Pressure independence of stereomutation and fragmentation in the bis-spirocyclobutanes, anti- and syn-2,9-dicyanodispiro[5.0.5.2]tetradeca-1,8- dienes

Article abstract of DOI:10.1021/ja053168z

Pressure dependent rate constants and volumes of activation for stereomutational interconversions of the cyclobutanes, anti- and syn-2,9-dicyanodispiro[5.0.5.2]tetradeca-1,8-dienes (anti-G and syn-6), and for their fragmentation to 1-cyano-3-methylenecycl

Full text of DOI:10.1021/ja053168z

Published on Web 12/03/2005
Pressure Independence of Stereomutation and Fragmentation
in the Bis-spirocyclobutanes, anti- and
syn-2,9-Dicyanodispiro[5.0.5.2]tetradeca-1,8-dienes
Frank-Gerrit Kla¨rner,*,† Frank Wurche,† William von E. Doering,*,‡ and Jiade Yang‡
Contribution from the Institut fu¨r Organische Chemie, UniVersita¨t Duisburg-Essen, D-45117
Essen, Germany, and the Department of Chemistry and Chemical Biology, HarVard UniVersity,
Cambridge, Massachusetts 02138-2902
Received May 14, 2005; E-mail: frank.klaerner@uni-essen.de; doering@chemistry.harvard.edu
Abstract: Pressure dependent rate constants and volumes of activation for stereomutational interconversions
of the cyclobutanes, anti- and syn-2,9-dicyanodispiro[5.0.5.2]tetradeca-1,8-dienes (anti-6 and syn-6), and
for their fragmentation to 1-cyano-3-methylenecyclohexene (5) have been determined. Although fragmenta-
tions might have been expected to have larger and thus more positive volumes of activation than
stereomutation, both processes have essentially identical volumes of activation within experimental
uncertainties at 50.1 °C over the pressure range, 1-3000 bar: ∆Vq ) (+7.4 to +9.9) ( 2.0 cm3 mol-1
.
While these positive values are consistent with the rate of entry into the hypothetical caldera being
determined by breaking the weakest bond in the cyclobutanes, the insensitivity of the product-determining
exit channels argues against the obligatory second bond-breaking being a significant factor in fragmentation.
Introduction
the dimerization of olefins and dienes (olefin-olefin unions),
its reverse, fragmentation of the related cyclobutanes (also
This work brings to a landing1 our investigation of pressure
dependence as a searchlight into the darkness of the caldera,
arguably the link between educt and product in not-obviously
concerted thermal reorganizations. Having passed through
various stages, distinction between concerted and not-obviously
concerted mechanisms on the basis of stereochemistry, energy
of concert, response to pressure, and negation of vestigial
influence of orbital symmetry on the fate of diradical intermedi-
ates, efforts continue to seek a measure of intellectual control
over the events that follow initial generation of energy-credible
diradicals and their passage from educt through caldera to
products. Questions are addressed to the relative importance of
the mode of entrance into the domain of diradicals (the caldera)
and of internal rotations on the determination of products. Do
diradicals continue on their way as diradicals-in-transit, or do
they pause in the caldera as diradicals-as-intermediates long
enough to establish, or approach, equilibria among product-
related conformations prior to exit?
dubbed cycloreversion or cleavage), which comprises a set of
exit channels not available to cyclopropanes; stereomutation,
the change from one stereoisomer to another; and ring enlarge-
ment, actualized in appropriate π-systems such as those contain-
ing proximate vinyl groups.
The effect of pressure is ideally studied in the distribution of
products from cyclobutanes where fragmentation competes with
stereomutation. If a larger increase in volume of reaction is
expected to be associated with fragmentation, a substantially
more positive volume of activation should be shown by
fragmentation than by stereomutation.4 Competition between
racemization of optically active 1,3,4,6-tetraphenylhexa-1,5-
diene (∆Vq ) -7.4 cm3 mol-1) and the mutual interconversion
of rac- and meso-1,3,4,6-tetraphenylhexa-1,5-diene (∆Vq
)
+13.5 and +11.5 cm3 mol-1, respectively) by way of the 1,3-
diphenylallyl radical may serve as an example.5
Studies of the effect of pressure on the ethene-ethene union
begin with the classical example of Stewart on the pressure
dependence of the products of the thermal dimerization of
chloroprene.6 A deeper probing through deuterium-labeling
follows in the work of Kla¨rner, Krawczyk, Ruster, and Deiters.7
Cyclobutanes have played a key role in these explorations,
not only because of a rich history in the field of thermal
reorganizations rivaling that of the cyclopropanes2,3 but also
because of the great disparities in the thermochemistry of their
various modes of reaction. These modes of reactions include
(4) (a) Kla¨rner, F.-G.; Wurche, F. J. Prakt. Chem. 2000, 342, 609-638. (b)
van Eldik, R., Kla¨rner, F.-G., Eds. High-Pressure Chemistry; Wiley-
VCH: Weinheim, 2002. (c) Matsumoto, K.; Hamana, H.; Iida, H. HelV.
