Remote substituent effects upon the rearrangements of housane cation radicals

Article abstract of DOI:10.1021/jo050087c

The rearrangement of a series of housane-derived cation radicals was investigated. Surprisingly, 2-aryl-substituted systems rearranged regioselectively and in a process whose selectivity proved to be independent of the electronic character of para substit

Full text of DOI:10.1021/jo050087c

Remote Substituent Effects upon the Rearrangements of Housane
Cation Radicals
†
‡
†
†
James B. Gerken, Selina C. Wang, Alejandro B. Preciado, Young Sam Park,
†
,‡
,†
Gisele Nishiguchi, Dean J. Tantillo,* and R. Daniel Little*
Department of Chemistry & Biochemistry, University of CaliforniasSanta Barbara, Santa Barbara,
California 93106, and Department of Chemistry, University of CaliforniasDavis, Davis, California 95616
little@chem.ucsb.edu; tantillo@chem.ucdavis.edu
Received January 14, 2005
The rearrangement of a series of housane-derived cation radicals was investigated. Surprisingly,
2
-aryl-substituted systems rearranged regioselectively and in a process whose selectivity proved
to be independent of the electronic character of para substituents. The major reaction pathway is
suggested to be the one that allows maximum delocalization, and allows it to be maintained for as
long as possible. Bridging is invoked to account for the regio- and stereoselectivity. When a
2
nonbridging trimethylsilylmethyl substituent is appended to C , the regioselectivity is eroded
entirely. B3LYP/6-31G(d) calculations corroborate the notion that bridging plays a role. While
bridging ought to stabilize an intermediate by allowing delocalization of the charge/spin, there
should be an accompanying entropic penalty. To determine the relative importance of enthalpic
and entropic factors in determining the product selectivity, the rearrangement of the p-
methoxyphenyl-substituted housane was investigated as a function of temperature. Enthalpic factors
dominated over the entire temperature range that was explored. Overall, the results indicate that
it is possible to influence the direction of migration in housane-derived cation radical rearrangements
even when the regiochemical control unit is not directly appended to the migration terminus. This
finding suggests that there may be other substituents that can be placed at C
2
that could do the
same, perhaps more efficiently.
Introduction
electronic properties of the substituents appended to the
bridgehead carbons control the regioselectivity.
1
,2
When oxidized, housanes rearrange to produce cyclo-
pentenes (e.g., 1 to 2). The reaction proceeds via the
Our interest in this chemistry was kindled by the
1
1,2
existence of important mechanistic insights and by the
removal of an electron from the strained framework to
afford a cation radical. The existence of the intermediate
has been verified using a number of techniques including
desire to explore the use of these interesting processes
in synthesis. Our long-standing efforts to devise novel
3
routes to [5.3.0] and [6.3.0] ring systems, coupled with
2
ESR. As shown in Scheme 1, a Wagner-Meerwein
the realization that the oxidatively promoted rearrange-
ment of substrates 6 could, in principle, lead to the
assembly of these frameworks (7-9), prompted the
investigation described herein. Of the several possible
ways to assemble housanes, the methodology of H u¨ nig
used extensively by Adam and co-workers, seemed most
direct and best suited to meet our immediate needs; its
migration followed by back-electron transfer converts the
initial cation radical to the product. As illustrated by the
conversion of 3 to 4, the transformation is regioselective.
A substantial body of information indicates that the
†
Department of Chemistry & BiochemistrysUniversity of California-
Santa Barbara.
‡
University of CaliforniasDavis.
1) Adam, W.; Heidenfelder, T. Chem. Soc. Rev. 1999, 28, 359-
(
(3) (a) Billera, C. F.; Little, R. D. J. Am. Chem. Soc. 1994, 116 (12),
5487-5488. (b) Ott, M. M.; Little, R. D. J. Org. Chem. 1997, 62 (6),
1610-1616. (c) Carroll, Georgia L.; Allan, A. K.; Schwaebe, M. K.;
Little, R. D. Org. Lett. 2000, 2 (16), 2531-2534. (d) Mikesell, P. J.;
Little, R. D. Tetrahedron Lett. 2001, 42 (25), 4095-4097.
3
65.
