Perturbations by phenyl on the 1,5-hydrogen shift in 1,3(Z)-pentadiene. Another chameleonic transition region?

Article abstract of DOI:10.1021/ja040106k

The acyclic 1,5-dienyl hydrogen shift is accelerated by radical-stabilizing phenyl substituents without regard to the type of position occupied in the 1,3(Z)-pentadiene system. 1-Phenyl-5-p-tolyl-1,3(Z)-pentadiene has a corrected energy of activation 5.8 kcal mol^-1 lower than that of the parent, while the 2-and 3-phenyl analogues, examined in cyclic systems specifically designed to obviate the otherwise general need for a thermochemical correction to the immediately precursory s-cis conformation, reveal stabilizing effects of 3.6 and 3.4 kcal mol^-1, respectively. These relatively small effects are consistent with a chameleonic conceptual scheme for the transition region.

Full text of DOI:10.1021/ja040106k

Published on Web 10/09/2004
Perturbations by Phenyl on the 1,5-Hydrogen Shift in
1,3(Z)-Pentadiene. Another Chameleonic Transition Region?
William von E. Doering,* Edmund J. Keliher, and Xin Zhao
Contribution from the Department of Chemistry and Chemical Biology, HarVard UniVersity,
Cambridge, Massachusetts 02138-2902
Received April 12, 2004; Revised Manuscript Received August 16, 2004; E-mail: doering@chemistry.harvard.edu
Abstract: The acyclic 1,5-dienyl hydrogen shift is accelerated by radical-stabilizing phenyl substituents
without regard to the type of position occupied in the 1,3(Z)-pentadiene system. 1-Phenyl-5-p-tolyl-1,3(Z)-
pentadiene has a corrected energy of activation 5.8 kcal mol-1 lower than that of the parent, while the 2-
and 3-phenyl analogues, examined in cyclic systems specifically designed to obviate the otherwise general
need for a thermochemical correction to the immediately precursory s-cis conformation, reveal stabilizing
effects of 3.6 and 3.4 kcal mol-1, respectively. These relatively small effects are consistent with a
chameleonic conceptual scheme for the transition region.
Introduction
The acyclic 1,5-dienyl hydrogen shift in penta-1,3(Z)-dienes,1
one of the more prominent among purely thermal rearrange-
ments, may provide another example of chameleonic behavior.
An earlier example, revealed experimentally in Cope’s concerted
rearrangement of hexa-1,5-dienes (Figure 1),2 has been given
theoretical support by the calculations of Hrovat, Beno, Lange,
Yoo, Houk, and Borden.3
The transition region of the acyclic 1,5-dienyl hydrogen shift
can be represented as a continuum between the two extremes
in Figure 2, one end being characterized by coplanarity of the
hydrogen atoms of the 1 and 5 methylene groups with a planar
pentadienyl skeleton,4 the other end, by their orthogonal
orientation. That the transition region in the paradigm is well
approximated by the central drawing in Figure 2 has been
supported by several theoretical investigations, to all of which
references are made in the publication of Hess and Baldwin.5
Under the chameleonic hypothesis, perturbation by radical-
stabilizing groups in positions 2 and/or 4 may move the
transition region toward the allyl end of the continuum (lower
drawing) or toward the pentadienyl end (upper drawing) when
occupying position 3. The core element responsible for the large
energy of concert of the rearrangement consists of the migrating
hydrogen atom being bonded by its 1s orbital to orbitals of the
Figure 1. Perturbations by phenyl groups on the activation parameters of
Cope’s rearrangement as a function of occupancy at active positions (red
“a”) in a hypothetical bis-allylic transition region and at active, secondary
radical positions (blue “a”) in a cyclohexadiyl transition region. Enthalpies
of activation are reported in kcal mol-1
.
giving and receiving 1 and 5 methylene groups.6,7 These can
be hybridized throughout the continuum between sp2 (planar,
trigonal) and sp3 (pyramidal, tetrahedral).
(1) Thorough reviews are found in Alubugin, I. V.; Monoharan, M.; Breiner,
B.; Lewis, F. D. J. Am. Chem. Soc. 2003, 125, 9329-9342, in Hess and
Baldwin,5 and in Dieter Hasselmann, Houben-Weyl, Methods of Organic
Chemistry; Thieme: Stuttgart, 1995; Vol. E 21d, 4421-4430.
(2) Doering, W. v. E.; Wang, Y.-h. J. Am. Chem. Soc. 1999, 121, 10112-
10118.
Crucial to the testing of the chameleonic model is a
comparison of radical-stabilizing groups, phenyl in this work,
at the 2- and 3-position. The effect is expected to be comparable
to that found in the Cope rearrangement disubstituted in
positions 1 and 4 in which the transition region has moved
(3) Hrovat, D. A.; Beno, B. R.; Lange, H.; Yoo, H.-y.; Houk, K. N.; Borden,
W. T. J. Am. Chem. Soc. 1999, 121, 10529-10537.
(4) This end of the continuum is unacceptably high in energy owing to the
hydrogen atom-hydrogen atom interpenetration of their van der Waals
radii and is not as easily acceptable as either of the two ends of the
continuum presented in Cope’s rearrangement. This end of the continuum,
perhaps better reflected in cyclopentadiene, hypothetically the dehydroge-
nation product of s-cis-1, 3(Z)-pentadiene, has an activation energy for
(6) The parent model for the 1,5-hydrogen shift is the transfer of a hydrogen
atom between methane and a methyl radical. A closer analogy would be
the uninvestigated transfer from the methyl group of propene to an allyl
radical.
(7) (a) Ma, X.-l.; Schobert, H. H. Ind. Eng. Chem. Res. 2003, 42, 1151-1161.
(b) Camaioni, D. M.; Autrey, S. T.; Salinas, T. B.; Franz, J. A. J. Am.
Chem. Soc. 1996, 118, 2013-2022.
intramolecular 1,5-hydrogen shift of Ea 25 kcal mol-1
.
(5) Hess, B. A., Jr.; Baldwin, J. E. J. Org. Chem. 2002, 67, 6025-6033.
9
14206
J. AM. CHEM. SOC. 2004, 126, 14206-14216
10.1021/ja040106k CCC: $27.50 © 2004 American Chemical Society

1,5-Hydrogen Shift in 1,3(Z)-Pentadiene
A R T I C L E S
atom and a pentadienyl radical in its U configuration can be
estimated at ∼48 kcal mol-1, a value that places the 1,5-dienyl
hydrogen shift among the most strongly concerted of the thermal
organic chemical rearrangements.
Perturbational Approach
A traditional approach to conceptualizing a transition region
relies on the response of a paradigm to perturbations by
substituents known to stabilize or destabilize simple intermedi-
ates. Recall the carbenium ion, the carbanion, and, as thought
relevant here, the free radical. In two unpublished dissertations,
one by Norbert Rohse,14 the other by Monika Schlief,15 both
under the guidance of the late Professor Wolfgang R. Roth, the
first systematic applications of the perturbational approach to
the 1,5-hydrogen shift have been conducted. Their studies have
been guided by the Dewar model of a planar pentadienyl radical
as transition region, in which polarization toward a pentadienyl
carbanion/proton, resonance contributor determines the response
to substituents.16 Of the two types of position, 1 and 5 and 3,
both active under this mechanistic hypothesis, substituents at
the 3-position have been the focus of Rohse’s investigations
(see the upper representation in Figure 2). Substitution at the
C3-position by F, OCH3, Cl, H, CN, CH3, C6H5, CF3, and
COOCH3 encompasses relative rate constants at 200 °C ranging
from 0.57 to 13.4 (H ) 1.0). The Arrhenius parameters for this
series are reproduced in Table SI-1 of the Supporting Informa-
tion. DFT-based predictions by Saettel and Wiest of energies
of activation for the s-cis conformations of 1,3(Z)-pentadiene
(Ea ) 32.9 kcal mol-1) and its 3-methoxy (Ea ) 32.4 kcal
mol-1), and 3-fluoro (Ea ) 33.5 kcal mol-1) derivatives17 cannot
be convincingly compared with the experimental results of
Rohse on related s-trans conformations of hexa-1,3(Z)-diene
(Ea ) 32.8 kcal mol-1),and its 3-methoxy (Ea ) 33.4 kcal
mol-1) and 3-fluoro (Ea ) 32.7 kcal mol-1) derivatives.14
Kinetic data and derived Arrhenius parameters for a large
number of alkyl-substituted acyclic 1,3(Z)-pentadienes have been
in the literature for several decades (see Figures 4, 5, and SI-
1). Their comparison with the unsubstituted paradigm might
be expected to provide one test of the chameleonic conceptual
scheme. However, the radical-stabilizing power of alkyl groups
is small at best (the difference in bond dissociation energies of
primary and secondary C-H bonds, for example, being 2.4 kcal
mol-1).18 Comparisons are further complicated by experimental
activation energies that relate to the almost invariably more
stable, hydrogen-shift-incompetent, s-trans conformations of the
1,3(Z)-pentadienes, not to the s-cis conformations, which are
uniformly accepted to be the shift-competent, immediate starting
points for the rearrangement.19 This complication can be seen
Figure 2. Two extreme representations of the transition region of the 1,5-
hydrogen shift: the upper depicting the active 3-position in a pentadienyl
extreme, the lower depicting the active 2- and 4-positions in an allyl extreme;
the upper showing C1 and C5 as pure sp2, the lower as sp3.
toward the allylic end of the continuum. The small effect, 3.7
kcal mol-1, is comparable to the extra stabilization of ∼3 kcal
mol-1 given to an allyl radical by phenyl (as in the cinnamyl
radical) in the more transparent cis-trans thermal rearrangement
of a double bond.8 That given to a pentadienyl radical by phenyl
in the 3-position can hardly be much greater but has not been
independently evaluated.