Chim. Acta 2005, 88, 2033-2234.
† Universita¨t Duisburg-Essen.
‡ Harvard University.
(1) Arch., as at the end of a flight of stairs; Ger., Treppenabsatz.
(2) Northrop, B. H.; Houk, K. N. J. Org. Chem., published online Sept. 2,
2005 http://dx.doi.org/10.1021/jo051273L. Suhrada, C. P.; Selc¸uki, C.;
Nendel, M.; Cannizzaro, C.; Houk, K. N.; Rissing, P.-J.; Baumann, D.;
Hasselmann, D. Angew. Chem., Int. Ed. 2005, 44, 3548-3552.
(3) Lewis, D. K.; Charney, D. J.; Kalra, B. L.; Plate, A.-M.; Woodward, M.
H.; Ciancosi, S. J.; Baldwin, J. E. J. Phys. Chem. 1997, 101, 4097-4102.
(5) Doering, W. von E.; Birladeanu, L.; Sarma, K.; Blaschke, G.; Scheidem-
antel, U.; Boese, R.; Benet-Buchholz, J.; Kla¨rner, F.-G.; Gehrke, J.-S.;
Zimny, B. U.; Sustmann, R.; Korth, H.-G. J. Am. Chem. Soc. 2000, 122,
193-203.
(6) Stewart, C. A., Jr. J. Am. Chem. Soc. 1972, 94, 635-637.
(7) Kla¨rner, F.-G.; Krawczyk, B.; Ruster, V.; Deiters, U. K. J. Am. Chem.
Soc. 1994, 116, 7646-7657.
9
10.1021/ja053168z CCC: $30.25 © 2005 American Chemical Society
J. AM. CHEM. SOC. 2005, 127, 18107-18113
18107

A R T I C L E S
Kla¨rner et al.
Figure 2. The products of the thermal dimerization of cyclohexa-1,3-diene
at 70.5 °C are given with volumes of activation, ∆Vq, and steric energies,
SE, calculated by molecular mechanics (MM2).
be concerted by Woodward-Hoffmann rules,10 and likely
proceeding via a diradical intermediate. Particularly striking is
the apparent difference in mechanism involved in the generation
of the endo and exo Diels-Alder adducts.
A remarkable feature of both examples is the inability of
pressure to distinguish among products derived plausibly from
a single caldera. These include a pair of trans-1,2-cyclobutanes
and a vinylcyclohexene from the racemic diradical derived by
bond formation between the 1-positions of a pair of chloroprenes
(Figure 1) and the exo Diels-Alder adduct and meso-cyclobu-
tane plausibly derived from the meso diradical formed by the
union of two cyclohexadienes at the 1-positions (Figure 2).
Figure 1. Products of the thermal dimerization of (E)-1-deuterio-2-
chlorobutadiene at 50 °C are depicted with volumes of activation, ∆Vq,
and steric energies, SE, calculated by molecular mechanics (MM2).
Two of the products, 2- and 1-chloro-4-(1′-chlorovinyl)cyclo-
hexene, are constitutionally of the [4 + 2] Diels-Alder type
and have volumes of activation, ∆Vq ) -31 and -29 cm3
mol-1, respectively. A second set consisting of [2 + 2] cis-
and trans-1,2-dichloro-1,2-divinylcyclobutanes and, surprisingly,
a third product constitutionally of the Diels-Alder type, 1,4-
dichloro-4-vinylcyclohexene, all have significantly less negative
volumes of activation, ∆Vq ) -22 cm3 mol-1 (Figure 1).8 In a
compelling analysis, Kla¨rner concludes that the first set is
concerted in mechanism while the second involves diradicals
as intermediates, the former having a larger packing fraction
than the latter. These works exemplify the power of pressure
dependence to distinguish between a concerted set and a not-
obviously concerted, diradical set in situations where the two
are in competition with each other.
An example related to Stewart’s is the thermal dimerization
of cyclohexa-1,3-diene (Figure 2).9 Here again, two products,
an endo-bicycloo¨ctene constitutionally of the Diels-Alder type
and 5-(3′-cyclohexenyl)cyclohexa-1,3-diene, the result of a [6
+ 4] hydrogen transfer, are associated with more negative
volumes of activation, -28 and -32 cm3 mol-1, respectively,
while the exo-bicycloo¨ctene, and two [2 + 2] cyclobutane
dimers have more positive volumes of activation, ∆Vq ) -22,
-18, and -22 cm3 mol-1, respectively. The cyclobutanes serve
as anchors for not-obviously concerted reactions, forbidden to
The present work completes a study of cyclobutanes designed
to sharpen the focus on the effect of pressure on the ratio of
products reached through two types of exit channels, stereo-
mutation and fragmentation. The original choice of bis-
spirocyclobutanes was driven by a desire to keep the system
free of a third reaction mode of ring enlargement open to
unrestrained vinylcyclobutanes and yet satisfy the need for the
temperature of reaction to fall into a range experimentally
convenient for the evaluation of volumes of activation.
