Conjugative interaction in styrenes

Article abstract of DOI:10.1021/ja971040e

Conjugative interactions of the carbon-carbon double bond are fundamental in organic chemistry. In this work, equilibria are established among conjugated and unconjugated isomers of two β-substituted styrenes, 1-phenylbut-1-ene and 1-phenyl-3-methylbut-1-ene, and one α,β-disubstituted styrene, 2-phenyl-5-methylhex-2-ene, over a range of temperatures (the van't Hoff method) in hexamethylphosphoric triamide and potassium tert-butoxide. From the trans styrenes of the first two sets, an enthalpy of conjugative interaction of phenyl vis-a-vis alkyl (ΔΔH(ConjInter/Alk)) = -2.5 ± 0.2 kcal/mol [-5.1 kcal/mol defined as phenyl vis-a-vis hydrogen (ΔΔH(ConjInter/H))] is observed, while the cis styrenes reveal an attenuated ΔΔH(ConjInter/Alk) of -1.1 kcal/mol (ΔΔH(ConjInter/H)) = -2.7 kcal/ mol). The α-methyl group in the third set also leads to a reduced conjugative interaction. Entropy plays an important role in determining positions of equilibrium. Free energies of conjugation are reported for several sterically hindered o-methyl-substituted styrenes.

Full text of DOI:10.1021/ja971040e

J. Am. Chem. Soc. 1997, 119, 10947-10955
10947
Conjugative Interaction in Styrenes
W. von E. Doering,*,1a Johannes Benkhoff,1a Peter Smart Carleton,1b,2 and
Marco Pagnotta1a,3
Contribution from the Department of Chemistry and Chemical Biology, HarVard UniVersity,
Cambridge, Massachusetts 02138-2902, and the Sterling Chemistry Laboratory, Yale UniVersity,
New HaVen, Connecticut 06520
ReceiVed April 1, 1997X
Abstract: Conjugative interactions of the carbon-carbon double bond are fundamental in organic chemistry. In
this work, equilibria are established among conjugated and unconjugated isomers of two â-substituted styrenes,
1-phenylbut-1-ene and 1-phenyl-3-methylbut-1-ene, and one R,â-disubstituted styrene, 2-phenyl-5-methylhex-2-ene,
over a range of temperatures (the van’t Hoff method) in hexamethylphosphoric triamide and potassium tert-butoxide.
From the trans styrenes of the first two sets, an enthalpy of conjugative interaction of phenyl Vis-a`-Vis alkyl
(∆∆HConjInter/Alk) ) -2.5 ( 0.2 kcal/mol [-5.1 kcal/mol defined as phenyl Vis-a`-Vis hydrogen (∆∆HConjInter/H)] is
observed, while the cis styrenes reveal an attenuated ∆∆HConjInter/Alk of -1.1 kcal/mol (∆∆HConjInter/H ) -2.7 kcal/
mol). The R-methyl group in the third set also leads to a reduced conjugative interaction. Entropy plays an important
role in determining positions of equilibrium. Free energies of conjugation are reported for several sterically hindered
o-methyl-substituted styrenes.
Background to Conjugative Interaction
The great majority of conjugative interactions has been
obtained by the measurement of equilibrium constants in the
Thermochemical consequences of the interaction of a wide
variety of substituents with the carbon-carbon double bond have
a long history in organic chemistry. The subject has been
exhaustively reviewed by Hine and his co-workers.4 Thermo-
chemical information about styrenes being sparse, this paper is
focused on the enthalpy and entropy of conjugation of the phenyl
group and the influence of substituents and configuration on
deviations from coplanarity.
Analysis of thermochemical data in the literature5,6 relevant
to styrene affords various estimates of the enthalpy of conjuga-
tive interaction. From heats of formation (gas) of styrene (a0)
and ethylbenzene (a2) compared to vinylcyclohexane (m0) and
ethylcyclohexane (m2), an enthalpy of conjugation of phenyl
with ethene relative to that of alkyl with the carbon-carbon
double bond may be obtained.7 A value may also be derived
from the heat of hydrogenation of trans-stilbene (b0) to bibenzyl
(b2) by comparison with the mean of those of trans-1,2-
dialkylethenes (Vide infra).8,9 Bearing on conjugative interaction
in R-substituted styrenes are heats of hydrogenation of 2,5-
diphenylhexa-1,5-diene and 2,6-diphenylhepta-1,6-diene.10,11
These analyses bring to light substantial ambiguity in the
selection of a trustworthy value for the enthalpy of conjugation
of phenyl Vis-a`-Vis alkyl: somewhere between -0.6 and -3.3
kcal/mol seems credible.7,8,11-13
(7) For a0(PNK,5 p 267), values of ∆fH°(liq), +24.73 and +24.91 kcal/
mol (mean, 24.82 kcal/mol), and of ∆Hvap, +10.50 and +9.47 kcal/mol,
are given, from which ∆fH°(a0, gas) range from +35.32 to +34.29 kcal/
mol. For a2(p 268), ∆fH°(liq) are -2.95 and -3.55 kcal/mol and ∆Hvap is
+10.10 kcal/mol, from which ∆fH°(a2, gas) are +7.15 and +6.55 kcal/
mol, respectively. From these combustion data, values for ∆rHH (a0fa2,
2
gas) range from -27.14 to -28.77 kcal/mol. For m0(p 270), ∆fH°(liq),
-21.19 kcal/mol, and ∆Hvap, +9.50 kcal/mol, give ∆fH°(m0, gas), -11.69
kcal/mol. For m2(p 273), ∆fH°(liq), -50.70 kcal/mol, and ∆Hvap, +9.57
and +9.67 kcal/mol (mean, +9.62 kcal/mol), lead to ∆fH°(m2, gas), -41.08
kcal/mol. From these data, a value for ∆rHH (m0fm2, gas) of -29.39 kcal/
2
mol is derived. Values for the enthalpy of conjugation of styrene between
-0.62 and -2.25 kcal/mol are acceptable. If the lower value for ∆Hvap(a0)
(+9.47 kcal/mol) is ignored and mean values for ∆fH°(a0, gas) and ∆fH°-
(a2, gas) of +35.32 and +6.85 kcal/mol, respectively, is taken, ∆rHH (a0fa2,
2
gas) of -28.47 kcal/mol is derived and a value for ∆Hconj(a0) of -0.92
kcal/mol results. Experimental values for heats of hydrogenation (gas) of
a0 and a2 to m2 (-76.50 and -48.18 kcal/mol, respectively) lead to ∆rHH
-
2
(a0fa2, gas) of -28.32 kcal/mol and ∆Hconj(a0) of -1.07 kcal/mol. An
experimental determination of ∆rHH (m0fm2, liq) of -27.90 kcal/mol leads
2
to a positive value for ∆Hconj(a0) of +0.42 kcal/mol.8
(8) Williams, R. B. J. Am. Chem. Soc. 1942, 64, 1395-1404.
(9) From two values for ∆fH°(b0, solid), +32.72 and +33.58 kcal/mol
(PNK,5 p 304), and one for ∆Hsublim, +23.71 kcal/mol, two values for ∆fH°-
(b0, gas), +56.43 and +57.29 kcal/mol, are derived; from two values for
bibenzyl (b2) (p 305), ∆fH°(b2, solid), +11.87 and +12.32 kcal/mol, and
∆Hsublim, +21.84 kcal/mol (a much lower value, +20.10 kcal/mol, is not
taken), ∆fH°(b2, gas), +33.71 and +34.16 kcal/mol, are derived; resulting
∆rHH (b0fb2, gas) range from -22.27 to -23.58 kcal/mol. If a mean value
2
for trans-1,2-dialkylethenes of -27.25 ( 0.05 kcal/mol (Vide infra) is taken
as the model, ∆Hconj(b0) of -3.67 to -4.98 kcal/mol or ∆Hconj(a0) of -1.84
X Abstract published in AdVance ACS Abstracts, October 15, 1997.
(1) (a) Harvard University. (b) Sterling Chemistry Laboratory, Yale
University.
to -2.49 kcal/mol results. Hydrogenation data,8 ∆rHH (b0fb2, gas) ) -20.6
2
kcal/mol, lead to ∆Hconj(b0) of -6.65 kcal/mol or ∆Hconj(a0) of -3.33 kcal/
mol.
(2) Carleton, Peter Smart, Conjugative Interaction of Benzene Rings and
Double Bonds. Ph.D. Dissertation, Yale University, 1967; Diss. Abstr.
Intern. B 1967, 27, 4265 (Order No. 67-07001).
(3) Pagnotta, M. Conjugative Interaction in Olefins. Ph.D. Dissertation,
Harvard University, 1984; Diss. Abstr. Intern. B 1985, 45, 2163 (Order
No. 84-19388).
(4) (a) Hine, J.; Skogland, M. J. J. Org. Chem. 1982, 47, 4766-4770
(further references therein). (b) Hine, J. Structural Effects on Equilibria in
Organic Chemistry; Wiley: New York, 1975; pp 122-126, 270-276. (c)
Hine, J.; Flachskam, N. W. J. Am. Chem. Soc. 1973, 95, 1179-1185.
(5) Pedley, J. B.; Naylor, R. D.; Kirby, S. P. Thermochemical Data of
Organic Compounds, 2nd ed.; Chapman and Hall: London, 1986.
(6) Rogers, D. W.; McLafferty, F. J. Tetrahedron 1971, 21, 3765-3775.
(10) Roth, W. R.; Lennartz, H.-W.; Doering, W. v. E.; Birladeanu, L.;
Guyton, C. A.; Kitagawa, T. J. Am. Chem. Soc. 1990, 112, 1722-1732.
(11) 2,5-Diphenylhexa-1,5-diene (∆rHH , -56.65 kcal/mol) and 2,6-
2
diphenylhepta-1,6-diene (∆rHH , -55.67 kcal/mol) (mean, -56.16 kcal/
2
mol) may be compared with hexa-1,5-diene (∆rHH , -60.1 kcal/mol) and
2
hepta-1,6-diene (∆rHH , -60.2) or twice that of 1-alkylethenes (∆rHH
,
2
-59.82), from which conjugative interaction per phenyl Vis-a`-Vis hydroge2n
becomes -2.0 or -1.83 kcal/mol. These values are almost identical with
that resulting from replacement of vinyl hydrogen by alkyl, -2.01 kcal/
mol (cf. Figure 1), whence conjugative interaction in these examples might
appear to be +0.18 ( 0.5 kcal/mol.