The acyclic version of the 1,5-dienyl hydrogen shift was first
uncovered by Enklaar in the thermal conversion of ocimene to
alloo¨cimene9 and generalized by Wolinsky in 1962.10 His
proposed mechanism, “a six-membered transition state in which
five carbon atoms lie in a plane or near the plane with the
migrating hydrogen above or below this plane,” was later
legitimized theoretically by Woodward and Hoffmann in 196511
and given elegant experimental confirmation by Roth, Ko¨nig,
and Stein in 1970.12c
An early kinetic study by Roth and Ko¨nig of the unsubstituted
paradigm, 1,3(Z)-pentadiene, its degeneracy broken by deuterium-
labeling, provided the Arrhenius parameters, Ea ) (36.1 ( 2.2)
kcal mol-1 and log A ) (11.34 ( 1.0).12a,13 An energy of concert
derived by comparing this experimental energy of activation
with that of a nonconcerted model consisting of a free hydrogen
kcal mol-1! Extrapolation of the values of kH/kD at the three temperatures
given (°C), 195.1°, 5.22; 200.0°, 4.97; 205.3°, 5.11, to 25 °C is not
justifiable. This footnote may be of interest to the several theoretical
chemists who have been stimulated by this apparently large deuterium
isotope effect.
(8) Doering, W. v. E.; Birladeanu, L.; Sarma, K.; Teles, J. H.; Kla¨rner, F.-G.;
Gehrke, J.-S. J. Am. Chem. Soc. 1994, 116, 4289-4297.
(9) Enklaar, C. J. Recl. TraV. Chim. Pays-Bas 1907, 26, 157-179.
(10) Wolinsky, J.; Chollar, B.; Baird, M. D. J. Am. Chem. Soc. 1962, 84, 2775-
2779.
(14) Rohse, N. Substituenteneffekte auf die 1,5-Pentadienylwasserstoff-ver-
schiebung, PhD Dissertation, Abteilung fu¨r Chemie der Ruhr-Universita¨t,
Bochum, Germany (Roth, W. R, research sponsor), 1986.
(15) Schlief, M. Substituenteneffekte auf die 1,5-Pentadienylwasserstoff-ver-
schiebung in Cycloheptatriene, PhD Dissertation, Abteilung fu¨r Chemie
der Ruhr-Universita¨t, Bochum, Germany (Roth, W. R, research sponsor),
1990.
(11) Woodward, R. B.; Hoffmann, R. J. Am. Chem. Soc. 1965, 87, 2511-2513.
(12) (a) Roth, W. R.; Ko¨nig, J. Liebigs Ann. Chem. 1966, 699, 24-32. (b) Roth,
W. R Chimia 1966, 20, 229-264. (c) Roth, W. R.; Ko¨nig, J.; Stein, K.
Chem. Ber. 1970, 103, 426-439.
(13) Recalculation from the specific rate constants given on the last page of ref
12a leads to the following Arrhenius parameters (standard error; temperature
range, 185-205 °C): for kH, Ea ) (36.13 ( 2.16) kcal mol-1, log A )
(11.35 ( 1.0) (Houk, Li, and Evanseck36d report identical values, and may
be presumed to have carried out the same recalculations) and for kD, Ea )
(37.7 ( 0.7) kcal mol-1, log A ) (11.36 ( 0.3). Unjustified recalculation
of kH without the 190.2° point improves matters, Ea ) (34.83 ( 0.78) kcal
mol-1, log A ) (11.75 ( 0.37), but lowers the energy of activation by 1.3
(16) Dewar, J. M. S.; McDougherty, R. C. The PMO Theory of Organic
Chemistry, Plenum Press: New York, 1975.
(17) Saettel, N. J.; Wiest, O. J. Org. Chem. 2000, 65, 2331-2336.
(18) Seetula, J. A.; Russell, J. J.; Gutman, D. J. Am. Chem. Soc. 1990, 112,
1347-1353.
(19) Carpenter, J. E.; Sosa, C. P. THEOCHEM 1994, 311, 325-330.
9
J. AM. CHEM. SOC. VOL. 126, NO. 43, 2004 14207

A R T I C L E S
Doering et al.
Figure 3. Experimental activation parameters for the simplest, paradigmatic
1,5-hydrogen shift in the more stable s-trans,penta-1,3(Z)-diene are shown
along with the hypothetical energy of activation of the rearrangement-
competent s-cis conformation, estimated from the MM2-calculated, steric
energy difference. Error limits are standard errors; thermochemical quantities
in kcal mol-1
.
already in the paradigm (Figure 3). The ideal correction required
to extract the activation parameters for rearrangement of the
s-cis conformation from the s-trans related, experimental values
(e. g., 3 to 4 in Figure 3) is the difference between their
enthalpies and entropies of formation. These are not experi-
mentally accessible. In their stead, we have used the difference
in the steric energies (∆SE) calculated for the two conformations
by the molecular mechanics program, MM2, of Allinger as made
available in Chem 3D Pro 5.20
Figure 4. Depicted are 1,5-hydrogen shifts in the more stable s-trans
conformations of two methyl-substituted 1,3(Z)-pentadienes designated by
the symbol, for example, (5 S 1), to indicate the reversible rearrangement
of the substituent between the 5- and 1-positions. Heats of formation, ∆fH°,
are derived by Benson’s method of group equivalent values (cf. caption of
Figure 3 for additional explanations).
With 1,3(Z)-pentadiene as example, the calculated value of
∆SE is +3.35 kcal mol-1 and leads to a corrected energy of
activation for the 1,5-hydrogen shift from the s-cis conformation
of 32.7 (36.1-3.4) kcal mol-1. Saettel and Wiest, and most
recently, Hayase, Hrovat, and Borden, have used the density
functional theory of B3LYP, to calculate the difference in
enthalpies of formation of the two conformations and have,
reassuringly, found values of 3.7 and 3.6 kcal mol-1, respec-
tively.17,21 The base value of 32.7 kcal mol-1 for the paradigm
is used in the following comparisons with the conformationally
corrected values of the perturbed systems (but please note again
the gnawing reservation implied in ref 13).
A second complication in making the comparisons arises from
interactions of the substituents with carbon-carbon double
bonds, the so-called energy of conjugation or conjugative
interaction. This is a perturbation on the educt, irrespective
whether taken as s-trans or s-cis. In the domain of alkyl groups,
values for conjugative interaction of methyl groups have been
proposed in Figure 1 of Doering et al.22 They range between
-2.66 and -0.87 kcal mol-1 depending on the total degree of
substitution. We hesitate to apply this type of correction to the
alkyl perturbations because the radical-stabilizing power of alkyl
groups is comparable in magnitude to the conjugative interaction
operating in the educt. What fraction of the conjugative
interactions may already be contained in the calculated steric
energy of the educt is also unknown.
Perturbation at the 5(1)23 position by unconjugated meth-
yl14,15,24 (Figure 4) or ethyl25 (Figure SI-1) lowers the confor-
mationally corrected energy of activation by -3.2 kcal mol-1
in both instances. We make an exception and suggest that much
of the 2.7-kcal mol-1 conjugative interaction of alkyl in the
trans-1,2-disubstituted olefins ultimately destined to become the
products, 1(E)-alkyl-3(Z)-pentadienes, may already be making
itself felt in the transition region. In this connection, examination
of 1,5-dimethyl-1(E),3(Z)-pentadiene might be illuminating, the
more so since Hess and Baldwin have predicted, also by B3LYP,
that it should have the same energy of activation as that of the
unsubstituted paradigm (uncorrected for the s-trans/s-cis con-
formational difference).5
Perturbations by methyl26 or ethyl27 at the 2(4)-position, or
by two methyl groups at the 2,3(4,3)-positions,28 do not
significantly lower the conformationally corrected energies of
activation (-0.2, +1.0, and -0.5 kcal mol-1, respectively;
Figures 5 and SI-1). Unfortunately lacking, but perhaps under-
(22) Doering, W. v. E.; Benkhoff, J.; Carleton, P. S.; Pagnotta, M. J. Am. Chem.
Soc. 1997, 119, 10947-10955.
(23) As a short-hand, idiosyncratic notation expressing the two positions between
which substituents are distributed by the reversible 1,5-hydrogen shift; 5(1),
for example, indicating reaction from the educt, 1(5) indicating the reverse
reaction of the product.
(24) Frey, H. M.; Pope, B. M. J. Chem. Soc. A 1966, 1701-1702.
(25) O’Neal, H. E.; Frey, H. M.; Pope, B. M. Int. J. Chem. Kinet. 1970, 2,
343-347.
