observed values.
However, the Boltzmann weighted average of the two
provides a very good prediction of the observed KIEs: all
of the 13C KIEs and three of the ®ve 2H KIEs are predicted
Isotope effects were determined for the reaction of dimethyl
maleate with isoprene in xylenes at 80±1008C.32 The
unreacted isoprene from reactions taken to 98.0^0.2%
and 94^1% conversion was recovered by vacuum transfer
followed by fractional distillation and analyzed compared to
a standard sample 3vide supra) using the methyl group as the
internal standard for the NMR integrations.
2
within experimental error, and the other two H KIEs are
slightly underpredicted 3as discussed above). Considering
the differences between the gas-phase calculated reaction
and the solution-phase experimental reaction, as well as
the likely contribution of exo transition states to the
observed KIEs, the values predicted for 22123 agree
The resulting KIEs 3Table 4) closely resemble those found
above with acetylenedicarboxylates. As such, their qualita-
Table 4. Experimental and theoretically predicted isotope effects 3k12C/k13C or kH/kD, 1008C) for the Diels±Alder reaction of isoprene with dimethyl maleate
Experimentala
Predicted for
Weighted
averageb
Predicted for
1
2
22
23
21
C1
C2
C3
C4
H1in
H1out
H3
H4in
H4out
1.021 32)
1.006 37)
1.003 33)
1.017 35)
0.945 310)
0.975 38)
0.999 38)
0.952 35)
0.978 35)
1.021 34)
1.004 32)
1.003 34)
1.018 33)
0.954 38)
0.980 311)
0.995 37)
0.960 36)
0.988 39)
1.028
1.003
1.001
1.009
0.937
0.938
0.978
0.975
1.005
1.010
1.002
1.002
1.028
0.982
1.008
0.991
0.933
0.945
1.020
1.003
1.001
1.016
0.955
0.966
0.983
0.958
0.980
1.018
1.002
1.002
1.018
0.937
0.969
0.982
0.937
0.969
a
Experiments 1 and 2 were taken to 98.0^0.2% and 94^1% conversion, respectively.
A 59.3:40.7 weighting, based on a Boltzmann distribution of predicted isotope effects for 22 and 23.
b

5156
D. A. Singleton et al. / Tetrahedron 57 -2001) 5149±5160
Figure 4. Transition structures for the Diels±Alder reaction of maleic acid with isoprene. Relative energies 3Becke3LYP/6-31Gp1zpe) are given in kcal/mol
and estimated relative entropies 3258C) are in e.u.
quite well with the experimental values. Clearly, the
observed KIEs are consistent with the predicted highly
asynchronous transition states.
simple unactivated Diels±Alder reactions. Houk and
coworkers have studied both concerted and stepwise
mechanisms for this reaction in 3U)B3LYPcalculations.
On the concerted pathway is the synchronous transition
state 24. The stepwise pathway involves sequential
transition states 25 and 26 for breaking the s bonds of
norbornene, with a biradical intermediate in between. The
3U)B3LYP/6-31Gp calculations favor the concerted path-
way by over 12 kcal/mol.
Unfortunately, the KIEs are also well consistent with a
nearly synchronous transition state. This can be seen from
the predicted KIEs for the symmetrical structure 21 in Table
4. These KIEs do not match as well with the observed KIEs
as the weighted average from 22123. Even a small amount
of asynchroneity, however, would make most of the isotope
effects expected for a nearly synchronous transition state
closely resemble the experimental values. Due to the use
of isoprene experimentally instead of 1,3-butadiene in 21,
some asynchroneity must be expected. For this reaction,
isotope effects would appear to be a very dull tool for
distinguishing a synchronous from a highly asynchronous
transition state.
To address the question of concerted versus stepwise
mechanisms for this reaction experimentally, we attempted
to determine the KIEs for the retro-Diels±Alder reaction of
norbornene. This proved to be especially experimentally
challenging, and was successful only indirectly.