(12) Compare Table 41 (p 1091) of ref 13, in which the value, -2.2
kcal/mol, is proposed.
S0002-7863(97)01040-8 CCC: $14.00 © 1997 American Chemical Society

10948 J. Am. Chem. Soc., Vol. 119, No. 45, 1997
Von E. Doering et al.
system of R,â-conjugated and â,γ-unconjugated isomers at a
single temperature and has therefore furnished free energies of
conjugation. Theoretical scrutiny has had to be based on the
assumption that free energies and enthalpies are equal for
practical purposes, that is, that entropies of conjugation can be
ignored.4 Evaluation of this assumption is but one justification
for dissecting free energy of conjugation into its components.
Another is the desirability of providing more accurate data
against which the enthalpic results of quantum and molecular
mechanical calculational methods may be tested more rigor-
ously.
Given a cleanly catalytic system for establishing equilibrium
constants over a wide range of temperature and an analytical
method of sufficient accuracy, the van’t Hoff method14 offers
an attractive experimental approach. Sensitivity is greatest when
ln K ∼ 0 and decreases as the absolute value of ln K increases.
Accuracy is ultimately limited by the analytical method and
usually begins to falter when ln K falls outside a range of (4.
A major advantage, as noted explicitly by Benson and Bose in
their work on the butenes,15 lies in an accuracy in ∆∆H
significantly higher than can be obtained in general from
differences between two enthalpies of formation determined by
combustion, hydrogenation, or other chemical transformations.
A major series of van’t Hoff studies by Taskinen and co-
workers has included an investigation of the phenylpropenes,
which is relevant to styrene.16 Allylbenzene being the com-
pound of reference, conjugative interaction is relative to
hydrogen. Resulting values of ∆∆fH and ∆∆fS for trans-1-
phenylpropene Vis-a`-Vis 3-phenylpropene are -5.57 kcal/mol
and -3.24 cal mol-1 K-1 (eu), respectively, but the method is
stretched in this instance because ln K ranges between 4 and 7.
Steric factors need to be considered in the interaction of
phenyl with the carbon-carbon double bond beyond that in
styrene itself and trans-â-substituted styrenes, in which they
are assumed to be negligible by definition. These factors are
less important in butadienes, for example, where special
constitutional perturbations are required to enforce a significant
departure from coplanarity.17-19 But in styrenes, a cis orienta-
tion of â-substituents, the presence of substituents in the
R-position or in the ortho positions of the phenyl ring, can be
expected sterically to cause deviations from the coplanar
conformation optimum for π-electron delocalization.
Figure 1. The various thermochemical data, available as Supporting
Information,5,20,21 are translated into enthalpies of hydrogenation
(∆∆HH ) and summarized as mean values (in kcal/mol) ( the standard
2
deviation of the mean, s/n0.5. In parentheses, the number of examples,
n, of each type and the sample standard deviation, s, are also given.
Heats of hydrogenation of the simplest methyl-substituted examples
are shown along side in parentheses. Also shown are values calculated
by Benson’s values for group equivalents.31c Enthalpies of conjugative
interaction, as each additional alkyl group is added, are synonymous
with differences in heats of hydrogenation.
Values for conjugative interactions (thermochemical perturba-
tions) of alkyl groups in mono-, di-, and trisubstituted olefins
(five types) are estimated by an exhaustive examination of the
data on enthalpies of hydrogenation from Rogers,20 enthalpies
of hydration from Wiberg,21 and heats of formation from Pedley,
Naylor, and Kirby.5,22 Data are expressed in Figure 1 as mean
values of enthalpies of hydrogenation (∆HH ), while differences
2
between heats of hydrogenation define the enthalpy of conjuga-
tive interaction of alkyl groups Vis-a`-Vis hydrogen as a reference
(∆∆HConjInter/H). Although no model can serve objectively as a
best reference for comparison, the mean value of several
compounds of the type appeals to us, because it has a standard
deviation of the mean less than the experimental uncertainty
associated with any single example that might be chosen, also
arbitrarily, as a model.
Enthalpies of Conjugative Interaction of Alkyl Ethenes
Because equilibria in simple allyl systems are frequently too
one-sided to furnish much more than qualitative indications of
relative conjugative ability, in this study phenyl is pitted against
sterically innocuous, but significantly conjugating, alkyl groups
to achieve more favorable equilibrium constants. The resulting
enthalpies of conjugation of phenyl Vis-a`-Vis alkyl, to be
translated into enthalpies of conjugation Vis-a`-Vis hydrogen,
require correction for enthalpy of conjugation of the apposite
alkyl groups.
In monosubstituted and trans-disubstituted ethenes, conjuga-
tive interaction of alkyl groups is -2.66 kcal/mol. However,
in the other types of olefins, ∆∆HConjInter/H is not a constant,
but decreases dramatically with substitution. Thus, in cis-
disubstituted ethenes, it is only -1.62 kcal/mol, the difference
of -1.04 kcal/mol being widely equated with an enthalpy of
steric repulsion. Also worthy of note is a decrease in 1,1-
disubstituted ethenes (∆∆HConjInter/H ) -2.01 kcal/mol) of 0.65
kcal/mol, presumably stemming from a smaller but still
significant steric interaction. The absence of this correction in
the Benson scheme leads to the only large discrepancy between
the generalized values in Figure 1 and the comparable Benson
(13) Roth, W. R.; Staemmler, V.; Neumann, M.; Schmuck, C. Liebig’s
Ann. 1995, 1061-1118.
(14) van’t Hoff, J. H. EÄ tudes de dynamique chimique, 1st ed.; Muller:
Amsterdam, 1884.
(15) Benson, S. W.; Bose, A. N. J. Am. Chem. Soc. 1963, 85, 1385-
1387.
(16) Taskinen, E.; Lindholm, N. J. Phys. Org. Chem. 1994, 7, 256-
258.
(17) Honegger, E.; Yang, Z.-z.; Heilbronner, E.; Doering, W. v. E.;
Schmidhauser, J. C. HelV. Chim. Acta 1984, 67, 640-653.
(18) Doering, W. v. E.; Schmidhauser, J. C. J. Am Chem. Soc. 1984,
106, 5025.
values [∆∆HH in bracketsB]. Trisubstituted ethenes, which
2
(20) Rogers, D. W.; Crooks, E. L. J. Chem. Thermodyn. 1983, 15, 1087-
1092. Rogers, D. W.; Crooks, E. L.; Dejroongruang, K. J. Chem. Thermodyn.
1987, 19, 1209-1215. Rogers, D. W.; Dejroongruang, K. J. Chem.
Thermodyn. 1988, 20, 675-680.
(21) Wiberg, K. B.; Wassermann, D. J.; Martin, E.; Murcko, M. A. J.
Am. Chem. Soc. 1985, 107, 6019-6022.
(22) Data involving tert-butyl and similarly sterically crowded groups
have been omitted, but their inclusion would not have influenced the
outcome.
(19) Roth, W. R.; Lennartz, H.-W.; Doering, W. v. E.; Dolbier, W. R.,
Jr.; Schmidhauser, J. C. J. Am Chem. Soc. 1988, 110, 1883-1889.

ConjugatiVe Interaction in Styrenes
J. Am. Chem. Soc., Vol. 119, No. 45, 1997 10949
Chart 1
the trans-â-alkylstyrene of interest to furnish adequately accurate
measurements of relevant equilibrium constants. For elucidation
of the steric effect of an R-methyl group, the set of five
2-methyl-5-phenylhexenes (series 3, Chart 1) has been selected.
The likelihood of residual interaction of the double bond with
phenyl in reference compounds 3b and 3d is much reduced
because they are γ,δ- and δ,ꢀ-unsaturated, respectively, relative
to the R,â-unsaturated isomers of interest.
In the prior literature, equilibrium between 1a(E) and 1b(E)
has been established at single temperatures by Bateman and
Cunneen in methanol at 165 °C catalyzed by sodium methox-
ide: ∆∆G ) -1.32 kcal/mol,23 by Doering and Bragole in
dimethyl sulfoxide and KOC(CH3)3 at 55 °C (92.2:6.6; ∆∆G
) -1.72 kcal/mol),24 and by Carleton at 55 °C in hexameth-
ylphosphoric triamide (HMPT) and NaOCH2CH3 (93.9:3.35;
∆∆G ) -2.18 kcal/mol).2
Present studies have been made over the range 62-111 °C,
in C6D6, with a catalyst developed by Wilson and Osborn,25a
ruthenium hydridonitrosotris(triphenyl)phosphine25b (Chart 1,
series 1, rows 1-3), and over the range 0-97 °C, in HMPT
(mp 7.2 °C after several recrystallizations) with 0.1 M KOC-
(CH3)3 as catalyst (olefin/base, 1:1) (series 1, rows 4-6). Under
the latter conditions, the highest temperature that can be usefully
employed is limited by the appearance of byproducts and by
such a rapid establishment of equilibrium that its position is
seriously disturbed during removal and cooling of samples for
analysis. That true equilibria have been reached is confirmed
by starting from two isomers. To confirm that equilibrium has
not been compromised by a kinetically controlled neutralization
of an appreciable concentration of intermediary, delocalized
carbanion, quenching by D2O is found to generate products
containing negligible amounts of deuterium. To reveal indi-
vidual experimental uncertainties, the natural logarithm of the
percentage of one isomer among all isomers is plotted against
the reciprocal of temperature (K). All data and their plots
relating to the variation of equilibrium concentrations with
temperature in the three series are given as Supporting Informa-
tion.
Resulting differences in free energies [∆∆G(298°C)], enthalpies
(∆∆fH), and entropies (∆∆fS) of formation with standard
deviations are recorded in Chart 1. They pertain to solution
and are related to ∆∆fHgas by differences in enthalpies of
solution and enthalpies of vaporization, which, although not
determined, can probably be assumed negligible within experi-
mental uncertainties in such a series of closely related com-
pounds of low polarity. The near identity of equilibrium
constants obtained in two such disparate solvents as apolar
benzene and aprotic, highly dipolar and polarizable HMPT26
are thought to lend support to the assumption.