The cyclobutane dimers of 3-methylenecyclohexene, syn- and
anti-2, and of 1-phenyl-3-methylenecyclohexene, syn-4 and
anti-4 (Figures 3 and 4, respectively), were the objects of the
first study.11 A modestly positive volume of activation (∆Vq >
0) was expected in the opening of the cyclobutane ring to a
diradical as the first step.12 Pressure might be expected to
influence the subsequent reactions of the intermediate as well,
fragmentation (to 3-methylenecyclohexene 1 or the 1-phenyl
derivative 3, respectively) and ring closure (detectable as
stereomutation between syn-2 and anti-2 and syn-4 and anti-4,
(8) As a first point, recall that the paradigm of the Diels-Alder reaction, the
cycloaddition of ethylene to butadiene, is but weakly concerted (∼7 kcal
mol-1). As a second, note that the two cyclohexenes of the first, highly
likely concerted, constitutionally Diels-Alder set would have involved the
less well stabilized diradicals had they been involved, whereas the not
obviously concerted constitutionally Diels-Alder cyclohexene of the second
set is derived, as are the two cyclobutanes, from the optimally stabilized
racemic and meso diradicals.
(10) Hoffmann, R.; Woodward, R. B. J. Am. Chem. Soc. 1965, 87, 2046-2048.
See also: Hoffmann, R. Angew. Chem., Int. Ed. 2004, 43, 6586-6590 and
references therein.
(11) Doering, W. von E.; Ekmanis, J. C.; Belfield, K. D.; Kla¨rner, F.-G.;
Krawczyk, B. J. Am. Chem. Soc. 2001, 123, 5532-5541.
(12) Diedrich, M. K.; Kla¨rner, F.-G. J. Am. Chem. Soc. 1998, 120, 6212-6218.
(9) Kla¨rner, F.-G.; Dogan, B. M. J.; Ermer, O.; Doering, W. von E.; Cohen,
M. P. Angew. Chem., Int. Ed. Engl. 1986, 25, 108-110.
9
18108 J. AM. CHEM. SOC. VOL. 127, NO. 51, 2005

Stereomutation/Fragmentation in Bis-spirocyclobutanes
A R T I C L E S
Table 1. Pressure Dependence of the Rate Constants and
Volumes of Activation for the Stereomutation, and Fragmentation
to 5, of anti-6 and syn-6 at 50.1 °C in n-Heptane/tert-Butylmethyl
Ether (9:1)
anti-6
f
syn-6
syn-6
f
anti-6
anti-6
f
a
5
syn-6
f
a
5
a
a
p [bar]
kas
ksa
kam
ksm
1
500
11.1 ( 0.06
8.94 ( 0.45
7.65 ( 0.46
7.37 ( 0.40
6.50 ( 0.22
4.20 ( 0.27
4.87 ( 0.25
25.2 ( 1.2
18.4 ( 0.8
16.4 ( 0.8
15.9 ( 0.8
13.0 ( 0.4
8.6 ( 0.4
9.5 ( 0.4
3.78 ( 0.17
2.38 ( 0.14
2.58 ( 0.14
2.30 ( 0.12
1.83 ( 0.11
1.60 ( 0.11
1.34 ( 0.10
2.86 ( 0.33
2.53 ( 0.25
1.84 ( 0.24
1.50 ( 0.20
1.34 ( 0.13
1.08 ( 0.16
0.76 ( 0.14
1000
1500
2000
2500
3000
∆Vqb
∆Vq
8.0 ( 3.0
9.1 ( 1.7
∆Vq
9.0 ( 3.0
9.9 ( 1.7
∆Vq
8.2 ( 3.2
7.4 ( 1.8
∆Vq
11.5 ( 1.8
9.7 ( 0.9
as
sa
am
sm
∆Vq c
l
∆Vq
d
q
Figure 3. Thermal rearrangements of the [2 + 2] dimerization products
of 3-methylenecyclohexene 1, anti-2 and syn-2, are shown, along with their
volumes of activation, ∆Vq, specific rate constants, and differences in
volumes of activation, ∆∆Vq, relative to the conversion of syn-2 to anti-2
taken as 0.0.
a Rate constants in units of 10-5 s-1 b Volumes of activation in cm3
.
mol-1 c Derived from kinetic data by the linear correlation, ln(k) p ) a +
.
bp; ∆Vlq ) -bRT. d Derived from the kinetic data by the nonlinear
q
correlation, ln(k) p ) a + bp + cp2; ∆Vq ) -bRT.
convenient separation and purification of syn-4 and anti-4
necessary for both to serve as educts.