ꢀ5†
2.5. Retro-Diels±Alder reaction of norbornene
Zewail and coworkers' femtosecond time-resolved
observation of an intermediate in the retro-Diels±Alder
reaction of norbornene7 has reawakened interest in the
question of a concerted versus stepwise mechanism in
The decomposition of norbornene into ethylene and cyclo-
pentadiene 3Eq. 35)) proceeds slowly over the course of
several days at 1958C. In order to accomplish the reaction
practically on a large scale at atmospheric pressure, the
vapor pressure of the reaction was lowered by dissolution
of the norbornene at 0.25 M in diphenyl ether. Under these
conditions the product cyclopentadiene reacts with norbor-
nene to form a Diels±Alder adduct. To allow the loss of the
cyclopentadiene from the reaction while minimizing loss of
norbornene, the condenser for the re¯uxed reaction was kept
at <458C. In addition maleic anhydride was added to the
reaction mixture to trap the cyclopentadiene as it formed.33
Nevertheless, a signi®cant amount of the cyclopentadiene
adduct with norbornene was formed. Because of the side
reaction and the uncertainty in the effective conversion,34
the ole®nic carbons and hydrogens of norbornene were
ignored and the actual KIEs for the retro-Diels±Alder
reaction were not calculable. Nevertheless, it was possible
to evaluate the predicted KIEs for the concerted and
stepwise mechanisms by comparing them with the changes
in isotopic composition during the reaction.

D. A. Singleton et al. / Tetrahedron 57 -2001) 5149±5160
5157
Figure 5. Graphs of experimental R/R0 values for the retro-Diels±Alder reaction of norbornene versus KIEs predicted for transition structures 24, 25, and 26
3from left to right).
Reactions of norbornene on an <1 mol scale were taken to
78% total conversion 3retro-Diels±Alder1cyclopentadiene
adduct1loss of norbornene), and the unreacted norbornene
was recovered by a ¯ash distillation followed by fractional
distillation at atmospheric pressure in a solids-distillation
apparatus. The relative changes in isotopic composition
with the retro-Diels±Alder reaction of norbornene) but
they might not distinguish similar transition states.
Despite the uncertainties for the individual mechanisms,
there is no sign of a substantial intrinsic bias for synchro-
neity in Diels±Alder reactions. When any factor disfavors a
synchronous transition state, for example a possible steric
interaction of oxygens in 21, asynchronous distortion of the
transition state occurs readily. This is not really surprising,
because transition states are inherently highly ¯exible. At
the extreme of a stepwise Diels±Alder reaction, the energy
is at most only a few kcal/mol higher than the concerted
process. One of the most important classes of Diels±Alder
reactions, Lewis acid-catalyzed reactions, involves
extremely asynchronous transition states that could hardly
be called `aromatic'.35 The effect of aromaticity on the
Diels±Alder transition state is not an overriding factor in
determining transition state geometry.
2
3R/R0 values, calculated as the ratio of H or 13C in recov-
ered norbornene compared to original norbornene) were
then determined by NMR using the C7 carbon and Hanti
on C7 as standards for comparison.
The R/R0 values are compared graphically with the
predicted isotope effects in Fig. 5. Ideally, these plots should
be straight lines going through the origin of the graph at
31,1). The predicted isotope effects for the synchronous
concerted transition structure 24 correlate very well with
the R/R0 values; with uncertainties in the R/R0 values of
<^0.003 for 13C and ^0.02 for 2H every point falls within
experimental error of the best ®t line with R2ˆ0.95. In
contrast, the predicted isotope effects for the stepwise
transition structures 25 and 26 do not correlate as well,
and the best-®t lines pass farther from 31,1). The experimen-
tal observations thus support a concerted transition state.