(26) (a) Frey, H. M.; Ellis, R. J. J. Chem. Soc. A 1965, 4770-4773. (b) Glass,
D. S.; Boikess, R. S.; Winstein, S. Tetrahedron Lett. 1966, 999-1008.
(27) Frey, H. M.; Solly, R. K. J. Chem. Soc. A 1969, 733-735.
(28) Frey, H. M.; Lamont, A. M.; Walsh, R. J. Chem. Soc. A 1971, 2642-
2646.
(20) CambridgeSoft Corporation, 100 CambridgePark Drive, Cambridge, MA
02140: ChemBats3D PRO, version 5.0.
(21) Hayase, S.; Hrovat, D. A.; Borden, W. T. J. Am. Chem. Soc. 2004, 126,
10028-10034.
9
14208 J. AM. CHEM. SOC. VOL. 126, NO. 43, 2004

1,5-Hydrogen Shift in 1,3(Z)-Pentadiene
A R T I C L E S
Figure 6. Two 1,5-hydrogen shifts of phenyl-substituted 1,3(Z)-pentadienes
(cf. captions of Figures 3 and 4 for additional explanations).
Figure 5. Depicted are the 1,5-hydrogen shifts of two additional methyl-
substituted 1,3(Z)-pentadienes (cf. captions of Figures 3 and 4 for additional
explanations).
standably so in view of its symmetry, is a datum for 3-methyl-
1,3(Z)-pentadiene. Data are available, however, for 2,3(4,3)-
dimethyl-1,3(Z)-pentadiene (+0.5 kcal mol-1; Figure SI-1) and
3,5(3,1)-dimethyl-1,3(Z)-pentadiene (-2.2 kcal mol-1 when
compared to the unsubstituted paradigm, but +0.9 kcal mol-1
when compared, in our opinion more appropriately, to 5(1)-
methyl-1,3(Z)-pentadiene)(Figure4).14,28 Althoughmethylgroups
in the 2- or 3-position do not appear to stabilize the transition
region, a methyl group in the sp3 5-position on its way to
becoming an sp2 1-position in the product makes a significant
contribution of 3.1 kcal mol-1, commensurate with its develop-
ing energy of conjugation.
Perturbation by the Ad Libitum Phenyl Group
Perturbation by strongly radical-stabilizing phenyl groups
seems more likely to lead to definitive conclusions. Four
examples have been examined in Roth’s laboratory. These are
shown in Figures 6 and 7, along with their calculated, SE
conformational corrections. Perturbation by a single phenyl
group at the 2(4)-position lowers the corrected energy of
activation energy by 3.2 kcal mol-1, while a phenyl group at
the 2(4)-position coupled with a methyl group at the 5(1)-
position has a lowering effect of 5.2 kcal mol-1 vis-a`-vis the
corrected unsubstituted paradigm or 2.1 kcal mol-1 vis-a`-vis
5(1)-methyl-1,3(Z)-pentadiene.15 The combination of 4(2)-
phenyl and 5(1)-methyl has an energy-raising effect of +0.7
kcal mol-1, probably owing to adverse steric interactions.15
There being no simple 3(3)-phenyl analogue, the disubstituted
3(3)-phenyl-5(1)-methyl must serve as the only representative.
Its lowering effect vis-a`-vis the unsubstituted paradigm is 3.2
kcal mol-1, the same as that of the 2(4)-phenyl analogue.14 But
Figure 7. Two further 1,5-hydrogen shifts of phenyl-substituted pentadienes
(cf. caption of Figures 3 and 4 for additional explanations).
once again, compared, perhaps more appropriately, to 5(1)-
methyl-1,3(Z)-pentadiene (Figure 4), the value drops to -0.1
kcal mol-1. Investigation of the degenerate 3(3)-phenylpenta-
9
J. AM. CHEM. SOC. VOL. 126, NO. 43, 2004 14209

A R T I C L E S
Doering et al.
1,3(Z)-diene for direct comparison with 2(4)-phenylpenta-1,3-
(Z)-diene would give a more convincing insight into the relative
importance of a phenyl group at the 2- and 3-position.
Correction for conjugative interaction by phenyl would lower
the estimated heat of formation of the educt and, taken alone,
should manifest itself as a higher energy of activation than would
otherwise be observed if the perturbation had no effect on the
heat of formation of the educt. The magnitude of the conjugative
interaction of an ad libitum phenyl group located at the otherwise
unsubstituted R-position of a styrene has been estimated to be
5.1 kcal mol-1, or 2.5 kcal mol-1 if located at an already
substituted â-position.22 Although the reliability of the applica-
tion of these values to butadienes has not been demonstrated,
they must serve if any corrections for conjugative interaction
in the educt seem advisable.
Figure 8. Interconversion of 1-phenyl-5-p-tolyl-1,3(Z)-pentadiene, (E,Z)-
5, and 1-p-tolyl-5-phenyl-1,3(Z)-pentadiene, (E,Z)-6. Error limits are at the
95% confidence level.
In the examples of Figures 6 and 7, phenyl groups are free
to assume whatever conformation is energetically most favorable
in the incompetent s-trans educt, the competent s-cis educt, and
the fleeting transition region; that is, these phenyl groups are
what we call ad libitum. They are all of the â-type and therefore
subject to the correction of 2.5 kcal mol-1. Thus, the single
phenyl group in 2(4)-phenyl-1,3(Z)-pentadiene (Figure 6) may
be corrected not only by the 1.8-kcal mol-1 difference in steric
energy but by a conjugative interaction of 2.5 kcal mol-1. The
result is a total stabilization in the transition region of 5.7 kcal
Table 1. Specific Rate Constants for the Interconversion of
1-Phenyl-5-p-tolylpenta-1,3(Z)-diene [(E,Z)-5] and
1-p-Tolyl-5-phenylpenta-1,3(Z)-diene [(E,Z)-6]
a
a
educt
k1
K
educt
k1
K
153.7 °C
185.2 °C
(E,Z)-5
(E,Z)-6
5.35 ( 0.04b
5.29 ( 0.06b
1.257
(E,Z)-5
(E,Z)-6
63.0 ( 0.6b
61.2 ( 0.9b
1.233
Arrhenius Parameters
Ea ) 30.10 ( 0.21 (( 0.92)b kcal mol-1; log A ) 10.14 ( 0.11 (( 0.91)b
∆Hq ) 29.2 ( 0.2 kcal mol-1; ∆Sq ) -14.9 ( 0.5 eu.
mol-1 21,29
2(4)-Phenyl-5(1)-methylpenta-1,3(Z)-diene (Figure
.
6) leads to a stabilization of 7.7 kcal mol-1 (32.7-27.5 + 2.5
kcal mol-1) vis-a`-vis the corrected paradigm or 4.6 kcal mol-1
(29.6-27.5 + 2.5 kcal mol-1) vis-a`-vis the corrected value for
5(1)-methylpenta-1,3(Z)-diene (Figure 4).
a k1 × 10-6 sec-1
.
b Error limits at the 95% confidence level.
to precedent. The resulting mixture of (E,E)-5 and the desired
(E,Z)-5 resisted separation to a point where its necessity was
questioned. Indeed the (E,E) isomer proved stable to heating at
200 °C, a temperature well above that required for study of the
kinetics of the 1,5-hydrogen shift in the (E,Z) isomer. The
unseparated (E,E) isomer could thus serve as an internal
standard. To our pleased surprise, rearrangement of the system
could be followed quantitatively by NMR spectroscopy. There
seeming to be little serious question about the first-order nature
of the forward and backward steps of the rearrangement,
determination of the kinetics at two temperatures (∆T ) 31.5
°C) was deemed likely to furnish more reliable Arrhenius
parameters than allocating the two of the four runs to intermedi-
ate temperatures. The data relating relative concentrations and
time of reaction are given in Table SI-2 of the Supporting
Information. Specific rate constants were calculated on the basis
of the standard equation for reversible first-order reactions.
Values of the equilibrium concentration, xe, were varied to
achieve a “best fit”. The means of the pairs of the xe values at
each temperature were then used to calculate the four specific
rate constants reported in Table 1. Equilibrium constants at 153.7
°C and 185.3 °C are 1.250 and 1.233, respectively. This slight
favoring at equilibrium of the p-methylstyryl isomer 6 over the
styryl 5 is consistent with the observations of Bushby and Farber
on related systems.30 Arrhenius parameters for k1 (5f 6) are
derived from the four specific rate constants in Table 1: Ea )
30.1 ( 0.2 (( 0.9) kcal mol-1, log A ) 10.14 ( 0.11 (( 0.9)
[( standard errors (( 95% confidence levels)]; ∆Hq ) 29.2 (
0.2 kcal mol-1; ∆Sq ) -14.9 ( 0.5 eu.31
3(3)-Phenyl-5(1)-methylpenta-1,3(Z)-diene (Figure 7) leads
to a stabilization of 5.7 kcal mol-1 (32.7-29.5 + 2.5 kcal mol-1
)
vis-a`-vis the corrected paradigm or 2.6 kcal mol-1 (29.6-29.5
+ 2.5 kcal mol-1) vis-a`-vis the corrected value for 5(1)-
methylpenta-1,3(Z)-diene (Figure 4).