Support for the reliability of the experimental method is
afforded by the close agreement of the difference between cis
and trans isomers 1b(Z) and 1b(E) (+1.06 kcal/mol) and the
generic difference (+1.04 kcal/mol; Figure 1); and between the
two reference compounds, 1,1-disubstituted 2c and 1,1,2-
trisubstituted 2b (-1.36 kcal/mol; Chart 1, series 2, row 2),
and the difference in enthalpies of hydrogenation (-1.52 kcal/
mol; Figure 1)27 of the olefin types.
might be expected to have a ∆∆HConjInter/H of -5.32 kcal/mol
(2 × -2.66) Vis-a`-Vis monosubstituted ethenes, in fact have
the value -3.53 kcal/mol, which is quite close to the -3.61
kcal/mol (-5.32 + 1.69 ) -3.63 kcal/mol) value predicted if
the two enthalpies of strain above (cis-, +1.04 kcal/mol; 1,1-,
+0.65 kcal/mol) acted additively and independently without
“buttressing”. There is no additional conjugative interaction
in tetramethylethene (∆∆HH ) -26.31 kcal/mol)!
2
Surrogates for Conjugative Interaction in Styrene
Two sets of â-alkyl-substituted styrenes have been selected
for reevaluation of the enthalpy of conjugative interaction in
styrene. Behind this approach lie two plausible assumptions:
that interaction of the double bond with the â-alkyl group has
not altered its interaction with phenyl; and that steric perturbation
in trans isomers, if not zero, is well within experimental
uncertainties. The set of 1-phenylbutenes (1a(E), 1b(E), 1a(Z),
and 1b(Z); series 1 in Chart 1) is otherwise uncomplicated, while
the set of 1-phenyl-3-methylbutenes (2a(E), 2b, 2a(Z), and 2c;
series 2 in Chart 1) incorporates two reference standards of
differing degrees of substitution, one of which, it is hoped,
should have a free energy of formation close enough to that of
(23) Bateman, L.; Cunneen, J. I. J. Chem. Soc. 1951, 2283-2289.
(24) Doering, W. v. E.; Bragole, R. Tetrahedron 1966, 22, 385-391.
(25) (a) Wilson, S. T.; Osborn, J. A. J. Am. Chem. Soc. 1971, 93, 3068-
3070. (b) Bradley, J. S.; Wilkinson, G. Inorg. Synth. 1977, 17, 73-74.
(26) Reichardt, C. SolVents and SolVent Effects in Organic Chemistry;
VCH: Weinheim, 1988; e.g., p 79.
(27) The position of equilibrium between 2a(E) and 2b (∆∆G ) +0.06
kcal/mol) calculated at 165 °C from the experimental ∆∆fH and ∆∆fS
accords satisfactorily with the value of +0.25 obtained by Batemann and
Cunneen at 165 °C.23

10950 J. Am. Chem. Soc., Vol. 119, No. 45, 1997
Von E. Doering et al.
To avoid confusion in discussion, distinction is made between
a theoretical enthalpy of conjugation (∆∆HConjEnth), which is
defined as π-electron delocalization alone, and an empirical
enthalpy of conjugative interaction (∆∆HConjInter), which is
comprised additionally of such factors as 1,4-π-repulsions,
enthalpy of strain, and differences in σ-bond strengths (sp2-
sp2 Vis-a`-Vis sp2-sp3).28
To generalize a value for ∆∆HConjInter/Alk from the difference,
for example, between ∆∆fH1a(E) and ∆∆fH1b(E), implies an
assumption that methyl and ethyl are indistinguishable from each
other and alkyl. Subject to this reservation, a weighted mean
of -2.55 kcal/mol for ∆∆HConjInter/Alk of phenyl Vis-a`-Vis alkyl
is derived directly (Chart 1, series 1, row 7).
and free energy at equilibrium30 quite comparable to those found
for (E)- and (Z)-2-phenylbut-2-ene.
Equilibrations over the range 81-154 °C in HMPT/KOC-
(CH3)3 are unexceptional, although a declining recovery limits
the highest practicable temperature. Equilibrium constants
involving 3a(E), 3a(Z), and 3b fall within the experimentally
favorable range of 2 > ln K > -2. Values for ∆HConjInter/Alk
are derived directly by comparisons of 3a(E) with 3b, -1.1
kcal/mol, 3c with 3d, -0.4 kcal/mol, and 3a(Z) with 3b, 0.0
kcal/mol. Corresponding values of ∆HConjInter/H are -3.7, -3.1,
and -2.7 kcal/mol, respectively. The sensitivity to what
intuitively seem to be very small steric differences in 3a(E)
and 3c is remarkable.
The cis-trans increment in the 3a pair deserves comment.
3a(E) contains a conventional unit of cis strain (methyl vs
isobutyl: +1.04 kcal/mol), which should vanish in passing to
3a(Z), but in fact is more than compensated by an increase of
+1.08 kcal/mol. This striking difference between expectation
and fact in 3a(E) and 3a(Z) of +2.12 kcal/mol is doubtless steric
in origin, but how it should be allocated between a decrease in
π-electronic factor attributable to significant divergence from
coplanarity and nonbonded steric repulsion between twisted
phenyl and isobutyl is elusive.
Entropy is not a negligible factor. In series 1, it translates to
a contribution of 0.5-0.7 kcal/mol to the differences in free
energy of 1a(E) and 1b(E) while in series 2 it translates to 1.0
kcal/mol or more; and in series 3, entropy contributes 0.3-0.4
kcal/mol to the differences between free energies and enthalpies.
To extract a similar ∆∆HConjInter/Alk from ∆∆fH of 2a(E), a
disubstituted olefin, and 2b, a trisubstituted olefin, requires
correction by -0.87 ( 0.19 kcal/mol, the difference in enthal-
pies of hydrogenation of trans-1,2-dialkylethenes corresponding
to 2a(E) and 1,1,2-trialkylethenes corresponding to 2b (Figure
1). A similar extraction from ∆∆fH of 2a(E) and 2c requires
correction by +0.65 ( 0.14 kcal/mol, the difference between a
trans-1,2- and a 1,1-disubstituted alkene. The weighted mean
of the two corrected differences, 2.44 ( 0.15 and 2.28 ( 0.14
kcal/mol, respectively, is -2.37 kcal/mol. Agreement with
∆∆HConjInter/Alk from series 1, -2.55 kcal/mol, is impressive.
The combined result from series 1 and 2 is ∆∆HConjInter/Alk
)
-2.5 ( 0.2 kcal/mol. This is the value recommended for trans-
â-alkyl-substituted styrenes and styrene itself. Correction by
-2.66 kcal/mol generates a value, -5.1 ( 0.2 kcal/mol, for
conjugative interaction Vis-a`-Vis hydrogen as reference
(∆∆HConjInter/H).
Translation into Enthalpies of Formation
Values of ∆∆fHexp can be transformed into useful enthalpies
of formation by focusing on the unconjugated isomerss1a(E),
2b and 2c, and 3b and 3d. Their enthalpies of formation can
be estimated from compounds already in the literature in a
straightforward and reliable manner. The preferred ways
involve (i) the empirically based method of group equivalents
of Franklin31a as elaborated by Benson;31b,c (ii) extraction of a
value for the change in enthalpy of formation, ∆∆fH°, on
replacement of a hydrogen in a methyl group by phenyl and
application of this value to the appropriate alkene; and (iii)
subtraction from the appropriate phenylalkane of the appropriate
enthalpy of hydrogenation given in Figure 1. From these
estimates, a mean value for ∆fH°est of trans-1-phenylbut-2-ene
(1b(E)), +24.32 ( 0.23 kcal/mol, can be derived (series 1, row
8).32 For 2-methyl-4-phenylbut-2-ene (2b) and 2-methyl-4-
phenylbut-1-ene (2c), mean values for ∆fH°est of +16.62 ( 0.37
and +18.49 ( 0.33 kcal/mol, respectively, are derived (series
2, row 4).33 Finally, for 2-methyl-5-phenylhex-2-ene (3b) and
2-methyl-5-phenylhex-1-ene (3d), mean values for ∆fH°est of
+16.62 ( 0.37 and +18.49 ( 0.33 kcal/mol, respectively, are
obtained (series 3, row 4).34 Addition of the appropriate values
A cis effect in the conjugated styrenes, 1a(E) and 1a(Z)
(series 1, row 7: +2.60 kcal/mol) and 2a(E) and 2a(Z) (series
2, row 2: +2.37 kcal/mol), is more than twice the value for
the type, cis-1,2-dialkyl-substituted olefins (+1.04 kcal/mol).
The effect is plausibly ascribed to a compromise between
maximization of π-electron delocalization by attainment of
coplanarity and minimization of nonbonded steric repulsion.
Values of ∆∆HConjInter/Alk for cis-â-alkyl-substituted styrenes are
correspondingly lower: -1.01 kcal/mol from 1a(Z) (series 1,
row 7) and -1.11 kcal/mol from 2a(Z) (+0.80 corrected by
-1.91 kcal/mol, Figure 1). The mean value for ∆∆HConjInter/H
is -2.7 kcal/mol (compare the value from Taskinen for cis-1-
phenylpropene, -2.94 kcal/mol).16
Steric Factor in r-Substituted Styrenes
In the 2-phenyl-5-methylhexenes (Chart 1, series 3), the
reference compound, 3b, is trisubstituted as is 3a(E) and 3a(Z),
while 3d like 3c is 1,1-disubstituted. As in series 1,
∆∆HConjInter/Alk is derived directly from values of ∆∆fH without
correction for differences in substitution. Establishment of
stereochemistry in 3a(E) and 3a(Z) is based, first, on comparison
of the chemical shifts of their â-hydrogens at 5.80 and 5.48
ppm, respectively, with the two â-hydrogens cis and trans to
the phenyl group in styrene (5.3 vs 4.8), p-chlorostyrene (5.73
vs 5.28), R-methylstyrene (5.34 vs 4.99), and p-isopropylstyrene
(5.32 vs 4.99), on comparison with the vinyl hydrogens in trans-
(7.03) and cis- (6.57) stilbenes, and the â-hydrogens in 1a(E)
(6.10) and 1a(Z) (5.53), and 2a(E) (6.08) and 2a(Z) (5.39);
second, by observation of a small but consistent nuclear
Overhauser effect upon saturation of the aromatic hydrogens
and the R-methyl groups; and third, differences in UV spectra29
(30) Fort, A. W.; Girard, C. A. J. Am. Chem. Soc. 1961, 83, 3449-
3453.