4. Experimental
4.1. Data for compounds
4.1.1. Diels±Alder reactions of bisꢀ9-borabicyclo[3.3.1]-
nonyl)acetylene ꢀ1). Example procedure. To 1.34 g
36.7 mmol) of 9-bromo-9-BBN 3prepared by concentration
under vacuum of a commercial solution in CH2Cl2) in a
pressure tube at 2788C was added dropwise 1.87 g
33.1 mmol) of bis3tributylstannyl)acetylene. The resulting
mixture was warmed to 258C, and 208 mg 31.60 mmol) of
2-phenylbutadiene was added. The pressure tube was then
heated to 1008C for 16 h. After cooling to 258C, 5 mL of
THF and 3 mL 33.15 g, 52 mmol) of glacial acetic acid were
added and the mixture was stirred at 258C for 2 h. The
mixture was then partitioned between 50 mL of saturated
aqueous NaHCO3 and 30 mL of hexanes. The aqueous layer
was extracted with two 30 mL portions of hexanes, and the
combined hexane layers were rinsed with saturated aqueous
KF and dried 3Na2SO4). The solvent was removed on a
rotary evaporator and the residue was chromatographed on
a short silica gel column using hexanes as eluent to afford
169 mg 368%) of the known36 1-phenyl-1,4-cyclohexadiene
32) 3contaminated with trace amounts of biphenyl): 1H NMR
3CDCl3) d 7.60±7.20 3m, 5 H), 6.20 3br s, 1 H), 5.85 3m, 2 H),
3.15 3m, 2 H), 2.95 3m, 4 H); 13C NMR 3CDCl3) d 141.8,
133.9, 128.4, 127.0, 125.0, 124.5, 123.8, 121.8, 27.8, 27.1.
3. Conclusions
The Becke3LYPcalculations predict unsymmetrical, highly
asynchronous transition structures for the Diels±Alder
reactions of bis3boryl)acetylenes, dialkyl acetylenedicarb-
oxylates, triazolinediones, and dialkyl maleates. The
predicted preference for an asynchronous transition state
is substantial for bis3boryl)acetylenes, and this is the
strongest theoretical case. Supported by the unique
observation that the `doubly-activated' dienophiles are
less reactive than mono3boryl)acetylenes, it seems highly
likely that an asynchronous transition state is involved in
reality. In the other reactions, the theoretical preference for
asynchronous over synchronous is less pronounced, ranging
from 0.1 to 1.6 kcal/mol. This margin is less certain. The 13C
2
and H isotope effects are in each case consistent with the
asynchronous transition states, but they only really signi®-
cantly disfavor
a synchronous transition state with
triazolinediones. Like other mechanistic probes, isotope
effects have a limited resolutionÐit appears they can
distinguish concerted from stepwise cycloadditions 3as

5158
D. A. Singleton et al. / Tetrahedron 57 -2001) 5149±5160
Analogous procedures employing 2-tert-butyl-1,3-buta-
diene and myrcene with differing reaction times 3see Eqs.
31)±33)) afforded the known 1-tert-butyl-1,4-cyclohexa-
diene37 and 1-34-methyl-3-pentenyl)-1,4-cyclohexadiene,38
respectively, in 53 and 55% yields.
synthetic lot as used in the reaction. The NMR samples of
standard and recovered material were in each case prepared
2
identically. Both 13C and H NMR spectra were taken on
neat samples. A T1 determination by the inversion-recovery
method was carried out on each NMR sample.
The 13C spectra were recorded at 100.58 MHz with inverse
gated decoupling. Isoprene, 2-t-butyl-1,3-butadiene, and
norbornene spectra used 120, 240, and 120 s delays,
respectively, between calibrated pulses, with acquisition
times of 3.5, 10.5, and 3.9 s, respectively, collecting
158 336, 524 288, and 158 720 points, respectively, in
each case zero-®lled to 512 K before Fourier transforma-
4.1.2. Diels±Alder reactions of dialkyl acetylenedicarb-
oxylates. Example procedure. A mixture of 34.03 g
30.5 mol) of isoprene, 71.9 g 30.42 mmol) of diethyl
acetylenedicarboxylate, and 200 mL of toluene was heated
to 50^58C with stirring under N2. Aliquots were
periodically removed and the progress of the reaction was
1
analyzed by H NMR based on the ratio of the isoprene to
2
the product cyclohexadiene. After 289 h the conversion was
84^2%. A vacuum transfer of <20 mL of volatiles on a
water aspirator followed by two successive fractional
distillations afforded 3.4 g of isoprene with no detectable
tion. The H spectra for isoprene, 2-t-butyl-1,3-butadiene,
and norbornene were recorded at 61.395 MHz, using acqui-
sition times of 7, 6.98, and 2.276 s and additional delays of
0, 8, and 0 s between acquisitions. 13C integrations were
determined numerically using a constant region for each
peak that was <5 times the peak width at half height
1
impurities by H or 13C NMR.