We conclude from these examples from the literature that
perturbation by phenyl in either the 2- or the 3-position
occasions a small lowering in the conformationally corrected
energies of activation. The firmness of the conclusion is
compromised by uncertainties in the accuracy of the corrections,
the magnitudes of which are comparable to expected π-elec-
tronic stabilizations.
Present Experimental Results
Perturbation by phenyl at the 1- or 5-position has not been
reported heretofore in the literature. This lacuna, the more acute
since, from the analysis outlined in Figure 2, an energy-lowering
interaction is predicted throughout the hypothetical continuum
of transition regions, is now filled by examination of the
reversible system, 1-phenyl-5-p-tolylpenta-1(E),3(Z)-diene [(E,Z)-
5] and 1-p-tolyl-5-phenylpenta-1(E),3(Z)-diene [(E,Z)-6] (Figure
8). If not as pristinely degenerate as 1,5-diphenyl-1(E),3(Z)-
pentadiene, (E,Z)-5 and (E,Z)-6 involve no more than a trivial
difference in their heats of formation.
Preparation of (E,Z)-5 from trans-cinnamaldehyde and tri-
phenyl-â-p-tolylethylphosphonium bromide proceeded according
(29) In a recent publication,21 Hayase, Hrovat and Borden calculate by B3LYP
values of ∆Hq ) 30.3 and 34.5 kcal mol-1, respectively, for the forward
and reverse rearrangements. These compare with 30.4 and 33.1 kcal mol-1
obtained experimentally by Schlief (Roth).15 Her measurements of equi-
librium constants over the temperature range, 238.1-277.7 °C lead to a
value for ∆∆H° ) -2.78 ( 0.14 kcal mol-1 (HHB: -4.2 kcal mol-1).
(30) Bushby, R. J.; Farber, A. M. J. Chem. Soc., Perkin Trans. II 1976, 1695-
1700.
9
14210 J. AM. CHEM. SOC. VOL. 126, NO. 43, 2004

1,5-Hydrogen Shift in 1,3(Z)-Pentadiene
A R T I C L E S
Table 2. Specific Rate Constants for the Thermal Rearrangement
of (Z)-2-Ethylidene-1-methylenecyclohexane, (Z)-7, to
1-Methyl-2-vinylcyclohexene (8)
a
a
temp,
°C
k1
temp,
°C
k1
117.9
117.9
121.6
121.6
131.7
131.7
5.86
5.60
7.60
7.60
27.5
27.0
165.1
165.1
175.7
175.7
185.7
185.7
185.8
185.8
40.5
40.4
116.
112.
193.
185.
209.
201.
Arrhenius Parameters
Ea ) 30.62 ( 0.18 (( 0.40)b kcal mol-1; log A ) 10.85 ( 0.09 (( 0.20)b
∆Hq ) 29.8 ( 0.2 kcal mol-1; ∆Sq ) -13.0 ( 0.5 eu.
a k1 × 10-6 sec-1
.
b Error limits at the 95% confidence level.
ethylidenecyclohexan-1-one, prepared according to Van-
Catledge from cyclohexanone and acetaldehyde.32 Irradiation
in degassed benzene generates a 0.9:1 mixture of (Z)- and (E)-
2-ethylidenecyclohexan-1-one. Without separation, this mixture
is converted by treatment with Wittig’s reagent to a 1:0.8
mixture of (E)-7 and (Z)-7, which can be separated by
preparative GC.
Rates of rearrangement of 7 to 1-methyl 2-vinylcyclohexene
(8) (Figure 9) in benzene-d6 are determined at six temperatures
over the range of temperatures, 117.8 to 185.7 °C. The product
has the same NMR spectrum in CDCl3 as that reported by
Kabalka et alii.33 Assignment of structure to compound 8 is
based on its NMR spectrum and method of preparation.
Quantitative analysis of mixtures of 7 and 8 is accomplished
by NMR employing resonances at 5.27 and 6.89 ppm, respec-
tively. The unrefined data relating time of reaction and relative
concentrations are collected in Table SI-3 of the Supporting
Information. Because the equilibrium constant relating 7 and 8
strongly favors 8, the data can be translated into specific rate
constants by a first-order model as well as by the model for
reversible first-order reactions. In one variant of the latter, the
value of the relative concentration at equilibrium xe is varied to
achieve a best fit for the specific rate constants; in another, the
mean value, 0.9800, of the “best fit” values for xe (there is no
discernible trend with temperature) is then used in the calculation
of the rate constants (k1 + k-1). From these rate constants
(recorded in Table 2 and Figure 9), Arrhenius parameters are
calculated:34 Ek1 ) 30.62 ( 0.18 [( 0.40 (95%)] kcal mol-1
and log A 10.85 ( 0.09 [( 0.20 (95%)] (translates at a mean
temperature of 425 K into ∆Hq ) 29.8 kcal mol-1 and ∆Sq )
-11.6 cal deg-1 mol-1).
Sufficient 8 is recovered from the combined products of
several kinetic experiments to provide material for a single
kinetic run. At 185.6-186.0 °C, a sample of 8 containing 0.9%
of 7 lead to 1.5% of 7 after 30 min and 2.5% after 60 min. A
rough specific rate constant, k-1 ) 2.6 × 10-6 s-1, compares
with that for k1 ) 1.8 × 10-4 s-1, and leads to a value of K )
∼70.
(31) In a recent publication,21 Hayase, Hrovat, and Borden calculate by B3LYP
values for the forward (and reverse) rearrangements of 1,5-diphenyl-1,3-
(Z)-pentadiene of ∆Hq ) 32.6 kcal mol-1. This value, offered without
assessment of probable uncertainties, is higher by +3.4 kcal mol-1 than
our experimental value. It seems unlikely that the discrepancy can be
attributed to the perturbation by p-methyl of one of the two phenyl groups.
(32) Van-Catledge, F. A.; Boerth, D. W.; Kao, J. J. Org. Chem. 1982, 47, 4096-
4106.
Figure 9. 1,5-Hydrogen shifts in three cyclic 1,3(Z)-pentadienes restricted
to s-cis conformations are shown. Error limits are at the 95% confidence
level; thermochemical quantities in kcal mol-1
.
In the second part of this work, the need for a conformational
correction to the experimental enthalpies of activation is
removed. Systems specifically designed to be frozen in the s-cis
conformation are realized by incorporating C2 and C3 of the
basic s-cis-1,3(Z)-pentadiene system into a cyclohexane ring and
fusing benzene rings to generate compounds 10 and 13 of Figure
9. The phenyl groups are simultaneously restricted to a geometry
more nearly coplanar with the five atoms of the rearranging
pentadiene and are no longer ad libitum.
The unsubstituted parent of this pair is 1(Z)-ethylidene-2-
methylenecyclohexane (7). Its synthesis takes off from (E)-2-
(33) Kabalka, G. W.; Mohammadi, M.; Hylarides, M.; Finn, R. D. Org. Prep.
Prod. Int. 1984, 16, 321-328.
9
J. AM. CHEM. SOC. VOL. 126, NO. 43, 2004 14211

A R T I C L E S
Doering et al.
Table 3. Specific Rate Constants for the Thermal Rearrangement
of (Z)-2-Ethylidene-1-methylene-1,2,3,4-tetrahydronaphthalene (10)
to 4-Methyl-3-vinyl-1,2-dihydronaphthalene (11)
Table 4. Specific Rate Constants for the Thermal Rearrangement
of (E)-1-Ethylidene-2-methylene-1,2,3,4-tetrahydronaphthalene
(13) to 3-Methyl-4-vinyl-1,2-dihydronaphthalene (14)
a,b
a,c
a,b
a,c
a,b
a,c
a,b
a,d
temp,
°
C
(k1
+
k- )
1
(k1
+
k- )
1
temp,
°
C
(k1
+
k- )
1
(k1
+
k- )
1
temp,
°
C
k1
k1
temp,
°
C
k1
k1
98.4
98.4
98.4
98.4
1.88
1.77
2.06
1.85
1.93
1.65
1.93
1.89
140.8
140.8
140.8
140.8
75.6
78.6
79.2
79.3
84.1
73.4
87.1
82.6
98.3
98.3
98.3
2.37
2.58
2.41
2.44
2.27
2.35
140.8
140.8
140.8
95.5
97.3
96.1
99.3
98.8
118.0
Arrhenius Parameters
Ead ) 26.40 ( 0.19 (( 0.52)f kcal mol-1; log Ab ) 9.92 ( 0.10 (( 0.29)f
Eae ) 27.00 ( 0.15 (( 0.42)f kcal mol-1; log Ac ) 10.26 ( 0.09 (( 0.29)f
∆Hq ) 26.2 ( 0.2 kcal mol-1; ∆Sq ) -15.5 ( 0.5 eu.