(31) (a) Franklin, J. L. Ind. Eng. Chem. 1949, 41, 1070-1076. (b)
Benson, S. W.; Cruikshank, F. R.; Golden, D. M.; Haugen, G. R.; O’Neal,
H. E.; Rodgers, A. S.; Walsh, R. Chem. ReV. 1968, 68, 1513-1524. (c)
Benson, S. W. Thermochemical Kinetics, 2nd ed.; Wiley: New York, 1976.
(32) (i) Benson value (p 272):31c +24.70 kcal/mol. (ii) For the change
in enthalpy of formation when benzene and CH3R are dehydrogenated to
C6H5CH2R (∆∆fH°(CH3RfC6H5CH2R), a mean value of +26.98 ( 0.14
kcal/mol can be derived from four pairs of ∆fH° (in kcal/mol) in the
literature:5 ethylbenzene (+7.15)/ethane (-20.03); 1-phenylpropane (+1.89)/
propane (-25.02); 1-phenylbutane (-3.13)/n-butane (-30.02); 2-methyl-
1-phenylpropane (-5.14)/2-methylpropane (-32.07). Application of this
value to trans-butene (-2.72 kcal/mol) leads to +24.26 kcal/mol. (iii) ∆fH°-
(1-phenylbutane) ) -3.135 + 27.25 (Figure 1) ) +24.12 kcal/mol;
alternatively, ∆fH°(n-butane) - 30.02 + 26.98 + 27.25 ) +24.21 kcal/
mol. A mean value for ∆fH° of trans-1-phenylbutene-2 [1b(E)] is +24.32
( 0.26 kcal/mol.
(28) Dewar, M. J. S.; Schmeising, H. N. Tetrahedron 1959, 5, 166-
178. Dewar, M. J. S.; Schmeising, H. N. Tetrahedron 1960, 9, 96-120.
(29) Cram, D. J. J. Am. Chem. Soc. 1949, 71, 3883-3889.

ConjugatiVe Interaction in Styrenes
J. Am. Chem. Soc., Vol. 119, No. 45, 1997 10951
of ∆∆fHexp (series 1, row 7; 2, row 2; and 3, row 2) to these
reference hydrocarbons generates estimated enthalpies of forma-
tion, ∆fH°est, of the remaining compounds in the series. In series
2 and 3, two sets of nearly identical ∆fH°est can be generated,
depending on which of two reference compounds is selected.
These new thermochemical data are potentially useful for
evaluating the accuracy of quantum and molecular mechanical
approaches to enthalpies of formation. We have applied the
most recent version of the MMEVBH program of Roth,13 which
combines the MM2 program of Allinger as modified in
MM2ERW35 with the program of the extended valence bond
Hamiltonian developed by Maynau and Malrieu36 and the MM3
program of Allinger.37 Corrections for TOP (torsional) and POP
(Boltzmann) have not been made.38 Although this omission
affects the calculated values of enthalpies of formation, except
as noted, it has little effect on the differences. No exhaustive
evaluation is contemplated here.39
Comparisons of cis-trans isomers seems most favorable, as
they involve “isoBensonian” geometrical isomers40 and no
significant corrections for POP and TOP. For the comparison
of 1b(E) and 1b(Z), both methods do very well, while
MMEVBH does rather less well than MM3 in the comparison
of 1a(E) and 1a(Z) [Chart 1, series 1: row 7, ∆∆fHexp ) -2.60
kcal/mol, versus row 9, ∆∆fH°MMEVBH ) -1.85 kcal/mol, and
row 11, ∆∆fH°MM3 ) -2.45 kcal/mol], and in the comparison
of 2a(E) and 2a(Z) [experimental difference of -2.37 kcal/
mol (Chart 2, row 2); calculated differences, -1.49 (MMEVBH,
row 5) and -2.30 kcal/mol (MM3, row 7)]. In the comparison
of 3a(E) and 3a(Z), MMEVBH is substantially superior to
MM3: experimental, -1.08 kcal/mol (Chart 3, row 2);
MMEVBH, -0.62 kcal/mol (row 5); MM3, -0.10 kcal/mol.
The experimental value for ∆∆H(ConjInter/Alk) (1a(E) minus
1b(E), Chart 1, row 7: -2.55 kcal/mol) compares with the
values of -1.47 kcal/mol calculated by MMEVBH (row 9) and
of -1.13 kcal/mol by MM3 (row 11). Comparison between
2a(E) and 2b is unsatisfactory, MMEVBH at least being in the
right direction. Other comparisons are left to the reader. There
is an apparent need for a critical evaluation of computational
methods41 in the prediction of the enthalpic component of
equilibria.
Taken together, the susceptibility of enthalpy of conjugative
interaction to the cis effect and R-substitution should serve as
a warning against uncritical application of a single value for
styrene conjugation. The ultimate usefulness of a concept of
“intrinsic” π-electron delocalization (enthalpy of conjugation)
in styrenes generally will depend on the ability to estimate
accurately the steric factor and the availability of a good force
field for the phenyl-double bond dihedral angle. There seems
to be considerable room for improvement in the continuing effort
to gain intellectual control over the connection between structure
and thermochemistry. At the interface of experiment and
theoretical constructions, there remains great need for congruity
at least at the level of (1.0 kcal/mol proposed by Michael
Dewar for “chemical accuracy” over a quarter of a century ago,42
if not more realistically for the immediate future at the level of
(0.3 kcal/mol.
(33) (i) Benson values for ∆fH°est(2b), +16.25 kcal/mol, and for ∆fH°est
-
(2c), 18.79 kcal/mol. (ii) From known heats of formation (in kcal/mol) of
2-methylbut-2-ene (-9.99) and 2-methylbut-1-ene (-8.44),5 the addition
of (∆∆fH°(CH3RfC6H5CH2R), +26.98, leads to ∆fH°est(2b), +16.99, and
∆fH°est(2c), +18.54, respectively. (iii) Addition of ∆∆fH°(CH3RfC6H5-
CH2R)32 to 2-methylbutane (-36.74 kcal/mol)5 leads to ∆fH°est(2-methyl-
4-phenylbutane) of -9.76 kcal/mol. Whence additions of the heats of
dehydrogenation of 1,1,2-trialkyl-substituted ethenes (Figure 1, +26.38 kcal/
mol) and 1,1-dialkyl-substituted ethenes (+27.90 kcal/mol), respectively,
lead to ∆fH°est(2b) of +16.62 and ∆fH°est(2c) of +18.14 kcal/mol. Mean
values of +16.62 ( 0.37 and +18.49 ( 0.33 kcal/mol are derived for
∆fH°est(2b)and ∆fH°est(2c), respectively.
Steric Effect of o-Methyl Groups
Interference with the attainment of coplanarity between the
benzene ring and the double bond can also be explored by
introducing methyl groups at positions 2′ and 6′.2,43 These
studies have included examination of the 1-(2′,4′-dimethylphe-
nyl)- and the 1-(2′,4′,6′-trimethylphenyl)-n-butenes and n-
pentenes. Because equilibrations have been established at a
single temperature only, 90 °C in (unpurified) HMPT with
sodium ethoxide or potassium tert-butoxide as catalyst, discus-
sion of the results collected in Chart 2 is limited. Results of
calculations by the two molecular mechanical programs above
are also included.
Introduction of a second o-methyl group into 4a(E) to
generate 5a(E) attenuates the free energy of conjugation in 4a(E)
by 0.9 kcal/mol consistent with the expected greater deviation
from coplanarity (bond angles indicated in the formulas come
from MMEVBH calculations). A similar attenuation in ∆∆G
is seen in the comparison of the cis-trans pairs, 4a(E) and
4a(Z), and 5a(E) and 5a(Z). Parenthetically, the slightly greater
difference in ∆∆G in the pair 1a(E) and 1b(E) (∼-2.0 kcal/
mol; Chart 1, series 1, rows 2 and 4) compared to that in 4a(E)
and 4b(E) (-1.57 kcal/mol; Chart 2) is in good agreement with
the finding of Hine and Skogland in the intramolecular
competition in the system of 1-phenyl-3-(o-tolyl)propenes
(∆∆G353 ) -0.46 kcal/mol).44
(34) (i) Benson values for ∆fH°est(3b), +5.60 kcal/mol, and for ∆fH°est
-
(3c), +7.54 kcal/mol. (ii)
A
mean value (in kcal/mol) for
∆∆fH°(RCH2R′fC6H5CHRR′) of +25.78 ( 0.26 is derived from three
pairs of ∆fH° in the literature:5 isopropylbenzene(+25.98)/propane; sec-
butylbenzene(+25.86)/n-butane; and phenylcyclohexane(+25.49)/cyclo-
hexane. From ∆HH of 2-methylhex-2-ene (-26.00 kcal/mol) and 2-meth-
2
ylhex-2-ene (-27.68 kcal/mol)20 to 2-methylhexane (∆fH° ) -46.51 kcal/
mol)5 and addition of the increment, +25.78 kcal/mol, ∆fH°est of 3b and
3d of +5.27 and 6.95 kcal/mol, respectively, are derived. (iii) Addition of
the increment, +25.78 kcal/mol, to the ∆fH° of 2-methylhexane (-46.51
kcal/mol)5 leads to ∆fH°est of the common product of hydrogenation,
2-methyl-5-phenylhexane, of -20.73 kcal/mol. Dehydrogenation based on
the appropriate enthalpies of hydrogenation of the type (Figure 1) leads to
∆fH°est of 3b and 3d of +5.65 and 7.17 kcal/mol, respectively. Mean values
of +5.51 ( 0.21 and +7.22 ( 0.30 kcal/mol are included in Chart 1, series
3, row 4.
(35) Roth, W. R.; Adamczak, O.; Breuckmann, R.; Lennartz, H.-W.;
Boese, R. Chem. Ber. 1991, 124, 2499-2521.
(36) Sa¨ıd, M.; Maynau, D.; Malrieu, J.-P.; Garcia Bach, M.-A. J. Am.
Chem. Soc. 1984, 106, 571-579.
(37) Allinger, N. L.; Yuh, Y. H.; Lii, J.-H. J. Am. Chem. Soc. 1989,
111, 8551-8566.
(38) We express our sincere appreciation to Professor Wolfgang Roth
for having checked both the EVBH and MM3 calculations. The EVBH
program is available on application to Prof. Dr. W. R Roth, Ruhr-Universita¨t
Bochum, Organische Chemie, Postfach 10 21 48, D-44780 Bochum,
Germany. He notes that TOP and POP corrections to the MM3 results have
not been made, and that, although Allinger has not discussed this matter
explicitly in connection with MM3, one can infer from published examples
that a correction for (TOP + POP) of 0.75 kcal/mol has been made for
each freely rotating bond (i.e., <7 kcal/mol barrier; excluding bonds to
methyl). In 2c, 3c, and 3d this correction would raise the calculated values
of ∆∆fH° by 0.75 kcal/mol, all others remaining unchanged.