(
2) (a) Adam, W.; Sahin, C.; Sendelbach, J.; Walter, H.; Chen, G.-
F.; Williams, F. J. Am. Chem. Soc. 1994, 116 (6), 2576-2584. (b)
Gerson, F.; Sahin, C. Helv. Chim. Acta 2001, 84 (6), 1470-1488.
10.1021/jo050087c CCC: $30.25 © 2005 American Chemical Society
4598
J. Org. Chem. 2005, 70, 4598-4608
Published on Web 04/21/2005

Rearrangements of Housane Cation Radicals
SCHEME 1
SCHEME 2
When EHOMO resides between -8.5 and -9.6 eV, cycload-
dition occurs. p-Methoxystyrene (-8.62), p-methylstyrene
(
-8.81), styrene (-9.00), cyclopentadiene (-9.08), and
p-bromostyrene (-9.10) are dienophiles whose HOMOs
fulfill this criterion and they undergo cycloaddition.
4
use afforded access to housanes 10a-d and 11. Their
chemistry, described below, provided unanticipated re-
sults.
One must not be too eager to apply the criterion
blindly, however. Thus, while EHOMO for allyl phenyl
sulfide is calculated to be -8.54 eV and therefore falls
within the specified range, it fails to react. This is easily
understood by realizing that the eigenfunction corre-
sponding to the HOMO is heavily localized on sulfur
rather than the alkene. Steric factors can also override
predictions based exclusively upon an examination of
HOMO energies. For example, styrene reacts reasonably
efficiently (53-56%), while 1,1-diphenylethylene and
R-methyl styrene fail to undergo cycloaddition despite the
fact that the HOMO energy for each dienophile resides
within the reactivity window defined above.
Results and Discussion
Synthesis of the Starting Materials. The key trans-
formations of the H u¨ nig chemistry, illustrated in Scheme
One dienophile we discovered to meet our qualitative
criteria was allyltrimethylsilane (16). We were gratified
to find that it and isopyrazole 13 react slowly, but
nonetheless efficiently, to undergo cycloaddition when
refluxed in the presence of TFA. As is described later in
this paper, the significance of this finding centers upon
the utility of the C-Si bond.7
2
, involve the reaction of a 1,3-diketone with hydrazine
hydrate, followed by a stereospecific, acid promoted,
inverse-electron-demand Diels-Alder reaction of the
resulting isopyrazole with a suitable dienophile. Unfor-
tunately, the number of “suitable” alkenes is limited.
While H u¨ nig and co-workers were able to expand the
scope of the cycloaddition by operating at elevated
4
5
pressures, we sought to do so while operating at atmo-
spheric pressure. Simple quantum calculations proved
useful in guiding our selection of systems to be studied.
In particular, we calculated the energy and eigenfunction
corresponding to the HOMO for a series of alkenes at
the AM1 level. From these results coupled with experi-
6
ment, we formulate the following guidelines: (a) When
E
HOMO is more negative than -10 eV, cycloaddition does
A series of C-2 aryl-substituted diazenes, 14a-c, was
assembled in the manner portrayed in Scheme 2 and
detailed in the Experimental Section. Diazene 14d
required special attention since the Diels-Alder reaction
of isopyrazole 13 and the p-carboethoxy substituted
not occur. Thus, phenyl vinyl sulfone (-10.19), vinyltri-
methylsilane (-10.24), and allyl bromide (-10.35) all
failed to react with the protonated isopyrazole 15. (b)
(4) Beck, K.; H u¨ nig, S. Chem. Ber. 1987, 120, 477-483.
(5) Beck, K.; H u¨ nig, S.; Klaerner, F.-G.; Kraft, P.; Artschwager-Perl,
U. Chem. Ber. 1987, 120, 2041-2051.
(7) Among a very large number of potential options, consider the
following: Lee, T. W.; Corey, E. J. Org. Lett. 2001, 3 (21), 3337-3339.
(b) Beignet J.; Cox L. R. Org. Lett. 2003, 5 (22), 4231-4234.
(6) Calculations were performed at the indicated level of theory
using Spartan ′02, Wavefunction, Inc.