Arrhenius Parameters
Eab ) 26.85 ( 0.21 (( 0.59)e kcal mol-1; log Ab ) 10.07 ( 0.13 (( 0.33)e
Eac ) 27.15 ( 0.09 (( 0.23)e kcal mol-1; log Ac ) 10.21 ( 0.05 (( 0.13)e
∆Hq ) 26.4 ( 0.1 kcal mol-1; ∆Sq ) -15.7 ( 0.3 eu.
a (k1 + k-1) × 10-6 sec-1
.
b “Best fit” by allowing each xe to float in
a k1 × 10-6 sec-1
.
b “Best fit” by allowing xe to float in [(k1 + k-1) )
[(k1 + k-1) ) ln[(xe - xo)/(xe - xt)]/t. c Using the mean value of 0.988 for
xe from “best fit”. d Based on the “best fit” values for (k1 + k-1)b. e Based
on the “mean value” values for (k1 + k-1)c. f Error limits at the 95%
confidence level.
(1/t)/ln[(xe - xo)/(xe - xt)]. c Using the mean value of 0.9462 for xe. d Using
the mean value of 0.945 05 for xe. e Error limits at the 95% confidence
level.
are calculated: Ea ) 27.15 ( 0.09 (( 0.23 [95%]) and log A
) 10.21 ( 0.05 (( 0.13 [95%]).
Compound 10 is prepared from a precursor, (E)-2-ethylidene-
3,4-dihydro-2H-naphthalene-1-one, itself available from R-
tetralone and acetaldehyde following a procedure of Riahi et
al..35a Agreement with the reported NMR spectrum in CDCl3
is good. Irradiation of this ketone generates a photostationary
mixture of the desired (Z) isomer and the undesired (E) isomer.
Conversion of (Z)-2-ethylidene-3,4-dihydro-2H-naphthalene-1-
one to 10 is effected by the more reliable method of Petasis
and Bwozej,36 employing dimethyltitanocene, the conventional
Wittig procedure having proved unsatisfactory.
Rates of reaction of 10 are determined in three independent
runs each at 98.3 and at 140.8 °C (Table 3; Table SI-4 contains
the unrefined data). Calculations based on the equation for
reversible first-order reactions are carried out by varying values
of the concentration of product at equilibrium, xe, to achieve a
best fit of the rate constant. This method leads to Ea ) 26.40 (
0.19 (( 1.03 [95% confidence level]) and log A ) 9.92 ( 0.10
(( 0.57 [95%]). The mean of the “best fit” values, xe ) 0.9880
corresponding to K ) 82, is then used to recalculate specific
rate constants. This variant leads to Ea ) 27.00 ( 0.15 (( 0.85
[95%]) and log A ) 10.26 ( 0.10 (( 0.58 [95%]).
Preparation of 13 begins with â-tetralone and acetaldehyde
to provide a mixture of (E)- and (Z)-1-ethylidene-3,4-dihydro-
1H-naphthalene-2-one, which is separated by flash chromatog-
raphy.37 The (E) isomer is converted to (E)-13 by treatment
with dimethyltitanocene.
Rates of reaction of 13 are determined similarly in four
independent runs each at 98.4 and at 140.8 °C (Table 4; Table
SI-5 contains the unrefined data). In calculations based on the
equation for reversible first-order reactions, values of the
equilibrium concentration of product (xe) are varied to achieve
best fits of the eight rate constants and lead to Ea ) 26.85 (
0.21 (( 0.59 [95% confidence level]) and log A ) 10.07 (
0.13 (( 0.33 [95%]). The mean of the “best fit” values of xe
(0.9462 at 98.4 °C and 0.94055 at 140.8 °C, corresponding to
K ) 17.6 and 15.8, respectively) are then used to recalculate
specific rate constants. Therefrom Arrhenius parameters for k1
Discussion
The activation energy of the rearrangement of 1-phenyl-5-
p-tolylpenta-1(E),3(Z)-diene [(E,Z)-5] is 30.1 kcal mol-1 or 6.0
kcal mol-1 below the uncorrected value (36.1 kcal mol-1) for
the paradigm, penta-1,3(Z)-diene. Corrections for the difference
between s-trans and s-cis conformations, both ∆SE ∼3.5 kcal
mol-1, leave the estimated energy of activation of the re-
arrangement-competent s-cis conformation unchanged at ∼6
kcal mol-1. Inclusion of the conjugative interaction of the phenyl
group at the 1-position lowers the energy of activation by ∼5.1
kcal mol-1 for a total of ∼11 kcal mol-1. However, the phenyl
group at the 5-position very likely introduces an opposing, rate-
accelerating effect of equal or comparable magnitude as it
becomes involved in the transition region. It is not clear how a
correction for conjugative interaction is to be incorporated in a
credible way. Hayase, Hrovat, and Borden have calculated
enthalpies of activation for the fully degenerate 1(E),5-diphenyl-
1,3(Z)-pentadiene of 32.6 kcal mol-1 from its s-trans conforma-
tion and 28.8 kcal mol-1 from its s-cis conformation.21
Conjugative interaction of phenyl in the educts and steric factors
are implicit in these DFT calculations. Vis-a`-vis their calculated
value of ∆Hq for the paradigm (32.0 kcal mol-1), the lowering
is 3.2 kcal mol-1. The apparent discrepancy of 2.8 kcal mol-1
with the experimentally based value may seem large (6-3.2
kcal mol-1), but taking realistic account of the uncertainties in
the experimental value, and the inaccuracies in the DFT-
calculated value and the MM-2-calculated conformational
correction, we consider the agreement satisfactory!31
Recognition of the hybrid relation of the transition region of
the 1,5-hydrogen shift to Cope’s rearrangement and the dyo-
tropic shift of two hydrogen atoms38 provides models for the
paltry size of the stabilization energies accruing from perturba-
tion by phenyl (Charts 1 and 2). A model for the intramolecular
1,5-hydrogen shift more realistic than a simplistic dissociation
into a pentadienyl radical and an hydrogen atom can be
constructed from the extensively investigated, intermolecular
methane/methyl radical transfer. Starting from the known heat
(34) The specific rate constant at 131.7 °C is omitted from the calculation because
its inclusion increases the standard error by more than a factor of 3.
(35) (a) Riahi, A.; Thorey, C.; He´nin, F.; Muzart, J. Synth. Commun. 1998, 28,
4339-4344. (b) Negishi, E.; Cope´ret, C.; Ma, S.; Mita, T.; Sugihara, T.;
Tour, J. M. J. Am. Chem. Soc. 1996, 118, 5904-5918. (c) Enders, D.;
Zhu, J.; Kramps, L. Liebigs Ann./Recueil 1997, 1101-1113.
(36) Petasis, N. A.; Bzowej, E. I. J. Am. Chem. Soc. 1990, 112, 6392-6394.
(37) Kinder, F. R., Jr.; Chen, C.-P.; Bair, K. W. Synth. Commun. 1997, 27,
871-881.
(38) (a) Doering, W. v. E.; Rosenthal, J. W. J. Am. Chem. Soc. 1967, 89, 4534.
(b) Grimme, W.; Pohl, V.; Wortmann, J.; Frowein, D. Liebigs Ann. 1996,
1905-1916. (c) Corey, E. J.; Pasto, D. J.; Mock, W. L. J. Am. Chem. Soc.
1961, 83, 2957-2958. (d) Houk, K. N.; Li, Y.; Evanseck, J. D. Angew.
Chem., Int. Ed. Engl. 1992, 31, 682-708. (e) Houk, K. N.; Li, Y.;
McAllister, M. A.; O’Doherty, G.; Paquette, L. A.; Siebrand, W.;
Smedarchina, Z. K. J. Am. Chem. Soc. 1994, 116, 10895-10914.
9
14212 J. AM. CHEM. SOC. VOL. 126, NO. 43, 2004

1,5-Hydrogen Shift in 1,3(Z)-Pentadiene
A R T I C L E S
Chart 1
Chart 2
of formation of s-trans cis-piperylene (+19.5 kcal mol-1 39
) and
the credible difference of the s-cis conformation (+3.6 kcal
mol-1), proceeding to the interposition of an interaction-blocking
barrier (the red line in the lower part of Chart 1) by a 90° twist
about the C1-C2 terminal double bond (+64.9 kcal mol-1),40
introducing the energy of allylic stabilization (-15.7 kcal
mol-1),41 and adding the energy of activation of the methane/
methyl-radical hydrogen transfer (+14.6 kcal mol-1)7 lead to
an estimated energy of concert of 31.1 kcal mol-1. Theoretical
calculations of the intermolecular transfer have uniformly found
a 180°, linear transition structure to be the energy minimum.
By contrast, the optimum structures of the cyclic transition
regions in the 1,5-hydrogen shifts of pentadienes and the
dyotropic transfer are bent and likely higher in energy than the
linear optimum.42,43 The energy of concert may well be a lower
estimate.
transfer (25.9 kcal mol-1), and the most apposite, toluene/
benzyl-radical hydrogen transfer (15.8 kcal mol-1).44 The latter
is so close to the +14.6 kcal mol-1 of the methane/methyl-
radical hydrogen transfer that the small effect observed in 5 is
more easily understandable. Consistently, a symmetrical per-
turbation by two phenyl groups has little effect on the transition
region.
Compounds 10 and 13 are so constructed as to need no
thermochemical corrections for the s-trans/s-cis conformational
factor. Simultaneously the fused benzene groups are no longer
ad libitum but are more nearly coplanar with the pentadienyl
system. Their capacity for π-electronic interaction may thereby
be enhanced but at the expense of a flexibility concordant with
the minimization of adventitious steric interactions.