(39) “...it’s one of those things that once you have done it with
supervision of someone who really understands it, then you will understand
it. But working from the literature, it is not so easy.”
In series 6, 7, and 8, the o-methyl factor has been combined
with the R-methyl factor. The near identity of series 7 and 8
confirms that a methyl group in the 4′ position is without effect.
This result agrees with that of Bushby and Ferber in the series
(41) For example: Gundertofte, K.; Liljefors, T.; Norrby, P.-O.; Pet-
tersson, I. J. Comput. Chem. 1996, 17, 429-449.
(42) Dewar, M. J. S. XXIIIrd International Congress of Pure and Applied
Chemistry; Butterworths: London, 1971; Vol. 1, pp 1-30.
(40) Compounds describable by identical summations of group equivalent
values. Indeed, the entire evaluation of several molecular methods focuses
on differences of conformational energies.41
(43) Proctor, George R., unpublished work on the 1-phenyl-3-mesityl-
propene system; Yale University, 1963, supported by the Norman Fund.
(44) Hine, J.; Skogland, M. J. J. Org. Chem. 1982, 47, 4758-4766.

10952 J. Am. Chem. Soc., Vol. 119, No. 45, 1997
Von E. Doering et al.
Chart 2
Vis series 6. On the assumption that this difference would persist
if differences in enthalpy had been measured, this apparent
increase is more likely the result of a large increase in steric
factor in 7b(E) than an enhancement of π-electron delocalization
in 7a(Z)! A qualitative explanation is possibly to be found in
the inability of substitutents on the sp3 R-carbon to escape steric
interference from the methyl groups at 2′ and 6′ positions of
the benzene ring. How far out of the plane awaits crystal-
lographic analysis; meanwhile the dihedral angles revealed by
MM at globally minimized steric energies in system 7 are close
to 90°.
Conclusions
The main lesson is clear: reliance on a single value for
conjugative interaction in styrene is quite unjustified for
predictive purposes in more complicated systems. The cis
effect, R-substituents, and ortho substituents lead to marked and
variable attenuations of conjugative interaction. A second lesson
teaches the inadequacy of equating enthalpies and free energies
of conjugative interaction: entropies can be large and adversely
influential on the position of equilibrium of conjugated isomers
relative to their unconjugated isomers. Although it may be
useful to define an intrinsic enthalpy of π-electron delocalization,
empirical conjugative interaction is not its measure, but is rather
a composite thermochemical quantity thought to consist of an
interplay between an enthalpy-raising, nonbonded steric interac-
tion per se and its attenuation of enthalpy-lowering π-electron
delocalization by enforcement of a deviation from coplanarity.
At the present time, this interplay seems not to be well
calculable by molecular mechanical programs with an accuracy
appropriate to a useful prediction of positions of equilibria. The
thermochemical data brought to light here on three sets of
isomeric compounds of simple structure offer one of few
opportunities to test the accuracy of molecular mechanical
programs that is neither redundant nor has already been
consumed by parametrization.46 Additional sets are desirable.
Experimental Section
General Procedures. NMR spectra are determined on Bruker AM
Series 250, 300, 400, or 500N in C6D6 unless otherwise noted, and are
referenced to C6D5H and reported in ppm, δ scale; coupling constants
(J) are reported in Hz. Spin-lattice relaxation times (T1) are determined
by the inversion recovery method. IR spectra, recorded on a Mattson
Galaxy Series FTIR 3000, are measured, unless otherwise noted, in
KBr pellets and reported in cm-1
. UV spectra are measured on a
Hewlett Packard 8452A diode array spectrophotometer in spectrograde
hexane and are reported as λmax in nm (log ꢀ). GC-EIMS (70 eV) are
obtained on a Hewlett Packard 5890 Series II gas chomatograph (HP
5 MS, 30 m × 0.25 mm) and a Hewlett Packard 5972 Series mass
selective detector and are reported as m/z and intensity (as %) of major
peak.
Analytical GC are effected on a Hewlett Packard 5890 A gas
chromatograph with J & W Scientific columns: (A) Megabore, DB-1,
30 m × 1.5 mm; and (B) DB 1701, 30 m × 1.0 mm. Preparative
GLC are performed on a Varian Aerograph Model 90-P, column C:
20% Carbowax on Anakrom AS 50/60; 3 m, i.d. 5 mm, 20 psi He (2
mL/s); injector 175 °C, column 175 °C, detector 220 °C unless
otherwise noted. Melting points are uncorrected. Solvents are
redistilled before use: THF from sodium/benzophenone; mixed hexanes
and ethyl ether from LiAlH4; CHCl3 and CH3CN from CaCl2; methanol
from Mg; pyridine and 2,6-lutidine from KOH; tert-butyl alcohol from
CaH2; and CH3I through a column filled with copper wire.
of 1-phenyl-3-(p-tolyl)propenes.45 Assignment of configuration
to the cis-trans pairs in series 6, 7, and 8, illustrated in detail
for 8a(E) and 8a(Z) in Figure 5 included in the Supporting
Information, is based on analogy with chemical shifts in the
NMR spectra of styrene and o-methylstyrene, in which the vinyl
hydrogens trans to the benzene ring are upfield of the cis
hydrogen but downfield in 2′,4′,6′-trimethylstyrene.
Materials. In series 1-3, Wittig reactions are carried out by
preparing the ylide by addition of a 2.5 M solution of n-butyllithium
(10% molar excess) in hexanes under argon over a period of 1 h to a
In series 7 (and 8), the second o-methyl group appears to
have occasioned a dramatic increase in ∆∆GConjInter/alk, Vis-a`-
(45) Bushby, R. J.; Ferber, G. J. J. Chem. Soc., Perkin Trans. 2 1976,
1683-1688.
(46) Cieplak, P.; Howard, A. E.; Powers, J. P.; Rychnovsky, S. D.;
Kollman, P. A. J. Org. Chem. 1996, 61, 3662-3668.

ConjugatiVe Interaction in Styrenes
J. Am. Chem. Soc., Vol. 119, No. 45, 1997 10953
vigorously stirred suspension of the apposite triphenylphosphonium
halide in anhydrous THF maintained at 0 °C in an ice/water bath. After
the mixture is stirred for an hour at 0 °C, aldehyde or ketone in
anhydrous THF is added over a 1-h period. The reaction mixture is
warmed to room temperature and stirred for 1.5 h. Workup procedure
A includes filtration over a bed of Celite (pentane), concentration in
Vacuo (<15 °C), and subsequent purification over 10 g of silica
(pentane); procedure B includes concentration in Vacuo, thorough
extraction of the residue with hexanes (total, 500 mL), filtration, drying
over K2CO3, and evaporation of solvent to leave an oil, which is then
purified by vacuum distillation.
Grignard reagents are prepared in the usual manner, treated with
aldehyde or ketone under argon (<40 °C), and worked up convention-
ally.
Series 1. (E)-1-Phenylbut-2-ene (1b(E)). 1b(E) was obtained from
(Aldrich Chemical Co.) and purified by distillation, but it still remained
contaminated by 5% of 1b(Z): 1H NMR (300 MHz) 7.25-7.00 (m,
5H), 5.51 (m, 1H), 5.37 (m, 1H), 3.19 (d, 2H, J ) 6.5, H-1), 1.54 (dm,
3H, J ) 5.9, H-4).
(E)-1-Phenylbut-1-ene (1a(E)). This compound is prepared from
1.5 g of 1b(E) by equilibration in 10 mL of a 0.05 M solution of
potassium tert-butoxide in HMPT in a Schlenk flask with stirring under
argon for 30 min at room temperature. The mixture is then worked
up by quenching with 50 mL of water and extraction with 100 mL of
pentane. The extract is washed twice with H2O (50 mL each) and 50
mL of brine and concentrated in Vacuo at 0 °C. Distillation then affords
0.9 g (60%) of 1a(E) in a purity of 96% (GC): 1H NMR (500 MHz)
7.25-7.00 (m, 5H), 6.30 (d, 1H, J ) 15.8, H-1), 6.10 (dt, 1H, J ) 6.6,
H-2), 2.04 (m, 2H, J ) 7.5, 1.4, H-3), 0.95 (t, 3H, H-4).
(E)- and (Z)-1-Phenylbut-1-ene (1a(E) and 1a(Z)). 1a(E) and
1a(Z) are preapred by the Wittig reaction (Vide supra) from 16.5 g of
isopropyltriphenylphosphonium bromide and 2.8 g of benzaldehyde,
which after distillation in Vacuo affords 2.7 g (77%) of a 43:57 mixture
of 1a(E) and 1a(Z): 1H NMR (500 MHz) 7.25-7.00 (m, 5H), 6.37 (d,
1H, J ) 11.6, H-1), 5.53 (dt, 1H, J ) 7.3, H-2), 2.21 (dm, 2H, J )
7.5, 1.8, H-3), 0.89 (t, 3H, H-4). 1a(Z) has also been prepared by
hydrogenation of 1-phenylbutyne47 with 10% Pd on CaCO3 in methanol
(interrupted after the uptake of 0.95 mol equiv of H2).
separation of evolved water. The usual workup (washing with saturated
NaHCO3 (three 50-mL portions) and brine, drying over anhydrous
K2CO3, filtering, and concentrating in Vacuo at 0 °C), affords, after
distillation, 3.8 g (86%) of an 84:16 mixture of 2b and 2c (bp 40 °C/
0.1 mm). Pure 2b is obtained by preparative GLC (column C) and
subsequent vacuum distillation (0.8 g from 1.2 g): 1H NMR (300 MHz)
7.2-7.0 (m, 5H), 5.34 (tm, 1H, J ) 7.3, 1.5), 3.26 (d, 2H), 1.63 (d,
3H), 1.55 (s, 3H); IR 3100-2800, 1604, 1494, 740, 697; UV 252 (2.88),
208 (4.07).