J. Org. Chem, Vol. 70, No. 12, 2005 4599

Gerken et al.
styrene failed. Treatment of the p-bromo-substituted
diazene 14e with 2 equiv of t-BuLi and ethyl chlorofor-
mate, however, provided access to the desired framework
tional theory calculations (see the Computational Meth-
ods section) show that the cation radical of the major
product, 18, is also more stable than that of 19, by 1.4
kcal/mol at the B3LYP/6-31G(d) level of theory. Conse-
quently, we wondered whether the product ratio simply
reflected thermodynamic control. To address this issue,
a sample of the minor adduct, 19, was resubjected to the
reaction conditions. No change was observable by capil-
lary GC analysis after 10 min, 2, 3, or 19 h. We conclude
that the products do not interconvert, a result that is
consistent with the fact that the cation radicals of 18 and
19 are computed to be 27.4 and 26.0 kcal/mol lower in
energy than the initially formed cation radical, and
therefore that the ratio does not reflect thermodynamic
control.
1
4d. When irradiated, each diazene was efficiently
converted to the corresponding housane, 10a-d. A UV
hand lamp (∼365 nm) proved sufficient to effect the
transformation, albeit slowly because of the lamp’s low
intensity; no specialized photochemical equipment was
required.
Selection of the Oxidizing Agent. Both photoelec-
tron transfer and reagent-induced oxidation of housanes
have been shown to afford a cation radical.1 Of the
possible oxidizing reagents from which to choose, we
selected tris(4-bromophenyl)aminium hexachloroanti-
,2
•
+
monate (TBA ; Eox +1.06 V vs SCE) rather than the
more powerful oxidant tris(2,4-dibromophenyl)aminium
8
hexa-chloroantimonate (Eox +1.50 V vs SCE). The former
SCHEME 3
has been shown to be capable of oxidizing a variety of
housanes, including, for example, 1,4-dimethylbicyclo-
[
(
2.1.0]pentane and the methyl phenyl tricyclic system 3
1/2 1.42 V vs SCE). That it can, despite the fact that
E
the substrates have an oxidation potential more positive
than the oxidizing agent, is undoubtedly due to the
existence of a fast chemical reaction following the electron-
transfer event (e.g., rearrangement), a phenomenon that
is often observed in redox-mediated processes.8,9
The Rearrangement. As indicated in the Introduc-
tion, the regioselectivity of the rearrangement of housane
cation radicals is governed by the nature of the substit-
uents appended to the bridgehead carbons.1 Our expec-
tation, therefore, was that housanes 10a-d would not
display any selectivity since each bridgehead bears the
same substituent. Such was not the case.
,2
Treatment of housane 10a, dissolved in dry methylene
•
+
chloride, with 10 mol % of TBA led to the near-
quantitative production of two products, formed in un-
equal amounts (76:24), after only 5 min at room temper-
ature. Separation, isolation, and characterization con-
firmed the structures to be 18 and 19, the former being
dominant. Two spectral features greatly simplified the
identification process. In the major product (18), the
benzylic proton is also allylic and therefore appears
farther downfield than the benzylic proton in 19 (viz. 3.68
vs 3.18 in 19). The stereochemical assignment in the
minor isomer is also simplified by the fact that the
bridgehead methyl group is significantly shielded by the
phenyl group located on the adjacent carbon; it appears
at 0.6 ppm in 19. The regio- and stereochemical outcome
was confirmed from extensive NOE experiments; details
are found in the Supporting Information. Based upon
these results, we conclude that there is a preference for
migration toward the bridgehead carbon that is distal to
the phenyl substituent (viz. toward CR′ of 10a).
While the ca. 3:1 preference is not large, it is real and
reproducible. One cannot help but wonder why the
preference should exist at all. Clearly, the phenyl group,
while not positioned at one of the bridgehead carbons, is
not an innocent bystander. The phenyl group may be
involved in several ways: (a) it may be the site of the
initial oxidation, rather than the strained σ framework
of the housane, and (b) it may determine both the regio-
and stereochemical outcome of the rearrangement by
interacting directly with the site of migration. The local
ionization potential map illustrated in Scheme 3 is in
accord with point a.10 This map reflects the relative ease
of electron removal (ionization) from any location in a
(
8) (a) Adam, W.; Sahin, C. Tetrahedron Lett. 1994, 35 (48), 9027-
9
030. (b) Wend, R.; Steckhan, E. Electrochim. Acta 1997, 42 (13-14),
027-2039. (c) Steckhan, E. Anodic Oxidation of Nitrogen-Containing
2
Compounds. In Organic Electrochemistry, 4th ed.; Lund, H., Hammer-
ich, O., Eds.; Marcel Dekker: New York, 2001; Chapter 15.