As calibration for perturbation by phenyl, note the claimed
energies of activation for the toluene/methyl-radical hydrogen
transfer (9.5 kcal mol-1), the methane/benzyl-radical hydrogen
(39) Pedley, J. B.; Naylor, R. D.; Kirby, S. P. Thermochemical Data of Organic
Compounds, 2nd ed.; Chapman and Hall: London, 1986.
(40) Doering, W. v. E.; Roth, W. R.; Bauer, F.; Breuckmann, R.; Ebbrecht, T.;
Herbold, M.; Schmidt, R.; Lennartz, H.-W.; Lenoir, D.; Boese, R. Chem.
Ber. 1989, 122, 1263-1275.
The parent 7 has an energy of activation of 30.6 kcal mol-1
,
which is lower by 2.1 kcal mol-1 than the 32.7 kcal mol-1 of
the s-trans/s-cis-corrected value being taken for the s-cis,
unsubstituted paradigm (Figure 3). No explanation for the
discrepancy is immediately apparent, beyond the already
(41) Korth, H. G.; Trill, R.; Sustmann, R. J. Am. Chem. Soc. 1981, 103, 4483-
4489.
(42) In this connection, regard the much lower energy of activation (Ea ) 21.5
kcal mol-1; Roth, W. R Tetrahedron Lett. 1964, 1009) reported by Baldwin
and Reddy for the 1,7-hydrogen shift in hepta-1,(Z)3(Z)5-triene. Although
the comparison of the antarafacial, helical eight-membered transition region
with the suprafacial, planar six-membered 1,5-hydrogen shift is problematic
in several ways, the crucial, core element has essentially become linear.
(43) Baldwin, J. E.; Reddy, V. P. J. Am. Chem. Soc. 1988, 110, 8223-8228.
(44) Like Ma and Schobert,7a we are not engaging in assessment of the accuracy
of these values.
9
J. AM. CHEM. SOC. VOL. 126, NO. 43, 2004 14213

A R T I C L E S
Doering et al.
Chart 3
mentioned high uncertainty in the experimental value taken for
the paradigm. The experimental activation energy for the
2-benzo analogue 10 is 27.0 kcal mol-1 or 3.6 kcal mol-1 below
7. A similar analysis of the 3-benzo analogue 13, Ea ) 27.2
kcal mol-1, reveals a comparable lowering of 3.4 kcal mol-1
.
The enthalpies of activation parameters from 10 and 13 may
also be corrected for conjugative interaction in the educt by
transferring values reported for similarly benzo-fused analogues.
Conjugative interaction of the more nearly coplanar benzo group
with a C-C double bond appears to be increased by ∼0.7 kcal
mol-1 45 This value, if added to the correction of 2.5 kcal mol-1
.
applied above to 2-phenyl-1,3(Z)-pentadiene, would increase the
estimated correction to 3.2 kcal mol-1
.
The observed activation energy for 10, corrected by this value,
becomes 23.8 kcal mol-1, implying a stabilization in the
transition region of 6.8 kcal mol-1. A similar analysis of the
3-benzo analogue 13 leads to an apparent stabilization of 6.6
kcal mol-1. By way of calibration, the experimental activation
energy for the thermal cis-trans isomerization of a benzo-fused
1,6-diphenylhexa-1,3,5-triene is 7.2 kcal mol-1 lower than a
comparable hexa-1,3,5-triene as model; that is, each phenyl has
lowered the energy of activation by 3.6 kcal mol-1
.
of additivity versus competition between phenyl groups at
opposite ends of the continuum.
Hayase, Hrovat, and Borden have very recently calculated
values of the enthalpies of activation by the DFT procedure,
BLYP/6-31G*, for the s-cis series, 7, 10, and 13.46
Conclusions
Their calculated values for ∆Hq are 33.5, 29.7, and 29.7,
respectively. Comparisons with the corresponding experimental
values of Ea, 30.6, 27.0, and 27.2 kcal mol-1, respectively, are
uniformly higher by ∼2.5 kcal mol-1, a discrepancy that exceeds
in our judgment plausible uncertainties in the experimental
values. The calculated differences between 7 and 10 and 7 and
13 are, however, both -3.8 kcal mol-1, in fine agreement with
The s-cis-locked compounds 10 and 13 of Figure 9 have
revealed energies of activation lowered essentially identically
by 3.6 and 3.4 kcal mol-1, respectively, regardless whether the
benzo group is positioned at C2 or C3. The small magnitude of
the stabilizations is in accord with the equally small size of the
difference between the stabilization energies of allyl and
cinnamyl radicals. Although other explanations can be enter-
tained, these stabilizations are consistent with the hypothesis
of a chameleon-like transition region that responds to a demand
for π-electron delocalization by substitution at the 2-position
by moving toward the allylic end of the continuum and to
substitution at the 3-position by moving toward the pentadienyl
end (Figure 2). Unlike in the Cope rearrangement where a large
change in geometry is involved (Figure 1), in the 1,5-hydrogen
shift change in geometry is negligible. Regardless of explanation,
it has been credibly established that radical-stabilizing phenyl
groups lower the energies of activation at all three types of
position in the transition region of the 1,5-hydrogen shifts in
acyclic 1,3(Z)-pentadienes.
the experimental differences of -3.6 and -3.2 kcal mol-1
respectively. Hayase, Hrovat, and Borden ascribe these differ-
ences to relief of a greater strain in the educts on passing to the
transition structures.
,
Several suggestions, some being pursued at present, that may
lead to further insight into the nature of the transition region
(Chart 3) are offered. The introduction of sufficient rigidity into
the 1,5-diphenyl system by way of compound 16 achieves a
minimal goal of removing the s-trans/s-cis correction. Com-
parison of the activation parameters for compounds 17 and 18
may help unravel the contribution to the transition region of
conjugative interaction in the educt. In respect of 2-phenyl
derivatives, advantage can be taken of the possibility of
introducing two phenyl groups at the same type of position.
The resulting degenerate 2,4-diphenyl derivative, compound 19,
is in every way comparable to the 1,5-diphenyl derivative
already examined in this work, while the more rigid compound
20 removes the s-trans/s-cis conformational factor and confines
both phenyl groups to near coplanarity without appearing to
place significant steric restrictions on the hydrogen atoms of
the 1,5-carbon atoms at the core of the hydrogen shift. As the
basis of a further assessment of the effect of a phenyl group in
the 3-position, compound 21 in conjunction with compound 22
may be helpful. Compound 23 bears on the interesting question
Experimental Section
1
General Procedures. H (500 MHz) and 13C (125.7 MHz) NMR
spectra were measured in benzene-d6, unless otherwise noted, on a
Varian Unity/Inova 55 Instrument. Chemical shifts are reported in ppm
1
(δ) with respect to residual C6D5H (7.15 ppm for H and 128.0 ppm
for 13C). Spin-lattice relaxation times (T1) were determined by the
inversion-recovery method in benzene-d6. Preparative GC on a Varian
Aerograph A90-P3 instrument employed a 3-m column of 20%
Carbowax 20M on Anachrom AS with He as a carrier gas. Analytical
gas chromatography was conducted on a Hewlett-Packard (HP) 5890A
gas chromatograph equipped with a J&W Scientific, Inc. DB-225 (0.53
mm i.d. × 30-m, 1 µm film thickness) megabore column and an HP
3393A integrator.
(45) Doering, W. v. E.; Benkhoff, J.; Shao, L.-m. J. Am. Chem. Soc. 1999, 121,
962-968.
Triphenyl-â-p-tolylethylphosphonium Bromide. To a suspension
of methyltriphenylphosphonium bromide (6.12 g, 17 mmol) in 60 mL
(46) Hayase, S.; Hrovat, D. A.; Borden, W. T. Unpublished results (with
permission to quote).
9
14214 J. AM. CHEM. SOC. VOL. 126, NO. 43, 2004

1,5-Hydrogen Shift in 1,3(Z)-Pentadiene
A R T I C L E S
of anhydrous tetrahydrofuran (THF) in a 100-mL, round-bottomed flask
under argon, 8.5 mL of a solution of n-butyllithium (2.0 M in pentane)
were added at 0 °C. After the mixture was warmed to room temperature,
stirred for 30 min, and recooled to 0 °C, a solution of p-tolylmethyl
bromide (1.57 g, 8.5 mmol) in anhydrous THF (5 mL) was added by
syringe over a 10-min period. After the mixture was stirred for 2 h at
0 °C and overnight at 25 °C, absolute ethanol was added dropwise
until no further solid precipitated. The filtered precipitate was washed
3 times with THF, dispersed in 6 mL of water, filtered, washed with
1 mL of water, and dried over phosphorus pentoxide to give 2.1 g
(53%) of a colorless solid: 1H NMR (CDCl3) 2.23 (s, 3H), 2.95-3.02
(m, 2H), 3.96-4.02 (m. 2H), 7.00 (d, J ) 8.05 Hz, 2H), 7.12 (d, J )
8.05 Hz, 2H), 7.65-7.85 (m. 15H).
mercury lamp for 10 h. GC analysis (50 °C for 15 min, 15 °C per min
to 200 °C, and finally 15 min isothermal) showed two major peaks:
starting material at 23.08 min and a new peak at 21.36 min in a ratio
of 1 to 0.9.