The importance of placing this apparent lack of pressure
dependence of products emerging from a single caldera on as
firm a footing as possible prompted a search for a pair of
cyclobutanes that would react at temperatures between the
unsubstituted and the too reactive phenyl-substituted. The cyano
group turned out to be a sufficiently less radical-stabilizing group
than phenyl to be close to “just right”! The derived anti- and
syn-2,9-dicyanodispiro[5.0.5.2]tetradeca-1,8-dienes, anti-6 and
syn-6, were stable enough to allow separation by high-pressure
liquid chromatography, thus making possible a study of both
isomers at the convenient temperature of 50.1 °C. Furthermore,
we were interested in the question whether additional radical-
stabilizing substituents such as the cyano group might lead to
the formation of a more stable diradical that would show the
reactive behavior expected of a conventional intermediate,
namely a pressure-dependent product ratio.
The preparation of the precursor, 1-cyano-3-methylenecy-
clohexene (5), followed the procedure of Cronyn and Goodrich
but required modification to become reliably reproducible.14 Its
photodimerization led to a mixture of anti-6 and syn-6 which
could be separated by medium-pressure liquid chromatography
(MPLC). The specific rate constants of the reversible stereo-
mutation between anti-6 and syn-6 and fragmentation of anti-6
to 5 and of syn-6 to 5 were calculated by numerical integration
from two sets of product ratios determined at 50.1 °C and
different pressures starting from either enriched anti-6 or syn-
6. Further details of the kinetic analyses are given in the
Experimental Section. Kinetic data are presented in the con-
ventional way (Table 1) and also as ratios of rate constants for
fragmentation relative to the rate of conversion of syn-dimer to
anti-dimer (ksa taken as 1.000) (Table 2). Estimates of relative
rate constants at 1 bar and 50.1 °C were obtained by linear
regression of the ratios in Table 2. The equilibrium constant, K
Figure 4. Thermal rearrangements of the [2 + 2] dimerization products
of 1-phenyl-3-methylenecyclohexene 3, anti-4 and syn-4, are shown, along
with their volumes of activation, ∆Vq, specific rate constants, and differences
in volumes of activation, ∆∆Vq, relative to the conversion of syn-4 to anti-4
taken as 0.0.
respectively). Presumably, the volume of activation for frag-
mentation should be significantly more positive than that for
reclosure to the cyclobutanes.13
In fact, the thermal reactions of syn-2 and anti-2 (Figure 3)
and syn-4 and anti-4 (Figure 4) showed a pressure-induced
deceleration of the overall reaction (∆Vq > 0) as expected. But
neither showed any significant decrease in the ratio of the
products of fragmentation and stereomutation. Perhaps, these
examples had not been optimal for revealing small differences.
A convenient temperature for the study of 1 and its dimers was
on the high side, while that for the study of 3 and its dimers
was too low both for kinetic studies of high precision and for
) ksyn-6/kanti-6 ) 2.21, translates to ∆∆G50° ) -0.5 kcal mol-1
.
Steric energies calculated by MM2 for syn-6 and anti-6, +41.6
and +40.0 kcal mol-1, respectively, are in the same direction.
There is nothing surprising about the thermochemistry of the
system.
(13) In the case of the [2+2] cycloaddition of 1,1-difluoroallene, there is good
evidence that the ratio of products derived from a diradical intermediate
can indeed be pressure-dependent. At high pressure the product resulting
from ring closure in the diradical intermediate is favored over that resulting
from bond rotation followed by ring closure. Dolbier, W. R., Jr.; Weaver,
S. L. J. Org. Chem. 1990, 55, 711-715.
(14) Cronyn, M. W.; Goodrich, J. E. J. Am. Chem. Soc. 1952, 74, 3331-3333.
9
J. AM. CHEM. SOC. VOL. 127, NO. 51, 2005 18109

A R T I C L E S
Kla¨rner et al.
Table 2. Pressure Dependence of the Rate Constants of syn-6
and anti-6 Relative to the Conversion, syn-6 f anti-6 (ksa Set to
1.000), and the Differences in Volumes of Activation at 50.1 °C in
n-Heptane/tert-Butylmethyl Ether
p [bar]
syn-6
f
anti-6
anti-6
f
syn-6
anti-6
f
5
syn-6 f 5
1
500
1.000
1.000
1.000
1.000
1.000
1.000
1.000
0.444
0.485
0.466
0.463
0.501
0.490
0.511
0.150
0.129
0.157
0.157
0.141
0.187
0.141
0.114
0.138
0.112
0.095
0.103
0.127
0.079
1000
1500
2000
2500
3000
a
a
a
a
∆∆Vq
0.0
∆∆Vq
∆∆Vq
∆∆Vq
sa
as
am
sm
-1.0 ( 0.3
-0.8 ( 1.2
+2.5 ( 1.7
a Differences in volumes of activation in cm3 mol-1 derived from the
ratios of rate constants by linear correlation, e.g., ln(kas/ksa) p ) a + bp;
∆∆Vq ) -bRT. For stereomutation ∆∆Vq ) ∆Vq - ∆Vqsa, and for
as
as
as
fragmentations ∆∆Vqam ) ∆Vqam - ∆Vqsa and ∆∆Vqsm ) ∆Vqsm - ∆Vq
.