(
9) (a) Steckhan, Ed. Angew. Chem. 1986, 98 (8), 681-699. (b)
Simonet, J. and Pilard, J.-F, Electrogenerated Reagents. In Organic
Electrochemistry, 4th ed.; Lund, H., Hammerich, O., Eds.; Marcel
Dekker: New York, 2001; Chapter 29. (c) Little, R. D.; Moeller, K. D.
Interface 2002, 11 (4), 36-42.
(
10) (a) Sjoberg, P.; Murray, J. S.; Brinck, T.; Politzer, P. Can. J.
Chem. 1990, 68, 1440-1443 (b) Murray, J. S.; Politzer, P. In Theoretical
Organic Chemistry. Theoretical & Computational Chemistry; Parkanyi,
C., Ed.; Elsevier: Amsterdam, 1998; Vol. 5, p 189. (c) Politzer, P.;
Murray, J. S.; Concha, M. C. Int. J. Quantum Chem. 2002, 88, 19-27.
Molecular mechanics calculations show that 18 is more
stable than 19, and hybrid Hartree-Fock/density func-
4600 J. Org. Chem., Vol. 70, No. 12, 2005

Rearrangements of Housane Cation Radicals
molecule. In this case, both the σ framework and the
phenyl group are obviously the most likely candidates.
One possibility posits the initial removal of the electron
from the aromatic unit to afford a cation radical (20)
wherein both the charge and odd electron are located
there and that this species is in equilibrium with another
cation radical (21) wherein the charge and odd electron
are localized on the strained σ bond that ultimately
cleaves. We were able to locate only one [2.1.0] bicyclic
cation radical species with B3LYP/6-31G(d) calculations,
however, and this species resembles 21 (Scheme 3), not
showing any significant delocalization onto the aromatic
ring.
Having made a case for bridging, the question re-
mains: Why should migration toward the distal carbon
be preferred? After all, bridging stabilizes a species that
is common to both paths (see 24). We suggest that the
differences show up after the common intermediate. The
major path may simply be the one that allows maximum
delocalization, and allows it to be maintained for as long
as possible.13 Migration away from the bridging unit
would allow the stabilization to be expressed until such
point as it is no longer “needed” (just prior to the back
electron transfer?), while migration to the proximal
carbon would immediately begin to disrupt bridging and
its attendant stabilization, thereby raising the activation
barrier for this process.
To account for the regioselectivity, we invoke the
existence of full or partial bridging between the phenyl
ring and the bridgehead carbon, C , to afford 22 (fully
R
bridged) or 23 (partially bridged), followed by the pref-
erential movement of the highlighted methylene unit
toward the distal bridgehead carbon (vide infra) (Scheme
1
1,12
4
).
In addition to providing a rationale for the regio-
selectivity, the existence of bridging blocks the bottom
face of the molecule thereby ensuring that the migration
occurs on the opposite side and accounting for the
stereochemical outcome that is observed in each of the
regioisomers.
SCHEME 4
To possibly supply insight into the factors that deter-
mine the regioselectivity, we examined the redox chem-
istry of the para-substituted housanes 10b-d. When each
was subjected to the conditions described previously,
rearrangement occurred. The results are summarized in
Table 1. Once again, there is a clear preference for the
migrating carbon to move away from the carbon that we
suggest is involved in bridging. Comparison of the two
most revealing cases, entry 1 where X is electron with-
drawing and entry 4 where it is strongly donating,
reveals nearly identical product distributions (80:20 vs
B3LYP/6-31G(d) calculations also provide evidence in
support of the notion that bridging plays a role. As
indicated above, the structure of the first formed cation
radical resembles that of the starting material with no
indication of involvement by the aryl substituent (21).
As the reaction proceeds (21 f 24), however, the angle
7
2:28). We conclude that the difference in energy between
between the ipso carbon-C
8° as the envelope flattens, thereby suggesting the
existence of a bridging interaction.
2 R
and C changes from 117° to
the transition states for distal and proximal migration is
insensitive to the nature of the substituents.
9
14
TABLE 1.