2
4.1.3. Diels±Alder reaction of 2-t-butyl-1,3-butadiene
with PTAD. Example procedure. A mixture of 26.3 g
30.150 mol) of freshly prepared39 PTAD in 550 mL of
CH2Cl2 was added slowly to a mixture of 22.0 g
distance on either side of the peak. H integrations were
determined using a deconvolution program to determine
the area of integration under partially overlapping peaks.
A zeroth order baseline correction was generally applied,
but in no case was a ®rst-order 3tilt) correction applied.
From the relative integrations the isotope effects were calcu-
lated as previously described.24
30.200 mol)
of
2-tert-butyl-1,3-butadiene,
8.10 g
30.055 mol) of 1,2-dichlorobenzene 3used as GC internal
standard), and 110 mL of CH2Cl2 in a dry ice±acetone
bath. After 2 h the reaction was found to be 71.3^1.0%
complete by GC. Vacuum transfer of the volatiles from
the reaction mixture using a water aspirator followed by
fractional distillation afforded 3.8 g of recovered 2-tert-
butyl-1,3-butadiene 3bp 101±1028C, .98% pure by GC).
4.3. Calculational procedures
The structures reported here were obtained with Gaussian
9840 using the Becke3LYP41 hybrid HF±DFT method with
the 6-31Gp basis set.42 Vibrational frequency analyses were
carried out on all stationary points. The energies obtained
for 3, 4, 10, 11, 12, 13, 18, 19, 20, 21, 22 and 23 were,
including zpe, 2284.085421, 2421.725170, 2649.569011,
2649.569010, 2649.567547, 2649.567585, 2704.462371,
2704.462686, 2611.511281, 2611.508767, 2650.802797,
and 2650.802351, respectively. A ®le containing the
complete geometries and energies for all calculated
structures is available from the corresponding author on
request.
4.1.4. Diels±Alder reaction of isoprene with dimethyl
maleate. Example procedure. A mixture of 338.9 g
34.98 mol) of isoprene, 736.1 g 35.11 mol) of dimethyl
maleate, 1000 mL of xylenes, and 2.00 mL of n-heptane
3internal standard) was heated under a dry-ice condenser
to 80±1008C32 for 10 d, at which time GC analysis of an
aliquot showed the conversion to be 98.0^0.2%. Vacuum
transfer of <70 mL of volatiles from the reaction mixture
using a water aspirator followed by fractional distillation
afforded 3.9 g of isoprene with no observable impurities
1
by H or 13C NMR.
Acknowledgements
4.1.5. Retro-Diels±Alder reaction of norbornene.
Example procedure. A mixture of 82.9 g 30.88 mol) of
norbornene, 4 kg of diphenyl ether, 50 g 30.51 mol) of
maleic anhydride, and 3.6 g of diglyme was heated to
195^58C under a positive pressure of argon in a 5 L ¯ask
equipped with a re¯ux condenser at 45^28C. Aliquots were
periodically removed and the progress of the reaction was
analyzed by 1H NMR based on the ratio of the norbornene to
the internal standard diglyme. After 192 h the conversion
was 78^2%. An initial distillate of 30 mL collected under a
water aspirator was redistilled twice to afford 10 g of
We thank NIH grant # GM-45617 and The Robert A. Welch
Foundation for support.