sa
Discussion
Much as expected, the radical-stabilizing substituents, cyano
and phenyl, decrease the Gibbs free energy activation barriers
relative to the thermal reactions of the syn and anti dimers of
the unsubstituted 3-methylenecyclohexene:
CN (50.1 °C): ∆∆Gqsyn ) ∆Gqsyn-2 - ∆Gq
)
syn-6
29.32 - 24.19 ) 5.13 kcal mol-1
∆∆Gqanti ) ∆Gqanti-2 - ∆Gq
)
anti-6
29.62 - 24.64 ) 4.98 kcal mol-1
Ph (43.6 °C): ∆∆Gqsyn ) ∆Gqsyn-2 - ∆Gq
)
syn-4
29.32 - 23.44 ) 5.88 kcal mol-1
∆∆Gqanti ) ∆Gqanti-2 - ∆Gq
)
anti-4
29.62 - 23.80 ) 5.82 kcal mol-1
Although the stabilizing effect of the cyano group is only slightly
less than that of phenyl, the small difference has been just
helpful enough to allow completion of an experimentally
satisfying investigation of the effect of pressure.
Figure 5. Thermal rearrangements of the [2 + 2] dimerization products
of 1-cyano-3-methylenecyclohexene 5 (anti-6 and syn-6) are shown, along
with their volumes of activation, ∆Vq, specific rate constants, and differences
in volumes of activation, ∆∆Vq, relative to the conversion of syn-6 to anti-6
taken as 0.0. A hypothetical diradical serving as an intermediate is also
shown.
Ratios of stereomutation to fragmentation in the unsubsti-
tuted parent system, anti-2 and syn-2, at 72.1 °C are kas/kam
)
0.49/0.70 ) 0.70 and ksa/ksm ) 0.70/0.93 ) 0.75 at 72.1 °C,
respectively, while the values from the phenyl analogues, anti-4
and syn-4, at 43.6 °C are kas/kam ) 5.67 and ksa/ksm ) 5.79,
respectively. Those from the cyano analogues, anti-6 and
syn-6, at 50.1 °C are kas/kam ) 2.94, and ksa/ksm ) 8.81,
respectively. This enhancement of stereomutation over frag-
mentation (or this reduction of fragmentation relative to
stereomutation!) by increasing stabilization of the diradical has
already been noted when the activating substituents are basically
butadienyl (stereomutation/fragmentation ∼25 at 43.6 °C)15 and
hexatrienyl (stereomutation/fragmentation ∼4000 at -25.1
°C).16
system 6 could be investigated starting from both stereoisomers.
The ratio of stereomutation to fragmentation in the rearrange-
ment of syn-6 is greater by a factor of 3.00 (∆∆Gq ) 0.70 kcal
mol-1) than in the rearrangement of anti-6. On the assumption
that anti-6 has a lower enthalpy of formation than syn-6, it is
tentatively proposed that a steric driving force for anti-6 to
undergo the necessary internal rotation prior to ring closure may
be less than that for syn-6.
The most significant result of the present work is a more
rigorous confirmation of the insensitivity of the ratio of
stereomutation to fragmentation to increasing pressure. The
anticipated repression of fragmentation vis-a`-vis stereomutation
by increasing pressure has not been observed. Presenting the
available data as ratios emphasizes the differences in volumes
of activation. The absence of any significant result favoring one
way or the other seen in Figures 3, 4, and 5 emerges with
striking clarity. The quite disparate geometry of the two exit
channels has no effect.
Quantitative comparisons of the relative ratios of stereomu-
tation to fragmentation in the syn isomers are relatively
unreliable because there is but a single observation starting from
syn-2 and none starting from syn-4. By contrast, the cyano
(15) Doering, W. von E.; Belfield, K. D.; He, J.-n. J. Am. Chem. Soc. 1993,
115, 5414-5421.
(16) Doering, W. von E.; He, J.-n.; Shao, L.-m. J. Am. Chem. Soc. 2001, 123,
9153-9161.
9
18110 J. AM. CHEM. SOC. VOL. 127, NO. 51, 2005

Stereomutation/Fragmentation in Bis-spirocyclobutanes
A R T I C L E S
All the examples can be accommodated by a conceptual
scheme of diradicals in transit. In this scheme each educt opens
to a diradical consisting of an idiosyncratic distribution of
conformations each individual among which is predestined to
proceed to a unique product by its point of entry into the
multidimensional potential energy surface representing the
diradical (shades of a pin-ball machine!). An elegant, lucid
account of the application is found in the analysis of vinylcy-
clopropane given by Doubleday, Li, and Hase.17 The goal of
trajectory analysis is the calculation of the trajectories followed
to exit by a sufficiently large number of entry states to provide
a statistically acceptable approximation of the experimental
distribution of products apposite to the system. No simple
(qualitative or pseudo-quantitative) ways of thinking about these
pathways are expected to emerge, even though distributions
among multiple products will have been predicted correctly.