(
11) Adam, W.; Heidenfelder, T.; Sahin, C. J. Am. Chem. Soc. 1995,
1
17, 9693-9698. This paper describes an interesting example of
bridging in the rearrangement of the housane-derived cation radical
B. Unlike our system where bridging is invoked to temporarily stabilize
the bridgehead carbenium ion, with B it leads to the formation of a
σ-bond, and ultimately to D, E, and F.
entry
X
product ratio (25/26)
yield (%)
1
2
3
4
CO2Et
H
CH3
OCH3
80:20
76:24
76:24
72:28
87
>95
95
79 (GC)
Bridging ought to stabilize an intermediate by allowing
delocalization of the charge/spin. Accompanying it, how-
(
12) Pioneering studies of aryl bridging can be found in: (a) Cram,
(13) (a) Zimmerman, H. E.; Armesto, D. Chem. Rev. 1996, 96 (8),
3065-3112. (b) Hixson, S. S.; Mariano, P. S.; Zimmerman, H. E. Chem.
Rev. 1973, 73 (5), 531-551.
D. J. J. Am. Chem. Soc. 1952, 74, 2129-2137. (b) Cram, D. J. J. Am.
Chem. Soc. 1964, 86, 3767-3772.
J. Org. Chem, Vol. 70, No. 12, 2005 4601

Gerken et al.
TABLE 2.
reasonable: (1) The trimethylsilylmethyl group will have
no influence upon the regiochemical outcome since it is
a nonbridging alkyl substituent and (2) the trimethyl-
silylmethyl group will serve as the regiochemical control
unit and will direct the migration, through the existence
3
T (K)
25/26
ln(25/26)
10 /T
3
2
2
2
1
13
96
73
33
95
3
1.09
1.25
1.45
1.74
2.19
3.195
3.38
3.66
4.29
5.13
3.5
4.3
5.7
9
of a γ-silyl effect,15 toward the bridgehead carbon C
.
R
ever, would be a reduction in degrees of freedom, i.e.,
there is an entropic penalty associated with bridging that
could be manifested differently in the competing transi-
tion states leading to the two products.1 To determine
the relative importance of enthalpic and entropic factors
in determining the product selectivity, we investigated
the rearrangement of the p-methoxy substituted housane
4b
While the ability of an R-silyl group to stabilize either
a carbanion or a carbenium ion and a â-silyl group to
stabilize a carbenium ion is well-known,16 the capacity
of a γ-silyl unit to stabilize intermediates is less appreci-
ated. Whitmore and co-workers first described silicon’s
stabilizing interaction with a positive charge positioned
γ to it in 1946.15 Since then, many studies have focused
upon accelerating effects in solvolysis reactions of sys-
tems bearing a 1,3-relationship between a C-Si bond and
a leaving group.17 In general, authors account for dra-
matic rate increases by invoking the existence of a
stabilizing interaction between the back lobe of the γ-C-
Si σ-bond and the incipient positive charge. This so-called
1
0c as a function of temperature. The ratio of the major
to the minor product, 25/26 (X ) OMe), was determined
at five different temperatures. The results are sum-
marized in Table 2.
3
When ln(25/26) is plotted as a function of 10 /T, a
straight line results whose slope and intercept are 0.55
and -0.62, respectively. Since ln(25/26) ) -∆∆H /RT +
∆
q
q
∆S /R, it follows that the difference in enthalpies of
activation for the competing pathways is -1.09 kcal/mol,
while the differences in entropies of activation is only
-
1.23 eu. The magnitudes of these quantities indicate
“
percaudal interaction” is maximized when the interact-
that enthalpic factors dominate over the entire temper-
ature range that was explored.
ing centers are related by a W-like geometry similar to
that highlighted in the structure of the silyl housane 11,
18
Silylhousane 11. We wondered how a housane that
was devoid of an aryl substituent would behave since its
presence, we suggest, is responsible for the selectivity
observed in the reactions described thus far. Toward this
end, we converted diazene 17 to the trimethylsilyl-
methylhousane 11 in the manner portrayed in the
accompanying equation. A priori, two scenarios appear
and commonly invoked in discussions of long-range
coupling in NMR spectroscopy. The ability of a γ-silyl
substituent to stabilize a full or an incipient positive
charge should, therefore, direct the rearrangement of 11
R
so that the methylene unit will migrate toward C .