(Z)- and (E)-2-Ethylidene-1-methylenecyclohexane (Z)-7 and (E)-
7. In a 500-mL, four-necked, round-bottomed flask fitted with a 150-
mL, pressure-equalizing addition funnel, methyltriphenylphosphonium
bromide (17.6 g, 0.05 mol) was suspended in tetrahydrofuran (THF)
(150 mL, freshly distilled from sodium and benzophenone) under an
argon atmosphere. n-Butyllithium (24.5 mL, 2.0 M solution in pentane)
was added at -78 °C (became red/orange). The resulting Wittig reagent
was warmed to room temperature, stirred for 30 min, then cooled to
-78 °C. A solution of the 1:0.9 mixture of enones (6.1 g, 0.05 mol) in
THF (60 mL) was added with stirring over 1 h. After an additional
hour of being stirred, the reaction mixture was diluted with water (200
mL) and separated into an organic layer and an aqueous layer, which
was extracted with pentane (2 × 40 mL). The combined organic
solutions were washed with water (4 × 200 mL), dried over Na2SO4,
vacuum filtered, and concentrated by distillation. Distillation at reduced
pressure gave 2.66 g of a clear oil consisting of 30% of (Z)-7 and 39%
of (E)-7 by analytical GC. Preparative GC at 72 °C provided pure
samples. (Z)-7: 1H NMR 1.46-1.54 (m, 4H), 1.72 (dt, J ) 0.98 Hz,
6.84 Hz, 3H), 2.12-2.17 (m, 4H), 4.77 (d, J ) 2.93 Hz, 1H), 4.99 (t,
J ) 0.98 Hz, 1H), 5.27 (q, J ) 6.84 Hz, 1H); 13C NMR 14.48, 28.04,
28.39, 37.13, 38.28, 110.68, 118.15, 141.03, 146.81. (E)-7: 1.41 (m,
2H), 1.47 (m, 2H). 1.52 (d, J ) 6.84 Hz, 3H), 2.18 (m, 4H), 4.66 (d,
J ) 0.98 Hz, 1H), 4.98 (d, J ) 1.95 Hz, 1H), 5.64 (q, J ) 6.84 Hz,
1H); 13C NMR 13.05, 26.45, 27.50, 28.38, 35.82, 107.39, 117.49,
141.08, 151.51 (the missing 13C(C-sp2) resonance is presumed to be
hidden under the solvent).
(E,Z)- and (E,E)-1-Phenyl-5-p-tolylpenta-1,3-diene [(E,Z)-5 and
(E,E)-5]. Triphenyl-â-p-tolylethylphosphonium bromide (1.82 g, 3.95
mmol) was converted to its Wittig reagent as above. A solution of trans-
cinnamaldehyde (0.52 g, 3.95 mmol) in anhydrous THF (5 mL) was
added at 0 °C by syringe over a period of 20 min. After concentration
in a vacuum, the mixture was treated with 20 mL of water and extracted
twice with pentane (30 mL each). The combined extracts were washed
with water (40 mL) and brine, dried over Na2SO4, filtered, concentrated,
and subjected to column chromatography (neutral alumina, hexane) to
give a mixture of (E,Z)-5 (67%) and (E,E)-5 (33%) as a colorless oil
1
(66%). H NMR (E,Z)-5: 2.12 (s, 3H), 3.47 (d, J ) 7.81 Hz, 2H),
5.60-5.65 (m, 1H), 6.19-6.24 (m, 1H), 6.49 (d, J ) 15.63 Hz, 1H),
1
6.97-7.27 (m, 10H). H NMR (E,E)-5: 2.14 (s, 3H), 3.29 (d, J1 )
6.84 Hz, 2H), 5.77-5.83 (m, 1H), 6.11-6.16 (m, 1H), 6.34 (d, J )
15.63 Hz, 1H), 6.70 (dd, J1 ) 15.63 Hz, J2 ) 10.74 Hz, 1H), 6.91-
7.27 (m, 9H).
(E,Z)- and (E,E)-5-Phenyl-1-p-tolylpenta-1,3-diene (E,Z)-6 and
(E,E)-6. The conversion of triphenyl-â-phenylethylphosphonium bro-
mide (1.00 g, 2.2 mmol) and 4′-methylcinnamaldehyde (0.32 g) to a
mixture (0.35 g) of (E,Z)-6 (70%) and (E,E)-6 (30%) was effected by
the procedure above: 1H NMR (E,Z)-6: 2.09 (s, 3H), 3.46 (d, J )
7.81 Hz, 2H), 5.55-5.60 (m, 1H), 6.22-6.26 (m, 1H), 6.53 (d, J )
15.63 Hz, 1H), 6.95-6.97 (m, 2H), 7.06-7.25 (m, 8H); (E,E)-6 2.09
(s, 3H), 3.27 (d, J1 ) 6.84 Hz, 2H), 5.72-5.78 (m, 1H), 6.11-6.16
(m, 1H), 6.38 (d, J ) 15.63 Hz, 1H), 6.67-6.72 (m, 1H), 6.95-6.97
(m, 2H), 7.04-7.22 (m. 7H).
Kinetics. The sample (2-3 mg) in 0.2 mL of (benzene-d6) in an
NMR tube was degassed through three freeze-thaw cycles and sealed
under vacuum. The sealed tube was then heated in the vapors of boiling
solvents for the specified periods of time (data given in the Supporting
Information, Table SI-2): anisole, bp 153.6-153.8; diethyl oxalate,
bp 185.1-185.4.
(E)-2-Ethylidenecyclohexan-1-one. In a 250-mL, two-necked,
round-bottomed flask fitted with a 30-mL, pressure-equalizing addition
funnel, sodium ethoxide (430 mg, 6.4 mmol) was added to cyclohex-
anone (13.2 mL, 128 mmol), cooled to 0 °C, and stirred for 15 min, at
which time a solution of acetaldehyde (14.3 mL, 255 mmol) and
cyclohexanone (13.2 mL, 128 mmol) was added from the addition
funnel dropwise over 2 h. After being stirred at 0 °C for 1 h, the reaction
mixture was quenched with water (3 mL), transferred to a separatory
funnel, and treated with oxalic acid (0.21 g, 1.6 mmol) dissolved in
water (3 mL). After being washed with brine (2 × 20 mL), the organic
solution was distilled under reduced pressure, first at 95 °C and 95
mmHg to remove residual water and cyclohexanone, then at 110 °C
and 5 mmHg to afford 3.3 g of (E)-2-ethylidenecyclohexan-1-one in
21.9% of theoretical yield: 1H NMR 1.24 (m, 2H); 1.32 (m, 5H); 2.03
(m, 2H); 2.19 (t, J ) 6.59 Hz, 2H); 6.86 (m, 1H); 13C NMR 13.07,
23.38, 23.60, 26.35, 40.16, 133.16, 137.17, 198.03.
Heating (Z)-2-Ethylidene-1-methylenecyclohexane, (Z)-7. Samples
of (Z)-7 (7 mg, 7.5 mmol) in 0.7 mL of benzene-d6 in a lead-potash
glass tube were degassed through three freeze-thaw cycles and sealed
under vacuum. The tubes were heated in the vapors of the following
boiling solvents: n-butyl alcohol (117.4-118.3 °C), tetrachloroethylene
(121.3-121.8 °C), chlorobenzene (131.3-132.0 °C), mesitylene (164.9-
165.2 °C), p-methylanisole (175.5-175.9 °C), and diethyloxalate
(185.5-186.0 °C). For analysis, the tubes were opened. The contents
were transferred to NMR tubes and analyzed quantitatively by the
absorption at 5.27 and 4.77 for (Z)-7 and those at 6.89 and 5.15 ppm
for 8. A sample of the product, 1-methyl-2-vinylcyclohexene (8) was
isolated from the combined reaction products of the various kinetic
runs by preparative GC using a 3-m, 20% Carbowax 20M on Anakrom
column at 72 °C: 1H NMR 1.41-1.45 (m, 2H), 1.48-1.51 (m, 2H),
1.61 (s, 3H), 1.83-1.87 (m, 2H), 2.09-2.12 (m, 2H), 4.99 (d, J )
10.99 Hz, 1H), 5.15 (d, J ) 17.34 Hz, 1H), 6.89 (dd, J ) 17.34, 10.99
Hz, 1H); 13C NMR 19.15, 23.05, 23.16, 25.07, 33.00, 110.10, 133.10,
135.53.