2-Methyl-4-phenylbut-1-ene (2c). Preparation by a Wittig reaction
from methyltriphenylphosphonium bromide (30.5 g) and 4-phenyl-2-
butanone (Aldrich; 7.62 g) and distillation (bp 40 °C/0.25 mm) yields
4.98 g (66%) of 2c as a colorless liquid: 1H NMR (250 MHz) 7.0-
7.2 (m, 5 H), 4.77 (m, 1H, H-1), 4.74 (m, 1H, H-1), 2.62 (t, 2H, J )
8.0, H-4), 2.19 (t, 2H, H-3), 1.61 (s, 3H); IR 3100-2800, 1650, 1604,
1496, 888, 745, 698; UV 262 (2.11), 208 (3.92), 200 (3.99).
Series 3. (E)- and (Z)-2-Phenyl-5-methylhex-2-ene (3a(E) and
3a(Z)). (1-Phenylethyl)triphenylphosphonium bromide is prepared from
26.2 g of triphenylphosphine and 18.5 g of 1-phenylethyl bromide
(Aldrich Chemical Co.) in 30 mL of toluene under reflux for 10 h:
colorless crystals; 40.0 g (89%); mp 222 °C; 1H NMR (250 MHz,
CDCl3) 7.75-7.45 (m, 15H), 7.20-7.00 (m, 5H), 6.48 (m, 1H, J )
7.0, 7.0), 1.72 (dd, 3H, J ) 19.1). Wittig reaction (Vide supra) [40 g
of phosphonium bromide and 4.7 g of 3-methylbutanal (Aldrich
Chemical Co.)] affords a yellow oil, bp 65 °C/0.1 mm, 5.43 g (57%),
as an 84:16 mixture of 3a(E) and 3a(Z). This mixture (2 g) is separated
by preparative GLC (column C; injector 180 °C; column, 185 °C;
detector, 185 °C) and subsequent vacuum distillation to give 0.95 g of
3a(E) and 0.44 g of 3a(Z). 3a(E): 1H NMR (500 MHz) 7.34 (d, 2H),
7.18 (t, 2H), 7.09 (t, 1H), 5.80 (tq, 1H, J ) 7.4, 1.2), 1.99 (t, 2H), 1.90
(s, 3H), 1.61 (sept, 1H, J ) 6.7), 0.89 (d, 6 H); 13C NMR (125.8 MHz)
145.19, 136.31, 128,94, 128.04, 127.27, 126.59, 38.71, 29.96, 23.08,
16.56; IR 3100-2800, 1598, 1494, 755, 696; UV 246 (4.08), 210 (4.04);
GC-MS 174 (8) [M+], 131 (100) [M+ - C3H7], 118 (15) [C9H10+], 91
(55) [C7H7+]. 3a(Z): 1H NMR (500 MHz) 7.17 (m, 4H), 7.06 (t, 1H),
5.48 (td, 1H, J ) 7.3, 1.3), 2.00 (t, 3H), 1.94 (m, 2H, J ) 1.2), 1.55
(sept, 1H, J ) 6.7), 0.72 (d, 6H); 13C NMR (125.8 MHz) 143.37,
137.82, 128.86, 128.85, 127.33, 127.17, 38.96, 29.79, 26.35, 22.93;
IR 3100-2800, 1600, 1494, 762, 700; UV 236 (3.80), 204 (4.07); GC-
MS 174 (10) [M+], 131 (100) [M+ - C3H7], 118 (25) [C9H10+], 91
(53) [C7H7+].
4-Phenylbut-1-ene (1c). A sample from Aldrich Chemical Co. is
purified by distillation: 1H NMR (300 MHz) 7.20-6.95 (m, 5H), 5.73
(m, 1H, H-2), 4.98 (dm, 1H, H-1), 4.93 (dm, 1H, H-1), 2.52 (t, 2H, J
) 7.8, H-4), 2.21 (m, 2H, H-3).
2-Methyl-5-phenyl-hex-1-ene (3d). (a) 1-Bromo-2-phenylpropane
is prepared by a modification of the procedure of Brown and Lane.50
To a solution of 35.5 g (0.3 mol) of R-methylstyrene in 90 mL of
anhydrous THF is added 100 mL (0.1 mol) of a 1 M solution of BH3
in anhydrous THF under argon over a period of 1 h (0-5 °C). After
being stirred for 5 h at 25 °C, the reaction mixture is carefully quenched
with 2 mL of methanol and cooled to -10 °C. Addition of 20 mL
(0.4 mol) of bromine, followed by 100 mL (0.5 mol) of a 25% solution
of NaOCH3 in CH3OH (T <0 °C) with stirring for 30 min, addition of
saturated aqueous K2CO3, thorough extraction with hexanes (4 × 150
mL), washing the combined hexane extracts with water (3 × 100 mL)
and 100 mL of brine, drying the solution over MgSO4, and finally
filtering and concentrating the solution in Vacuo affords a yellow liquid
(33 g) consisting of a 75:25 mixture of 1-bromo-2-phenylpropane and
2-phenylpropanol. This mixture is heated under reflux in 200 mL of
hexanes with 1.2 g of LiAlH4 for 1.5 h, cooled, filtered, and
concentrated in Vacuo to a crude product, which is purified by vacuum
Series 2. 1-Phenyl-3-methylbut-1-enes (2a(E) and 2a(Z)). These
olefins are obtained by a Wittig reaction between benzyltriphenylphos-
phonium chloride (33.2 g) and isobutyraldehyde (3.7 g). Vacuum
distillation (bp 40 °C/0.25 mm) yields 4.62 g (31.6 mmol, 62%) of a
76:24 mixture of 2a(E) and 2a(Z). Separation of 1.8 g of product into
the pure isomers by preparative GLC (column C) and subsequent
distillation in vacuo provides 0.39 g of 2a(E) and 0.88 g of 2a(Z), the
latter also having been prepared by the following sequence, alkylation
of phenylacetylene48 to phenylisopropylacetylene and partial hydrogena-
tion of the latter49 (for details, see Carleton dissertation).2 2a(E): 1H
NMR (500 MHz) 7.26 (d, 2H), 7.14 (t, 2H), 7.04 (t, 1H), 6.30 (d, 1H,
J ) 16, H-1), 6.08 (dd, 1H, J ) 6.9, H-2), 2.29 (m, 1H, J ) 6.8, H-3),
0.99 (d, 6H, H-4); IR 3100-2800, 1651, 1597, 1493, 966, 789, 698;
UV 252 (4.15), 210 (4.11); GC-MS 146 (37) [M+], 131 (100) [M+
-
CH3], 91 (42) [C7H7+]. 2a(Z): 1H NMR (500 MHz): 7.25 (d, 2H),
7.16 (t, 2H), 7.05 (t, 1H), 6.31 (d, 1H, J ) 11.6, H-1), 5.39 (dd, 1H,
J ) 10.2, H-2), 2.90 (m, 1H, J ) 6.6, H-3), 0.93 (d, 6H, H-4); IR
3100-2800, 1598, 1494, 762, 789, 697; UV 242 (4.09), 208 (4.10).
3-Methyl-1-phenylbut-2-ene (2b). This compound is prepared by
the dehydration of 2-methyl-4-phenylbutan-2-ol (obtained from ben-
zylacetone and ethereal methyllithium) at 0 °C: bp 76 °C/0.1 mm; mp
distillation to afford 19.5
g (33%, bp 70 °C/0.1 mm) of
1-bromo-2-phenylpropane: 1H NMR (250 MHz) 7.15-6.80 (m, 5H),
3.17 (dd, 1H, J ) 9.8, 6.0), 3.03 (dd, 1H, J ) 8.0), 2.77 (m, 1H, J )
6.9), 1.11 (d, 3H). The bromide (12.2 g) is converted to its Grignard
reagent (Vide supra), treated with 6.0 g of freshly distilled methallyl
chloride in 10 mL of THF (40 °C), and work up by procedure B to
yield 5.5 g of crude product. Distillation in Vacuo provides 3.9 g (37%)
of pure 2-methyl-5-phenylhex-1-ene (3d) (bp 65 °C/0.1 mm): 1H NMR
(500 MHz) 7.20-7.00 (m, 5H), 4.75 (d, 1H, J ) 0.5), 4.71 (d, 1H),
2.53 (m, 1H, J ) 7.0, 7.0), 1.94 (m, 1H, J ) 20), 1.90 (m, 1H), 1.67
(m, 1H), 1.61 (m, 1H), 1.57 (s, 3H), 1.14 (d, 3H); 13C NMR (125.8
MHz) 148.18, 146.11, 129.15, 127.76, 126.70, 110.74, 40.47, 37.26,
1
44 °C from petroleum ether at -80 °C; H NMR (300 MHz) 7.25-
7.05 (m, 5H), 2.59 (m, 2H), 1.58 (m, 2H), 1.03 (s, 6H), 0.85 (s, 1H,
OH). A solution of carbinol (5.0 g) in 80 mL of toluene containing
0.4 g of p-toluenesulfonic acid is heated for 5 h under reflux with
(47) Johnson, J.; Schwartz, A.; Jacobs, T. L. J. Am. Chem. Soc. 1938,
60, 1882-1884.
(48) Grovenstein, E.; Lee, D. E. J. Am. Chem. Soc. 1953, 75, 2639-
2644. Straus, F. Ber. 1909, 42, 2866-2995.
(49) Schlubach, H. H.; Repenning, K. Liebig’s Ann. 1958, 614, 37-46.
(50) Brown, H. C.; Lane, C. F. Tetrahedron 1988, 44, 2763-2772.

10954 J. Am. Chem. Soc., Vol. 119, No. 45, 1997
Von E. Doering et al.
36.71, 22.93; IR 3100-2800, 1649, 1604, 1494, 886, 761, 700; UV
260 (2.31), 202 (3.94); GC-MS 174 (5) [M+], 159 (8) [M+ - CH3],
118 (100) [C9H10+], 105 (70) [C8H9+], 91 (25) [C7H7+].
saturated aqueous NH4Cl, decantation, and distillation to give a mixture
of 4-(2′,4′-dimethylphenyl)but-1-ene and starting chloride. The latter
is removed by refluxing over Mg in ether and hydrolyzing with saturated
aqueous NH4Cl. Distillation affords 4c: 5.0 g; bp 92-94 °C/10 mm;
IR 3090, 995, 910; UV 291.5 (3.20), 302.0 (2.95).