References
1. 3a) Dewar, M. J. S.; Olivella, S.; Stewart, J. J. P. J. Am. Chem.
Soc. 1986, 108, 5771. 3b) Dewar, M. J. S.; Pierini, A. B. J. Am.
Chem. Soc. 1984, 106, 203. 3c) Dewar, M. J. S.; Olivella, S.;
Rzepa, H. S. J. Am. Chem. Soc. 1978, 100, 5650. 3d) Telan,
L. A.; Firestone, R. A. Tetrahedron 1999, 55, 14269.
2. 3a) Loncharich, R. J.; Brown, F. K.; Houk, K. N. J. Org. Chem.
1989, 54, 1129. 3b) Birney, D. M.; Houk, K. N. J. Am. Chem.
Soc. 1990, 112, 4127.
1
norbornene which showed no detectable impurities by H
or 13C NMR.
4.2. NMR measurements
3. 3a) Bernardi, F.; Bottoni, A.; Field, M. J.; Guest, M. F.; Hillier,
I. H.; Robb, M. A.; Venturini, A. J. Am. Chem. Soc. 1988, 110,
3050. 3b) Bach, R. D.; McDouall, J. J. W.; Schlegel, H. B. J.
Org. Chem. 1989, 54, 2931. 3c) Houk, K. N.; Lin, Y.-T.;
For each reaction the NMR spectra of the samples of
recovered isoprene, 2-t-butyl-1,3-butadiene, or norbornene
were compared with standard material from the same

D. A. Singleton et al. / Tetrahedron 57 -2001) 5149±5160
5159
Brown, F. K. J. Am. Chem. Soc. 1986, 108, 554. 3d). Houk,
K. N.; Loncharich, R. J.; Blake, J. F.; Jorgensen, W. L. J. Am.
Chem. Soc. 1989, 111, 9172. 3e). Houk, K. N.; Gonzalez, J.;
Li, Y. Acc. Chem. Res. 1995, 28, 81. 3f). Li, Y.; Houk, K. N. J.
Am. Chem. Soc. 1993, 115, 7478. 3g). Goldstein, E.; Beno, B.;
Houk, K. N. J. Am. Chem. Soc. 1996, 118, 6036.
4. Sakai, S. J. Phys. Chem. A 2000, 104, 922.
mono- and doubly-activated transition states one cannot
strictly expect the second substituent to have the same
activating effect as the ®rst. Only in extreme cases is the
interpretation clear.
24. Singleton, D. A.; Thomas, A. A. J. Am. Chem. Soc. 1995, 117,
9357.
25. 3a) Bigeleisen, J.; Mayer, M. G. J. Chem. Phys. 1947, 15, 261.
3b) Wolfsberg, M. Acc. Chem. Res. 1972, 5, 225.
5. Hancock, R. A.; Wood, Jr., B. F. J. Chem. Soc., Chem.
Commun. 1988, 351.
26. The calculations used the program quiver 3Saunders, M.;
Laidig, K. E.; Wolfsberg, M. J. Am. Chem. Soc. 1989, 111,
8989) with Becke3LYPfrequencies scaled by 0.9614 3Scott,
A. P.; Radom, L. J. Phys. Chem. 1996, 100, 16502).
27. Bell, R. P. In The TunnelEffect in Chemistry ; Chapman and
Hall: London, 1980; pp. 60±63.
6. Gajewski, J. J.; Peterson, K. B.; Kagel, J. R.; Huang, Y. C. J.
J. Am. Chem. Soc. 1989, 111, 9078.
7. Horn, B. A.; Herek, J. L.; Zewail, A. H. J. Am. Chem. Soc.
1996, 118, 8755.
8. For a countering view, see: Hrovat, D. A.; Fang, S.; Borden,
W. T.; Carpenter, B. K. J. Am. Chem. Soc. 1997, 119, 5253.
9. Beno, B. R.; Wilsey, S.; Houk, K. N. J. Am. Chem. Soc. 1999,
121, 4816±4826.