Little has been revealed about the effect of temperature on
the distribution among products emerging from multichanneled
caldera. The only study in this series of cyclobutanes relates to
the parent 2. But these data are not of a precision sufficient to
allow significant conclusions about the effect of temperature
on product ratio to be drawn.18 In the cyclopropane series,
1-cyano-2(E)-propenylcyclopropane at 207.1 and 387.4 °C
reveals no significant effect of temperature on the ratios of cis-
and trans-3-methyl-4-cyanocyclopentene, the products of ring
enlargement.19 Neither does any significant effect of temperature
emerge in the studies by Baldwin and Keliher of bicyclo[3.1.0]-
hexene.20 Further kinetic studies of sufficiently high precision
are desirable in this and other systems believed to involve
multiple exit channels from the caldera. But there is of course
a contradiction between any existing small differences in
enthalpy of activation and the experimental complications
involved in measuring temperature dependencies over neces-
sarily very large ranges of temperature.
In these types of reaction, the overall energy of activation is
consumed in the process of bond-breaking to the diradicals. The
caldera may be generated essentially at thermal equilibrium with
no significant excess of energy. If barriers to exit are of a
magnitude comparable to 3/2RT, no further activation by
intermolecular collision is required and the Arrhenius equation
is inapplicable.21
Among the conventional ways of thinking about product
distribution, three are not helpful for explaining or predicting
product ratios for conversions possibly involving diradical
structures traversing a caldera: neither energies nor volumes
of activation nor application of orbital symmetry theory are
fruitful guides for anticipating or understanding product ratios.
Perhaps the irrelevance of these three approaches may serve as
a useful criterion that a reaction leading to two or more products
may be proceeding by way of a caldera. Nonetheless, good
Figure 6. The fraction in % of each of the three components, anti (a), syn
(s), and monomer (m), emanating from a purely hypothetical intermediate
in common is calculated for the reactions in Figures 3, 5, and 4, from three
equations: a/m ) kas/kam; s/m ) ksa/ksm, and a + s + m ) 1.
Were the reactions under study here to consist of two
competing exit channels proceeding from an intermediate in
common, the behavior of the caldera could then (and always)
be depicted in a graphical form (the pie representation in Figure
6), provided the unobserVable return reactions were given the
same relative values they had in the observable reactions. This
virtual formulation is not applicable to systems comprising more
than two exit channels. To give it visibility in this way does
not imply support for a common intermediate as the full or even
a partial description of the behavior within the caldera. What is
seen is the remarkable decline in the proportion of fragmentation
as the overall enthalpy of activation (stabilization in the
diradical) decreases.16
In the examples of this and the preceding paper11 as well all
the other examples we have studied of not-obviously concerted
rearrangements leading to multiple products but having es-
sentially indistinguishable enthalpies of activation, none has been
successfully accommodated by a single diradical intermediate
in common that reaches equilibrium regardless of origin.
“Reaches equilibrium” implies intramolecular vibrational re-
laxation being faster than sudden death by reclosure.
(17) Doubleday, C.; Li, G.-s.; Hase, W. L. Phys. Chem. Chem. Phys. 2002, 4,
304-312.
(18) From the temperature-dependent data in Table 1 of ref 11, including the
data in Table 2 at 100.5 °C, the activation energies for the product ratios,
kas/kam and ksa/ksm, are calculated to be 1.5 ( 0.8 and -3.1 ( 6.2 kcal
mol-1, respectively (standard errors!). Recalculation affords somewhat
altered values for the energies of activation that emphasize the poorer quality
of the data relating to syn-2: anti-2 f syn-2, 31.4 ( 1.1; syn-2 f anti-2,
34.2 ( 8.2; anti-2 f 1 (methylenecyclohexene), 33.0 ( 1.9; syn-2 f 1,
30.3 ( 6.4 (95% confidence level).
(19) Doering, W. v. E.; Barsa, E. A. J. Am. Chem. Soc. 2004, 126, 12353-
12362.
(20) Baldwin, J. E.; Keliher, E. J. J. Am. Chem. Soc. 2002, 124, 380-381.
(21) Menzinger, M.; Wolfgang, R. Angew. Chem., Int. Ed. Engl. 1969, 8, 438-
444.
9
J. AM. CHEM. SOC. VOL. 127, NO. 51, 2005 18111

A R T I C L E S
Kla¨rner et al.
Table 3. Pressure Dependence of the Relative Concentrations of
anti-6, syn-6, and 5 at 2000 bar and 50.1 °C in %
control of the effect of structural perturbations on the overall
enthalpy of reaction remains.