2-Methyl-5-phenyl-hex-2-ene (3b). Compound 3d (0.5 g) is added
to 10 mL of a 0.1 M solution of potassium tert-butoxide in HMPT and
stirred under argon in a Schlenk flask for 2 days at 50 °C. The deep
purple solution is quenched with saturated NH4Cl (150 mL) and
extracted with hexanes (4 50-mL portions). The extract is washed with
100 mL of brine and 100 mL of water, dried over anhydrous K2CO3,
filtered, and concentrated. Vacuum distillation affords 0.45 g (90%)
of an 8:92 mixture of 3d and 3b: 1H NMR (300 MHz) 7.25-7.00 (m,
5H), 5.14 (tm, 1H, J ) 7.6, 7.0), 2.64 (m, 1H, J ) 7.0, 7.0, 7.0), 2.30
(dd, 1H), 2.20 (dd, 1H), 1.59 (s, 3H), 1.45 (s, 3H), 1.18 (d, 3H); 13C
NMR (125.8 MHz) 148.21, 132.46, 129.03, 127.80, 126.63, 124.03,
41.24, 37.80, 26.21, 22.06, 18.23; IR 3100-2800, 1603, 1494, 759,
699; GC-MS 174 (8) [M+], 105 (100) [C8H9+], 91 (6) [C7H7+]; UV
260 (2.34), 202 (4.13) (corrected for the absorbance of 3d).
2-Phenyl-5-methylhex-1-ene (3c). (a) A 1-L flask fitted with an
efficient condenser and a stirring bar is charged with 94.4 g (0.8 mol)
of freshly distilled R-methylstyrene, 90 g (0.5 mol) of N-bromosuc-
cinimide (NBS), and 50 mL of freshly distilled CCl4. The reaction
mixture is heated to 130-140 °C; finally at 180 °C spontaneous,
vigorous boiling begins. The exothermic reaction is moderated by
intermittent cooling in an ice/water bath until all of the NBS is dissolved
and reaction ceased. Cooling to room temperature precipitates suc-
cinimide, which is removed by filtration. The mother liquor is distilled
in Vacuo (0.1 mm): first to remove solvent and excess R-methylstyrene
(25-70 °C) and then to give 68 g (69%) of a colorless oil consisting
of 3-bromo-2-phenylprop-1-ene and 1-bromo-2-phenylprop-1-ene in the
ratio 63:37 (GC).
(E)-1-(2′,4′-Dimethylphenyl)but-1-ene (4a(E)). Compound 4c (0.3
M) is isomerized on a preparative scale with NaOEt (0.5 M) in HMPT
at 90 °C for 1 h. Separation by preparative GC (column C) affords
(E)-1-(2′,4′-dimethylphenyl)but-1-ene (4a(E)) as the major component
(UV 291.5 (3.20), 302.0 (2.95); IR 1380, 1620, 960, 845) and (Z)-1-
(2′,4′-dimethylphenyl)but-1-ene (4a(Z)) as the first minor eluent (UV,
broad of weaker intensity than that of 4a(E); IR 1615, 1405, 830).
The second minor eluent is (E)-1-(2′,4′-dimethylphenyl)but-2-ene
(4b(E)):53 UV 276.6 (2.53); IR 1620, 1380, 960.
Series 5. (E)-1-Mesitylbut-1-ene (5a(E)). A stirred solution of
the reaction mixture from n-propylmagnesium bromide and mesityl-
aldehyde54 (1-mesitylbutan-1-ol, 25 g, 0.13 mol) in 100 mL of
anhydrous ether is treated with phosphorus tribromide (11.0 g, 0.04
mol) in a cooled (<0 °C) 500-mL, 3-necked flask bearing a reflux
condenser, mechanical stirrer, addition funnel, and drying tube. After
the addition, the solution is warmed to 25 °C, boiled under reflux for
2 h, and left standing for 12 h. Addition of ice and water and extraction
with ether (50 mL) gives an ethereal extract, which is washed
successively with 5% aqueous NaHCO3, 5% H2SO4, and ice water,
and dried (MgSO4). Distillation affords a slightly red material, which
is dissolved in ether, washed with dilute aqueous NaOH, dried, and
redistilled to give colorless 5a(E) (13.3 g (60%); bp 108-109 °C/17
mm; lit. bp 90-100 °C/12 mm;53 114 °C/21 mm;55 110 °C/20 mm56 ).
IR 970 cm-1, (lit. 96853). The olefinic coupling constant (J ) 16.5
Hz) confirms the trans structure.
(b) Following the general procedure, a Grignard reagent is prepared
from 8.4 g of isobutyl bromide and treated with the 63:37 mixture of
8.0 g of bromides above in 10 mL of THF. Removal of solvent in
Vacuo affords a mixture of 3c and 1-bromo-2-phenylprop-1-ene, which
is added in 5 mL of THF to a mixture of 5 g of Mg turnings in 25 mL
of THF and 0.3 mL of 1,2-dibromoethane. Stirring for 2 h, quenching
with aqueous NH4Cl, and extraction with hexanes provides R-meth-
ylstyrene and 2-phenyl-5-methylhex-1-ene (3c), distillation of which
in Vacuo affords 2.2 g of pure 3c as a colorless liquid: bp 65 °C/0.1
mm; 1H NMR (500 MHz) 7.33 (d, 2H), 7.15 (t, 2H), 7.08 (t, 1H), 5.28
(d, 1H, J ) 1.7), 5.04 (dd, 1H, J ) 1.4), 2.43 (dt, 2H, J ) 7.9), 1.46
(m, 1H, J ) 6.6), 1.27-1.37 (m, 2H), 0.81 (d, 6H); 13C NMR (125.8
MHz) 149.4, 141.94, 128.53, 127.51, 126.47, 112.06, 37.90, 33.68,
28.07, 22.65; IR 3100-2800, 1627, 1600, 1494, 893, 777, 703; UV
240 (3.97), 206 (4.19); GC-MS 174(8) [M+], 131 (15) [M+ - C3H7],
118 (100) [C9H10+], 91 (15) [C7H7+].
(E)- and (Z)-2-Methyl-5-phenylhex-3-ene (3e(E) and 3e(Z)). A
Wittig reaction (Vide supra) between 27 g of isobutyltriphenylphos-
phonium bromide51 in 150 mL of THF and 5.5 g of 2-phenylpropi-
onaldehyde (Aldrich Chemical Co.) in 25 mL of THF affords, after
distillation (bp 65 °C/0.1 mm), 6.4 g (90%) of 3e(E) and 3e(Z) in the
ratio 12:88: GC-MS (of the mixture) 174 (15) [M+], 131 (100) [M+
- C3H7], 118 (85) [C9H10+], 105 (90) [C8H9+], 91 (65) [C7H7+]. The
NMR spectra can be distinguished by virtue of the large difference in
relative intensities. 3e(E): 1H NMR (300 MHz) 7.30-7.00 (td, 5H),
5.57 (m, 1H, J ) 15.4), 5.41 (m, 1H), 3.33 (m, 1H, J ) 6.9), 2.19 (m,
1H, J ) 6.7), 1.28 (d, 3H), 0.93 (d, 6H). 3e(Z): 1H NMR (300 MHz)
7.25-7.00 (td, 5H), 5.40 (td, 1H, J ) 10.1), 5.17 (td, 1H,), 3.69 (dq,
1H, J ) 7.0, 9.5), 2.61 (dsept, 1H, J ) 6.6, 9.6), 1.25 (d, 3H), 0.94 (d,
3H), 0.84 (d, 3H).
4-Mesitylbut-1-ene (5c). An ethereal solution of chloromethyl-
mesitylene57 (5.0 g) is added dropwise to allyl magnesium bromide in
diethyl ether (from 4.0 g of allyl bromide). Stirred and boiled under
reflux for 6 h and washed with saturated aqueous NH4Cl, the reaction
mixture affords 4-mesitylbut-1-ene (2.25 g) after distillation, as shown
by the infrared and NMR spectra.
Series 6. (E)-2-(2′,4′-Dimethylphenyl)pent-3-ene (6b(E)). Drop-
wise addition of R,γ-dimethylallyl bromide58 (15.0 g) in ether to the
Grignard reagent from 2,4-dimethylbromobenzene (15.0 g) and mag-
nesium turnings (2.4 g, 0.10 g-atom), boiling under reflux for 3 h, and
the usual workup gives 6b(E) (38%; bp 101-102 °C/10 mm). All
GC columns tested indicate a single compound; NMR and IR at 970
cm-1 indicate the trans, unconjugated structure.
Series 7. (E)- and (Z)-2-(2′,6′-Dimethylphenyl)pent-3-ene (7b-
(E) and 7b(Z)). A Grignard reagent in THF from 2,6-dimethylbro-
mobenzene (10.0 g) is treated dropwise with R,γ-dimethylallyl bromide
(9.0 g). The resulting mixture is stirred and heated at 60 °C for 12 h,
cooled, and treated with saturated aqueous NH4Cl. The light yellow
THF layer is decanted, washed with dilute NaOH, dried over Mg2SO4,
and distilled to yield 3.6 g of 7b(E) and 7b(Z) (bp 77-79 °C/2.5 mm).
Series 8. (E)- and (Z)-2-Mesitylpent-3-ene (8b(E) and 8b(Z)).
Both stereoisomers have been prepared and characterized by Goering
et al.59 Following the procedure above, mesityl magnesium bromide
and R,γ-dimethylallyl bromide yield 1.0 g (0.0053 mol) of a colorless
liquid (bp 115-118 °C/5 mm). The major product is (E)-2-mesitylpent-
3-ene (8b(E)), about 10% of which is 8b(Z) and a small amount is
(E)-1-mesitylbut-2-ene.
(E)- and (Z)-2-Mesitylpent-2-ene (8a(E) and 8a(Z)). The mixture
of 8b(E) and 8b(Z) above is completely rearranged in KOC(CH3)3/
HMPT at 90 °C under kinetic control to 8a(E), the ratio to 8a(Z) being
Series 4.52 4-(2′,4′-Dimethylphenyl)but-1-ene (4c). To a solution
of allyl magnesium bromide from allyl bromide (30.5 g) in ether at 0
°C is added dropwise, 2,4-dimethylbenzyl chloride57 (38.8 g). After 6
h (slow warming to 25 °C), the reaction is worked up by shaking with
(54) Fuson, R. C.; Horning, E. C.; Rowland, S. P.; Ward, M. L. In
Organic Syntheses; Horning, E. C., Ed.; Wiley: New York, 1955, Collect.
Vol. III, pp 549-551.
(55) Klages, A. Chem Ber. 1902, 35, 2245-2262.