28. Gillis, B. T.; Hagarty, J. D. J. Org. Chem. 1967, 32, 330.
29. exo Transition structures were predicted by Houk and Foote to
be much higher in energy 3see Ref. 12) and were not
considered here.
10. Wilsey, S.; Houk, K. N.; Zewail, A. H. J. Am. Chem. Soc.
1999, 121, 5772.
30. Huisgen, R. Angew. Chem., Int. Ed. Engl. 1963, 2, 633.
31. Sauer, J.; Sustmann, R. Angew. Chem., Int. Ed. Engl. 1980, 19,
779.
11. McCarrick, M. A.; Wu, Y.; Houk, K. N. J. Am. Chem. Soc.
1993, 58, 3330.
12. Chen, J. S.; Houk, K. N.; Foote, C. S. J. Am. Chem. Soc. 1998,
120, 12303.
32. Early in the reaction the temperature was limited by the
boiling point of the reaction mixture. As isoprene reacted,
the boiling point rose and was limited at 1008C for the later
stages of the reaction. Because later stages of the reaction
weigh most heavily in the isotopic fractionation process
3see Ref. 24 and refs therein), the observed KIEs are best
considered as measured at <1008C.
33. This reaction with maleic anhydride was not expected to be
reversible 3see: Tobia, D.; Harrison, R.; Phillips, B.; White, T.
L.; DiMare, M.; Rickborn, B. J. Org. Chem. 1993, 58, 6701)
but based on the ineffectiveness of the trapping this
assumption may not have been correct.
13. Froese, R. D. J.; Coxon, J. M.; West, S. C.; Morokuma, K.
J. Org. Chem. 1997, 62, 6991.
14. 3a) Singleton, D. A.; Merrigan, S. R.; Liu, J.; Houk, K. N.
J. Am. Chem. Soc. 1997, 119, 3385±3386. 3b) DelMonte,
A. J.; Haller, J.; Houk, K. N.; Sharpless, K. B.; Singleton,
D. A.; Straûner, T.; Thomas, A. A. J. Am. Chem. Soc. 1997,
119, 9907±9908. 3c) Keating, A. E.; Merrigan, S. R.;
Singleton, D. A.; Houk, K. N. J. Am. Chem. Soc. 1999, 121,
3933. 3d) Meyer, M. P.; DelMonte, A. J.; Singleton, D. A.
J. Am. Chem. Soc. 1999, 121, 10865. 3e) Singleton, D. A.;
Hang, C. J. Am. Chem. Soc. 1999, 121, 11885. 3f) Singleton,
D. A.; Hang, C. J. Org. Chem. 2000, 65, 895.
34. In principle this could be allowed for by adjusting the %
conversion used to calculated isotope effects; for example,
using a nominal % conversion of 61% for the second reaction
above brings all of the isotope effects found for that reaction
within experimental error of those predicted for the concerted
reaction. However, such an adjustment is arbitrary and
unnecessary for the purpose of comparing the calculated
KIEs for the differing pathways with experiment.
15. Beno, B. R.; Houk, K. N.; Singleton, D. A. J. Am. Chem. Soc.
1996, 118, 9984±9985.
16. Singleton, D. A.; Leung, S.-W. J. Org. Chem. 1992, 57, 4796.
17. Leung, S.-W.; Singleton, D. A. J. Org. Chem. 1997, 62, 1955±
1960.
18. Ketone, ester, and nitrile substituents activate ethylene by a
factor of <105, while a second such substituent accelerates
reactions by <10±102. See Refs. 1c and 5.
35. Singleton, D. A.; Merrigan, S. R.; Thomas, A. A. Tetrahedron
Lett. 1999, 40, 639±642.
19. The steric effect of the BBN group does not appear to be
signi®cantly deactivating, as the bis3dimethylboryl)acetylene
was found to be much less reactive than 1.