t (s)
anti-6
syn-6
5
Σa
t (s)
anti-6
syn-6
5
Σa
Experimental Section
0
75.26 19.22 5.52 100.0
0 19.83 78.40 1.76 100.0
2460 67.50 23.28 9.22 92.8 2040 28.60 68.28 3.12 100.8
5580 61.29 25.45 13.26 94.2 4080 35.62 59.38 5.00 101.1
8160 57.05 26.05 16.90 94.0 6180 40.47 52.97 6.56 96.9
10440 53.87 25.79 20.34 93.7 8100 46.51 44.25 9.24 97.9
12840 50.76 25.39 23.85 94.1 10080 49.76 38.92 11.32 95.8
15360 48.72 24.33 26.95 93.3 12540 50.85 35.95 13.20 95.8
18060 46.38 23.79 29.82 94.3 14580 51.66 31.67 16.67 91.5
20340 44.12 22.54 33.34 91.9 16620 51.90 29.72 18.38 93.3
22740 42.03 22.29 35.68 91.4 18660 51.73 28.41 19.86 95.7
25500 41.41 20.86 37.72 93.5 20700 49.48 29.42 21.10 96.8
28500 38.98 19.63 41.39 92.4 22740 49.76 27.41 22.84 97.9
31500 37.05 18.81 44.14 93.9 24840 50.29 26.23 23.48 100.7
33660 35.31 18.32 46.36 93.4 26640 48.00 26.32 25.68 100.3
35880 33.39 17.28 49.33 90.9 28620 45.80 25.66 28.54 96.8
1
General. H NMR spectra were taken on a Bruker DRX 500 mHz
instrument. High-pressure equipment consisted of a Dieckers, Willich
4-kbar instrument. HPLC apparatus was by Jasco, Gross-Umstadt
(column: Nucleosil 100-5 NO2 Vario Prep ET 250/10 column
(Marcherey-Nagel), n-heptane/tert-butylmethyl ether (90-10); flow of
1.0 mL/min; UV-detector: 0-10 min at 254 nm, 10-30 min at 225
nm). MPLC: Kronlab, Sinsheim (column: YMC-Silica gel 120 Å,
operated as above).
1-Cyano-3-methylenecyclohexene (5). A solution of 19.6 g (0.10
mol) of sodium metabisulfite in 32 mL of water was added dropwise
at 25 °C over a 15-min period to 31.2 g (0.20 mol) of 1,4-dioxa-spiro-
[4,5]decan-7-one in a 250-mL, round-bottomed flask accompanied by
vigorous stirring. After being stirred for 2 h more, the homogeneous
solution was added to 10.6 g of sodium cyanide (0.21 mol) in 22 mL
of water. After being stirred for 20 h at 25 °C, the solution was diluted
with 100 mL of ethyl acetate, stirred for 10 min, and filtered. The
organic layer was separated from the water layer, which was extracted
twice each with 20 mL of ethyl acetate. The combined organic layers
were dried over MgSO4 and concentrated in a vacuum to a residue
which was treated with 30 mL of 3 N HCl. The resulting homogeneous
solution was allowed to stand for 30 h, the initially homogeneous
solution having separated into two layers. An extract of the aqueous
layer with ethyl acetate (3 × 50 mL) was combined with the separated
organic layer, washed with aqueous NaHCO3 (2 × 75 mL), dried over
MgSO4, and concentrated to a yellow oil. Flash chromatography
(petroleum ether/ethyl acetate: 3:1) afforded 17.6 g of 3-cyanocyclohex-
2-enone: bp 83 °C at 2.2 mmHg (reported bp 105 °C at 3.8 mmHg);
1H NMR (400 mHz, CDCl3) 6.51 (s, 1H), 2.58-2.56 (m, 2H), 2.52-
2.49 (m, 2H), 2.16-2.10 (m, 2H).
To a suspension of the Wittig reagent prepared from methyltriph-
enylphosphonium bromide (8.0 g) in 60 mL of dry THF and 10.5 mL
of 2.0 M butyllithium, a solution of 2.46 g of the ketone above in 10
mL of THF at 0-5 °C was added dropwise over a period of 5 min.
After being stirred for 1 h at room temperature, the solution was
quenched with 0.5 mL of methanol. The resulting suspension was
decanted slowly into 200 mL of pentane containing 15 g of Celite,
stirred for 15 min, filtered, and concentrated by distillation through a
35-cm, helix-packed column. Passage through a short silica gel column
(petroleum ether/ether: 6:1), concentration by fractional distillation,
and crystallization overnight at -84 °C yielded a crystalline product
(1.52 g), which could be separated from the supernatant liquid at low
temperature. By 25 °C, the product had melted: 1H NMR (400 mHz,
CDCl3) 6.83 (s, 1H), 5.13 (s, 2H), 2.40-2.32 (m, 2H), 2.32-2.28 (m,
2H), 1.80-1.72 (m, 2H); Exact mass (electron spray) calcd for C8H9N-
(NH4+), 137.1079; found, 137.1079.
a Σ represents recovery vis-a`-vis mesitylene and p-dicyanobenzene as
internal standards.
by MPLC and stored in solution at -30 °C for use in the kinetic studies
below. Attempts to remove solvent under reduced pressure at -20 °C
for the purpose of obtaining NMR spectra on the pure dimers were
unsuccessful owing to their still high reactivity. A spectrum of a mixture
obtained by irradiation of 5 in toluene-d8 follows: 1H NMR (toluene-
d8) 6.27 (m, 5, 1H), 6.21 (m, syn-6, 1H), 6.14 (m, anti-6, 1H), 1.64-
1.59 (m), 1.44-1.35 (m), 1.30-1.21 (m), 1.15-1.08 (m), 1.02-0.95
(m), 0.93-0.87 (m). UV-vis: anti-6, λmax ) 233 nm; syn-6, λmax
)
223 nm.