When 11 was subjected to the oxidative conditions
described previously, nearly equal amounts of 27 and 28
were obtained in an 82% yield. That no regiochemical
preference was observed supports the suggestion that full
or partial bridging to the aryl substituent controls the
regiochemical outcome for the aryl-substituted housanes
(14) (a) At this point, neither of the structures represented by G
and H for the intermediate immediately preceding the alkyl shifts can
be definitively ruled out. Previous studies have suggested that the
preferred direction of migration for housane cation radicals is, in
general, toward the bridgehead that bears the bulk of the positive
charge, thus suggesting that H may be involved. This view is consistent
with the substituent effects described in the text, since both electron-
donating and electron-withdrawing substituents stabilize radicals.
However, our preliminary B3LYP calculations suggest that G best
represents the structure formed by ring-opening/flattening of the
initially formed housane cation radical. In addition, our calculations
suggest that the generalization described above may not correctly
predict the preferred direction of migration in some other systems as
well. Additional studies are necessary to settle this issue, however,
and these will be reported in due course. (b) So far, we have not been
able to rationalize the observed regiochemical preferences using our
preliminary B3LYP/6-31G(d) calculations. We should note here that
it appears that the potential energy surfaces in the vicinity of the
housane and “open” cation radicals and the various rearrangement
transition structures are rather flat, complicating the analysis. Further
explorations using other theoretical methods (including attempts with
CI methods) are in progress. We are also considering the effects of
entropy and the possibility of dynamic control. The results of these
studies will be reported in due course.
1
0a-d.
The result also indicates that a γ-silyl effect is not
operative, an outcome that appears to be at odds with
the commentary presented above. Several factors may
(
15) Sommer, L. H.; Dorfman, E.; Goldberg, G. M.; Whitmore, F. C.
J. Am. Chem. Soc. 1946, 68, 488-489. (b) For a review concerning the
interaction of silicon with positively charged carbon at the R, â, γ, and
δ positions see Lambert, J. B. Tetrahedron 1990, 46 (8), 2677-2689.
(
16) (a) Ushakov, S. N.; Itenberg, A. M. Zh. Obshch. Khim. 1937, 7,
2
495-2498. (b) Lambert, J. B.; Wang, G.-T.; Finzel, R. B.; Teramura,
D. H. J. Am. Chem. Soc. 1987, 109, 7838-7845. (c) Brinkman, E. A.;
Berger, S.; Brauman, J. I. J. Am. Chem. Soc. 1994, 116 (18), 8304-
8
310.
(
17) (a) Green, A. J.; Pigdon, T.; White, J. M.; Yamen, J. J. Org.
Chem. 1998, 63, 3943-3951. (b) Bentley, T. W.; Kirmse, W.; Llewellyn,
S o¨ llenb o¨ hmer, F. J. Org. Chem. 1990, 55, 1536-1540. (c) Engler, T.
J.; Reddy, J. P. J. Org. Chem. 1991, 56, 6491-6494.
(
18) Kaimakliotis, C.; Fry, A. J. J. Org. Chem. 2003, 68 (26), 9893-
9
898 reports a novel silicon-aryl effect upon redox potential.
4602 J. Org. Chem., Vol. 70, No. 12, 2005

Rearrangements of Housane Cation Radicals
contribute to the absence of an observed γ-silyl effect in
this case: (1) a carbenium ion formed at C would be
R
pentene-stabilized form. Basic alumina was purchased from
a commercial supplier and dried in an oven heated to 130 °C.