36. Grisdale, P. J.; Regan, T. H.; Doty, J. C.; Figueras, J.;
Williams, J. L. R. J. Org. Chem. 1968, 33, 1116.
37. Singleton, D. A.; Martinez, J. P. Tetrahedron Lett. 1991, 32,
7365.
20. MP2/6-31Gp
calculations
identically
predict
an
unsymmetrical transition structure similar to 3, favored over
the symmetrical 4 by 2.6 kcal/mol. In the MP2 calculation the
barrier for reaction of bis3boryl)acetylene is 2.2 kcal/mol
higher than for reaction of HCCBH2.
38. Ono, N.; Kamimura, A.; Kaji, A. J. Org. Chem. 1988, 53, 251.
39. Adams, C. E.; Aguilar, D.; Hertel, S.; Knight, W. H.; Paterson,
J. Synth. Commun. 1988, 18, 2225.
40. Frisch, M. J.; Trucks, G. W.; Schlegel, H. B.; Scuseria, G. E.;
Robb, M. A.; Cheeseman, J. R.; Zakrzewski, V. G.;
Montgomery, J. A.; Stratmann, Jr., R. E.; Burant, J. C.;
Dapprich, S.; Millam, J. M.; Daniels, A. D.; Kudin, K. N.;
Strain, M. C.; Farkas, O.; Tomasi, J.; Barone, V.; Cossi, M.;
Cammi, R.; Mennucci, B.; Pomelli, C.; Adamo, C.; Clifford,
S.; Ochterski, J.; Petersson, G. A.; Ayala, P. Y.; Cui, Q.;
Morokuma, K.; Malick, D. K.; Rabuck, A. D.; Raghavachari,
K.; Foresman, J. B.; Cioslowski, J.; Ortiz, J. V.; Stefanov, B.
B.; Liu, G.; Liashenko, A.; Piskorz, P.; Komaromi, I.;
Gomperts, R.; Martin, R. L.; Fox, D. J.; Keith, T.; Al-Laham,
M. A.; Peng, C. Y.; Nanayakkara, A.; Gonzalez, C.;
Challacombe, M.; Gill, P. M. W.; Johnson, B.; Chen, W.;
21. Singleton, D. A. J. Am. Chem. Soc. 1992, 114, 6563.
22. Based on RHF/3-21G outer coef®cients of 0.39 for B versus
0.18 for C. In the ground-state conformation for bis3boryl)-
acetylene the BH2 groups are twisted by 908 from each other.
23. There is a general problem in many kinds of studies that have
interpreted the symmetry of transition states from the rates of
reaction of mono- and doubly-activated compounds. The
problem is that a mono-activated reaction 3e.g. the Diels±
Alder reaction of butadiene with methyl propiolate)
necessarily involves an unsymmetrical transition state. Thus,
whether or not the transition state for a doubly-activated
reaction is symmetrical, from the differing geometry of the

5160
D. A. Singleton et al. / Tetrahedron 57 -2001) 5149±5160
Wong, M. W.; Andres, J. L.; Gonzalez, C.; Head-Gordon, M.;
Replogle, E. S.; Pople, J. A. Gaussian 98, Revision A.6;
Gaussian: Pittsburgh, PA, 1998.
1991. 3d) In Density FunctionalMethods in Chemistry ;
Labanowski, J., Andzelm, J., Eds.; Springer: Berlin, 1991.
3e) Parr, R. G.; Yang, Y. Density-FunctionalTheory of
Atoms and Molecules; Oxford University: New York, 1989.
42. Hehre, W. J.; Radom, L.; Schleyer, P. v. R.; Pople, J. A. Ab
Initio Molecular Orbital Theory; 1986; Wiley: New York, 1986.
41. 3a) Becke, A. D. J. Chem. Phys. 1993, 98, 5648. 3b) Lee, C.;
Yang, W.; Parr, R. G. Phys. Rev. B 1988, 37, 785. For reviews
of density-functional methods, see: 3c) Ziegler, T. Chem. Rev.