High-Pressure Kinetics. To portions of the standard solutions of
anti-6 and syn-6 prepared by irradiation and separation above by MPLC,
mesitylene (7) and p-dicyanobenzene (8) as internal standards and
catalytic amounts of bis-(3-tert-butyl-4-hydroxy-5-methylphenyl)-sulfide
(BHMPS, Aldrich, No. 12,462-5) as radical scavenger were added. At
each of the seven pressures indicated in Table 1, all at 50.1 °C, and at
fifteen different time intervals, starting from each of two isomers,
products are analyzed by HPLC with the Nucleosil column at a flow
rate of 1 mL per min. Retention times in min are 7, 2.74; 5, 3.98; 8,
12.66; anti-6, 14.67; and syn-6, 18.87. A single example at 2000 bar is
given in Tables 3.
Specific rate constants are calculated from the two sets of product
ratios by numerical integration using the program KINETIK of Dr. R.
Fink which employs a Runge-Kutta procedure of fourth order and
estimation of experimental uncertainties at the 95% confidence level
by the method of Marquardt.22 The program permits optimization of
kinetic schemes with as many as seven components.
Concentrations of 5, anti-6, and syn-6 were calculated from the
HPLC chromatograms using the following equations: c(anti-6)t )
[fanti-6 × Aanti-6,t × c(8)0]/A8,t, c(syn-6)t ) [fsyn-6 × Asyn-6,t × c(8)0]/
A8,t, and c(5)t ) [f5 × F5,t × c(7)0]/A7,t, where fanti-6, fsyn-6, and f5 are
HPLC UV factors for anti-6, syn-6, and 5, respectively, and Acpd,t are
HPLC peak areas of the various compounds. To determine the HPLC
UV factors of anti-6 and syn-6, a solution of 61.3 mg of 5 and 9.9 mg
of 8 in 0.7 mL of acetone-d8 was prepared, degassed, and irradiated
for 1 h at 0 °C. After irradiation, the ratios among anti-6, syn-6, and 8
Photodimerization of 3-Cyano-1-methylenecyclohex-2-ene (5). A
solution of 243 mg of 3-cyano-1-methylenecyclohex-2-ene and 30 mg
of benzophenone in a mixture of 3 mL of n-heptane and 3 mL of
acetone was partitioned among three ampules (5 mm × 150 mm). The
ampules were freed of oxygen by three pump-freeze cycles, sealed under
argon, and placed in a Pyrex irradiation apparatus that could be cooled
by circulating ethanol at 0 °C. Irradiation with a high pressure mercury
lamp (150 W) for 45 min generated only a small amount of insoluble
polymer, which was removed by filtration through a short column of
Al2O3. Analysis of the resulting mixture by HPLC on a Nucleosil 100-5
column with n-heptane/tert-butylmethyl ether at 254 nm showed only
a single main product (tR ) 4.21 min) after 16.7 min of irradiation.
Upon changing the UV-vis detector to 220 nm, a second photodimer
became observable. In the routine procedure, 254 nm was used for the
1
were determined by H NMR and used in the HPLC analysis for the
determination of the UV factors of anti-6 and syn-6. For 5, a solution
of 19.95 mg of 5 and 117.5 mg of 7 in 100 mL of n-heptane/tert-
butylmethyl ether (9:1) was prepared and analyzed three times by
HPLC, the UV factors being evaluated by means of the following
equations: fanti-6 ) (F8/Fanti-6) × (Ianti-6/I8) ) 0.2388; fsyn-6 ) (F8/
F
syn-6) × (Isyn-6/I8) ) 0.2398; and f5 ) (F7/F5) × (n1/n7) ) 0.1302,
where Icpd equals the integrated value of the NMR signal, and n, the
molar amount of 5.
first 10 min of analysis, followed by a shift to 225 nm: anti-6 (tR
)
15.38 min) and syn-6 (tR ) 19.78 min). Samples could be separated
(22) Marquardt, D. W. J. Soc. Ind. Appl. Math. 1963, 11, 431-441.
9
18112 J. AM. CHEM. SOC. VOL. 127, NO. 51, 2005

Stereomutation/Fragmentation in Bis-spirocyclobutanes
A R T I C L E S
Acknowledgment. W.v.E.D. and J.Y. thank the Norman Fund
in Organic Chemistry in memory of Ruth Alice Norman Weil
Halsband and Edward A. Norman for generous support of this
work. F.-G.K. expresses his thanks to the Deutsche Forschungs-
gemeinschaft, the Ministry for Science and Research of the Land
Nordrhein-Westfalen, and the Fond der Chemischen Industrie
for support.
JA053168Z
9
J. AM. CHEM. SOC. VOL. 127, NO. 51, 2005 18113