(Z)-2-Ethylidene-3,4-dihydro-2H-naphthalene-1-one. Following
the general procedure of Riahi et al.,35a R-tetralone (6.0 g, 0.04 mol)
and acetaldehyde (1.8 g, 0.04 mol) are condensed to yield 5.7 g of
crude product, which, after flash-chromatography (22 × 3.5 cm2 column
of silica gel, 14:1 petroleum ether/ethyl acetate), affords 2.8 g of (E)-
2-ethylidene-3,4-dihydro-2H-naphthalene-1-one (2.4 g after evaporative
distillation, mp 38 °C (lit. 38 °C35b)): 1H NMR35b,c (benzene-d6) 1.38
(d of t, J ) 6.84, 1.47 Hz, 3H), 2.29 (br t, J ) 6.35 Hz, 2H), 2.42 (t,
J ) 6.35 Hz, 2H), 6.82 (d, J ) 7.33 Hz, 1H), 7.01 (t of d, J ) 7.81,
0.98 Hz, 1H), 7.06 (t of d, J ) 7.32, 1.95 Hz, 1H), 8.42 (dd, J ) 7.81,
1.47 Hz, 1H) (one H is hidden under the H of the solvent benzene-d6);
1H NMR (CDCl3) 1.88 (d, J ) 7.32, Hz, 3H), 2.79 (br t, J ) 6.35 Hz,
2H), 2.95 (t, J ) 6.84 Hz, 2H), 7.03 (q of d, J ) 5.37, 1.47 Hz, 1H),
7.25 (t, J ) 6.35 Hz, 1H), 7.33 (t, J ) 7.33 Hz, 1H), 7.46 (t of d, J )
7.33, 1.47 Hz, 1H), 8.09 (dd, J ) 6.84, 0.98 Hz, 1H); 13C NMR
(benzene-d6) 14.10, 25.81, 29.54, 127.72, 128.89, 129.13, 133.30,
134.56, 134.93, 136.63, 144.21, 186.69.
Irradiation of (E)-2-Ethylidenecyclohexan-1-one. In a 500-mL,
single-necked, round-bottomed flask, a sample of ketone (7.24 g, 58
mmol) in benzene (450 mL) was degassed by passing through it a
stream of nitrogen for 20 min. The flask was then fitted with a reflux
condenser and irradiated with a Hanovia 450-W, medium-pressure
9
J. AM. CHEM. SOC. VOL. 126, NO. 43, 2004 14215

A R T I C L E S
Doering et al.
An X-ray crystallographic determination confirms the assigned
structure and affords a torsional angle of -7.8 ° between the carbonyl
group and the double bond.47
dihydro-1H-naphthalene-2-one and 4.3 g of (Z)-1-ethylidene-3,4-
dihydro-1H-naphthalene-2-one were isolated by flash chromatography
(15 × 6.5 cm2, silica gel, 14:1 petroleum ether/ethyl acetate), the (E)
isomer being the first to elute. 1H NMR (Z)-isomer: 2.09 (d, J ) 7.69
Hz, 3H), 2.57 (br t, J ) 6.59 Hz, 2H), 2.94 (t, J ) 3.30 Hz, 2H), 7.05
(q, J ) 7.69 Hz, 1H), 7.25 (m, 3H), 7.38 (d, J ) 6.59 Hz, 1H). (E)-
isomer: 2.29 (d, J ) 7.32 Hz, 3H), 2.62 (br t, J ) 6.59 Hz, 2H), 2.97
(t, J ) 6.59 Hz, 2H), 7.56 (q, J ) 7.32 Hz, 1H), 7.22 (m, 3H), 7.38 (d,
J ) 9.15 Hz, 1H).
Following the same procedure as above, bis(cyclopentdienyl)-
dimethyltitanium (6.2 g, 30 mmol) and (E)-1-ethylidene-3,4-dihydro-
1H-naphthalene-2-one (1.7 g, 10 mmol) were allowed to react to afford
1.64 g of dark-brown oil, which was flash-chromatographed over neutral
alumina to give, after evaporative distillation, 300 mg of 13: 1H NMR
1.85 (d, J ) 7.33 Hz, 3H), 2.33 (t, J ) 6.84 Hz, 2H), 2.66 (t, J ) 6.84
Hz, 2H), 4.94 (d, J ) 1.95 Hz, 1H), 5.03 (dd, J ) 2.20, 0.98 Hz, 1H),
6.08 (q, J ) 7.32 Hz, 1H), 6.91 (d, J ) 8.30 Hz, 1H), 7.04 (m, 2H),
7.46 (d, J ) 7.81 Hz, 1H).
Following the procedure of Riahi et al.,35a a solution of the ketone
(7.25 g, 0.04 mol) in 350 mL of benzene in a 500-mL, three-necked
Pyrex flask was degassed by a stream of nitrogen for 20 min and then
irradiated for 19 h with a Hanovia 450-watt medium-pressure mercury
lamp with stirring. After concentration, the resulting oil (7.22 g, E/(Z)
ratio 1:1 by NMR) was flash-chromatographed as above to yield 1.2 g
of (Z)-2-ethylidene-3,4-dihydro-2H-naphthalene-1-one (>99% of pu-
rity): 1H NMR 2.16 (d of t, J ) 7.32, 1.47 Hz, 3H), 2.26 (t of t, J )
6.35, 1.46 Hz, 2H), 2.48 (t, J ) 6.35 Hz, 2H), 5.66 (q of t, J ) 7.32,
1.47 Hz, 1H), 6.80 (d, J ) 7.33 Hz, 1H), 7.00 (t, J ) 7.06 Hz, 1H),
7.05 (t of d, J ) 7.33, 1.47 Hz, 1H), 8.40 (d, J ) 7.81 Hz, 1H).
(Z)-2-Ethylidene-1-methylene-1,2,3,4-tetrahydronaphthalene (10).
Following a general procedure of Petasis and Bzowej,36 a solution of
the ketone above (1.4 g, 9 mmol) in toluene (5 mL) added to a solution
of bis(cyclopentdienyl)dimethyltitanium (5.6 g, 27 mmol) in toluene
(62 mL) was stirred for 24 h at 65 °C, diluted with petroleum ether
(70 mL) at room temperature, and filtered from separated titanium salts.
The filtrate was concentrated to an oil, which was flash-chromato-
graphed (20 × 6.5 cm, neutral alumina, 14:1 petroleum ether/ethyl
acetate) to give (Z)-2-ethylidene-1-methylene-1,2,3,4-tetrahydronaph-
thalene (195 mg after evaporative distillation): 1H NMR 1.77 (d, J )
7.08 Hz, 3H), 2.34 (t, J ) 6.59 Hz, 2H), 2.68 (t, J ) 6.59 Hz, 2H),
5.08 (d, J ) 0.73 Hz, 1H), 5.39 (t (q?), J ) 7.00 Hz, 1H), 5.77 (s, 1H),
6.89 (m, 1H), 7.03 (m, 2H), 7.61 (m, 1H).
Kinetic studies were effected in the manner described above for 7,
using resonances at 5.13 ppm for (E)-13 and 6.50 for 14. From samples
that had reached equilibrium, the spectrum of the product, 3-methyl-
4-vinyl-1,2-dihydronaphthalene (14), was obtained: 1H NMR 1.78 (s,
3H), 1.97 (t, J ) 7.57 Hz, 2H), 2.51 (t, J ) 7.57 Hz, 2H), 5.27 (dd, J
) 17.82, 2.20 Hz, 1H), 5.32 (dd, J ) 11.35, 2.20 Hz, 1H), 6.74 (m,
1H), 7.07 (m, 1H), 7.29 (m, 2H), 7.43 (d, J ) 7.81 Hz, 1H).
Acknowledgment. This work has been supported generously
by the Norman Fund in Organic Chemistry in memory of Ruth
Alice Norman Weil Halsband. We are grateful to Professor
Weston Thatcher Borden for his many helpful comments and
his permission to cite prior to publication his apposite compu-
tational investigations. We express our appreciation to Professor
Dieter Hasselmann for his painstaking, critical examination of
the manuscript.
Kinetic studies were effected in the manner described above for 7,
using resonances at 5.77 ppm for (Z)-10 and 5.19 for 11. From samples
that had reached equilibrium, the spectrum of the product, 4-methyl-
3-vinyl-1,2-dihydronaphthalene (11), was obtained: 1H NMR 1.96 (d,
J ) 2.93 Hz, 3H), 2.27 (t of d, J ) 7.93, 1.47 Hz, 2H), 2.55 (t, J )
8.30 Hz, 2H), 5.10 (d, J ) 10.99 Hz, 1H), 5.19 (dd, J ) 17.09, 0.98
Hz, 1H), 6.96 (m, 2H), 7.05 (t of d, J ) 7.32, 1.22 Hz, 1H), 7.11 (t of
d, J ) 7.45, 1.47 Hz, 1H), 7.20 (d, J ) 7.57 Hz, 1H).
Supporting Information Available: Untreated data from the
kinetic studies summarized in Tables SI-1 through SI-5 and one
figure of three additional rearrangements of alkyl-substituted
1,3(Z)-pentadienes (eight pages). This material is available free
of charge via the Internet at http://pubs.acs.org.
(E)-1-Ethylidene-2-methylene-1,2,3,4-tetrahydronaphthalene (13).
Following a procedure described in detail using benzaldehyde,37
â-tetralone (10 g) and acetaldehyde (3.8 mL) were condensed to afford
12.5 g of crude product, from which 3.1 g of (E)-1-ethylidene-3,4-
(47) Donahue, M.; Staples, R. J. X-ray Crystallographic Laboratory, Harvard
University, unpublished.
JA040106K
9
14216 J. AM. CHEM. SOC. VOL. 126, NO. 43, 2004