(56) Nightingale, D.; Radford, H. D. J. Org. Chem. 1949, 14, 1089-
1093.
(57) Vavon, G.; Bolle, J. C. R. Hebd. Seances Acad. Sci. 1937, 204,
1826.
(58) Hurd, C. D.; Ensor, E. H. J. Am. Chem. Soc. 1950, 72, 5135-5137.
(59) (a) Goering, H. L.; Kantner, S. S.; Chung, Chyi J. Org. Chem. 1983,
48, 715-721. (b) Goering, H. L.; Kantner, S. S. J. Org. Chem. 1984, 49,
422-426.
(51) Gamet, P. M.; Nagel, D. L.; Reilly, P. J.; Lawson, T.; Sharpe, J.;
Toth, B. J. Org. Chem. 1988, 53, 1064-1071.
(52) In series 4-8, IR spectra are recorded on a Perkin-Elmer Model
221 spectrophotometer as 10% solutions in CCl4 and are calibrated by the
1560-cm-1 band in air. NMR spectra are on a Varian A-60 spectrometer,
taken in CCl4 and recorded in ppm from TMS (δ-scale). Photocopies of all
spectra are available in Carleton.2
(53) Wehrli, R.; Heimgartner, H.; Schmid, H.; Hansen, H.-J. HelV. Chim.
Acta 1977, 60, 2034-2061.

ConjugatiVe Interaction in Styrenes
J. Am. Chem. Soc., Vol. 119, No. 45, 1997 10955
80.4 to 10.6, whereas after 120 h, equilibrium is established, the ratio
becoming 22:75 in favor of 8a(Z).
temperature for the various periods of time. Quenching is effected
prior to cooling with 1 mL of water. Thereafter, the same procedure
as above is followed.
Equilibrations. Purification of HMPT. In Series 1-3, the system
for effecting the equilibrations consists of hexamethylphosphoric
triamide (HMPT) as solvent and freshly sublimed potassium tert-
butoxide (KO-t-Bu), strictly free of tert-butyl alcohol, as strongly basic
catalyst.60 An effort to remove the water, a serious poison of the
catalyst, from the various commercially available samples of HMPT
by distillation from CaH2 proves inadequate. Stirring for 2 days at 80
°C over LiAlH4 or heating with a small amount of n-butyllithium (2.5
M in hexane solution) at 100 °C for 3 h, followed by distillation in
Vacuo, only a middle cut being taken, provides HMPT, which now is
anhydrous below the limits of detection. Another worry involves the
possibility of the highly reactive, proton-containing pentamethylphos-
phoric triamide (PMPT) being one of several impurities detectable by
GC and NMR. A variable peak in the range 2-4 ppm is removed
after extraction with D2O but reappears upon addition of traces of H2O.
Even after thorough drying and distillation, two peaks (δ 2.52 and 2.45)
remain. Although the HMPT is ∼99.7% of purity, three impurities
are revealed by GC (column DB-1): retention times 20.37 (HMPT,
21.13), 24.01, and 27.22 min. Purification by multiple crystallization
of the neat solvent removes all but one of the impurities (24.01 min).
After ten recrystallizations, HMPT of >99.9% of purity is obtained.
Coinjection of a synthetic sample of PMPT61 shows it is not an impurity.
In summary, recrystallization of meticulously dried HMPT and use of
sublimed KO-t-Bu lead to a catalytic system more active by a factor
of ∼10 than a catalyst similarly prepared from dry but not recrystallized
HMPT.
The upper temperature is limited in the present instance by the wish
to stay with a single preparation of one catalyst system and not to reduce
catalytic activity to achieve higher useful temperatures by the addition,
for example, of tert-butyl alcohol. In one variation, sealed, prescored
ampules (Wheaton, 1 mL, borosilicate glass) are stored in triethylamine,
dried >12 h at 150 °C, evacuated, baked with a heat-gun, cooled, and
flushed with argon. These are loaded with a standard solution, sealed
and heated at the desired temperature for the specified length of time.
This procedure is generally quite convenient, but positions of equilib-
rium are not reproducible at the highest temperatures because cooling
cannot be accomplished fast enough to preclude significant displacement
of equilibrium during the process. Kinetically more rapid intercon-
versions are particularly susceptible (e.g., 3a(E) and 3b(E)). In a second
procedure, equilibrations are conducted in rubber-sealed, Pyrex tubes
(350 × 4 mm) so that the catalyst can be quenched prior to cooling by
the injection of 1 mL of water.
Procedure. A standard 0.1 M solution is prepared by pumping
HMPT under argon pressure into a flask containing the requisite amount
of KO-t-Bu, dissolved by magnetic stirring and used at once. The
mixture for equilibration is prepared by introducing into a large ampule,
by means of gas-tight syringes, ∼100 µL of the starting olefin, ∼10
µL of bicyclohexyl as standard, and ∼2.5 mL of the 0.1 M catalyst
solution.
In the first procedure, aliquots are distributed among 5 ampules at
-78 °C, one being quenched immediately with 1 mL of water and
analyzed as described below to fix a standard for determining the
percent recovery in the other ampules after heating. The remaining
ampules are then sealed in Vacuo and placed in the appropriate constant-
temperature environment. Progress toward equilibrium is monitored
by analyzing ampules at appropriate intervals. Ampules are quickly
cooled by immersion into a vigorously stirred bath of isopropyl alcohol/
dry ice before opening, after which the contents are quenched by the
addition of 1 mL of water and extracted with 1 mL of hexanes. After
being washed thrice with 1-mL portions of water and passed though
filter paper, the solution is analyzed directly.
Analysis. Samples (0.8-1.2 µL) of the hexane solutions of the
olefins are injected into the gas chomatograph and their areas are
recorded. The mean of 3-4 injections is determined with a precision
of about 3% of the concentration of the individual component expressed
as percent of the sum. Response factors are obtained for GC-purified
samples of the pure olefins (>99%) relative to the most stable isomer
(f ) 1). The differences were small, the largest being 0.10 (3a(E) and
3d) and 0.031 (2a(Z) and 2b). Concentrations are corrected by the
response factors and reported as ratios of concentrations to the sum of
all the isomeric olefins, ln(ci/∑ci-j), in the Supporting Information as
Tables S1, S2, and S3.
Retention Times (in min) (Relative Response Factors). Phenyl-
butenes [column DB-1701; initial T, 60 °C; final T, 125 °C; rate, 3°/
min]: 4-Phenylbut-1-ene (1c) 11.67; 1-phenylbutane, 11.85; (E)-1-
phenylbut-2-ene (1b(E)), 12.66; (Z)-1-phenylbut-1-ene (1a(Z)), 13.07;
(Z)-1-phenylbut-2-ene (1b(Z)), 13.38; 1-methylindane, 13.50; (E)-1-
phenylbut-1-ene (1a(E)), 15.66; bicyclohexyl, 21.55; (1-methylind-1-
ene, 17.05; not present). Phenylpentenes [column DB-1; initial T, 80
°C; final T, 150 °C; rate 4°/min; final time, 40 min]: (Z)-1-phenyl-3-
methylbut-1-ene (2a(Z)), 14.06 (0.991); 1-phenyl-3-methylbutane,
14.69; 4-phenyl-2-methylbut-1-ene (2c), 15.04 (1.012); 1-phenyl-3-
methylbut-2-ene (2b), 16.14 (1.022); (E)-1-phenyl-3-methylbut-1-ene
(2a(E)), 16.56 (1.000, st); bicyclohexyl, 22.11. Phenylheptenes [column
DB-1; initial T, 80 °C; final T, 250 °C; rate 4°/min; final time, 30
min]: (Z)-2-phenyl-5-methylhex-3-ene, 19.02; (Z)-2-phenyl-5-meth-
ylhex-2-ene, (3a(Z)), 19.29 (0.969); (E)-2-phenyl-5-methylhex-3-ene,
19.72; 2-methyl-5-phenylhexane, 19.86; 2-methyl-5-phenylhex-1-ene
(3d), 20.68 (0.934); 2-methyl-5-phenylhex-2-ene (3b), 20.98 (1.000);
5-methyl-2-phenylhex-1-ene (3c), 21.28 (1.034); bicyclohexyl, 22.11;
(E)-5-methyl-2-phenylhex-2-ene (3a(E)), 23.15 (1.000, st).
Calculations. Linear regression analysis (Microsoft program EX-
CEL, version 4.0) of the function ln(ci/∑ci-j) ) -“∆H”/RT + “∆S”/
R, gives values of “∆H” directly from the slope and “∆S” from the
intercept. Differences are ∆∆H (in kcal/mol) and ∆∆S (in eu (cal mol-1
K-1)). This method of handling the data gives an unambiguous measure
of the standard error of each individual component. Corresponding
values for the differences in Gibbs free energy are derived by the usual
equation, from which equilibrium constants for any pair can also be
calculated.
Acknowledgment. This work has been supported by the
National Science Foundation, Grant No. CHE 94-11248. M.P.
acknowledges with gratitude support by the Norman Fund in
Organic Chemistry in memory of Ruth Alice Norman Weil
Halsband. P.S.C. acknowledges the gift of a Sheffield Fellow-
ship by Yale University and the assistance of a traineeship of
the National Aeronautics and Space Administration. Support
to the Department of Chemistry and Chemical Biology of
Harvard University comes from the NSF (CHE 90-20043) and
the NIH (S10-RR06716) for its Mass Spectrometry Facility and
from the NIH (S10-RR01748 and S10-RR04870) and the NSF
(CHE 88-14019) for its NMR Facility.
Supporting Information Available: Thermochemical data
on which Figure 1 is based, data relating to equilibrations at
various temperatures for series 1, 2, and 3 and their plots
(Figures 2-4), and Figure 5 showing chemical shifts of
hydrogens in styrenes and compounds 8a(E) and 8a(Z) (13
pages). See any current masthead page for ordering and Internet
access instructions.
When rubber-sealed tubes are used at ambient pressure, they are
first pretreated as described above. The standard solution is prepared
as above in one of the ampules and then distributed under argon in
equal portions into four tubes, a small amount being reserved for
quenching and analysis to serve as reference. Rubber septa are taped
to the tubes, and the reaction vessels are kept at the specified
(60) Doering, W. von E.; Urban, R. S. J. Am. Chem. Soc. 1956, 78,
5938-5942.
(61) Mihal, C. P. Am. Ind. Hyg. Assoc. J. 1987, 48, 997-1000.
JA971040E