1
,1-Diacetylcyclopentane (Diketone 12). To 100 mL of
tertiary. In a study appearing in 1999, Tilley and Shiner
concluded that tertiary systems do not require, and are
consequently not significantly stabilized by, γ-silyl
DMSO was added 12.412 g (90 mmol) of powdered anhydrous
potassium carbonate, followed by 5.0 mL (4.88 g, 49 mmol) of
2
,4-pentanedione. The mixture was stirred under argon while
.5 mL (11.97 g, 55 mmol) of 1,4-dibromobutane was added
1
9
groups. (2) The geometry that is noted by the high-
lighted bonds in 11 may not correspond to the idealized
W-like geometry that is needed for maximum benefit. In
fact, by viewing an AM1 geometry-optimized structure
6
dropwise over 10 min. The mixture was allowed to stir at room
temperature under argon for 24 h, and then 50 mL of brine
was added followed by 100 mL of deionized water. The mixture
was extracted with 5 × 50 mL of diethyl ether. The combined
organic extracts were dried over magnesium sulfate, and the
solvent was removed under vacuum. The crude product was
then purified via flash chromatography over silica gel in 4:1
hexanes/ethyl acetate. After concentration in vacuo, 4.442 g
of 12 was isolated (28.8 mmol, 60%) and determined by GC/
6
of the reactant from the top, one observes that the W-like
orientation is accommodated by the C
than the C -CR′ bond that undergoes cleavage. In
particular, when the C -Si bond is anti-periplanar to the
-C bond, as it is in the geometry-optimized structure,
the dihedral angle C -C -C -CR′ is 120°, while the C
-C -CR′′ dihedral angle is a nearly ideal 178°. In
addition, the cation radical produced from oxidation of
R
-CR′′ bond, rather
R
γ
C
â
R
γ
-
13
γ
â
R
MS to be 94% pure. Proton NMR and C NMR data matched
4
C
â
R
the literature values of Beck and H u¨ nig.
3′,5′-Dimethylspiro[cyclopentane-1,4′-[4H]pyrazole]
(Isopyrazole 13). Diketone 12 (3.512 g, 22 mmol) was
weighed into a one-necked flask equipped with a water-cooled
condenser, a boiling stone, and 11 mL of dry methylene
chloride (distilled from calcium hydride). Hydrazine hydrate
1
1 (B3LYP/6-31G(d) with methyls on Si replaced by H’s,
9) shows no significant bending of the CH SiR group
2
2
3
toward the bridgehead carbon. Interestingly, however,
the CH SiR group does appear to interact somewhat
2
3
(
1.3 mL, 27 mmol) was added, and the mixture was heated to
with this carbon upon conversion to the “open” form of
the cation radical (30), thereby stabilizing positive charge
at that position. Apparently, the regiochemical preference
one would expect based on this interaction is somehow
overwhelmed, given the regiorandom nature of the
reflux overnight. The reaction mixture was then allowed to
cool, dried over magnesium sulfate, and filtered. The solvent
was removed under vacuum to give a white crystalline solid
(mp 111-113 °C) in 2.967 g isolated yield (19.78 mmol, 87%).
GC/MS of the product showed no detectable impurities, and
4
1
4b
NMR data matched the literature values of Beck and H u¨ nig.
transformation.
1,4-Dimethyl-anti-5-phenyl-2,3-diazaspiro[bicyclo[2.2.1]-
heptane-7,1′-cyclopentan]-2-ene (Phenyldiazene 14a).
Isopyrazole 13 (1.026 g, 6.8 mmol) was dissolved in 50 mL of
dry chloroform (distilled from calcium oxide) and transferred
to a 250 mL round-bottomed flask. Trifluoroacetic acid (0.7
mL, 9.1 mmol) was added with swirling. Styrene (20 mL,
1
75 mmol) was added, a water-cooled reflux condenser was
attached, and the system was purged with argon. The mixture
was heated to reflux for 3 days and allowed to cool to ambient
temperature. Powdered anhydrous potassium carbonate (2.007
g, 14.5 mmol) was added, and the mixture was stirred at room
temperature for 1 h. The mixture was filtered, solvents were
removed under vacuum, and the crude product was chromato-
graphed on silica gel in 4:1 hexanes/ethyl acetate. Phenyl
diazene 14a was isolated (973 mg, 3.83 mmol, 56%) as a white
powder.
Concluding Remarks
The present study provided surprises at every turn,
the most significant being that rearrangement of the
2
-aryl-substituted systems occurred regioselectively. Thus,
it is possible to influence the direction of migration even
when the substituent is not directly appended to the
migration terminus. This finding suggests that there may
The spectral data were: 1H NMR (500 MHz, CDCl
) 7.2 (m,
3
3
1
H), 7.0 (m, 2H), 2.94 (dd, 1H), 2.03 (dd, 1H), 1.74 (s, 3H),
.6-1.3 (m, 10H), 1.2 (m, 2H); NOE enhancements were
be other substituents that can be placed at C
do the same, perhaps more efficiently. We are intrigued
by functional groups such as OR, SAr, Si(OR′)R , NR
and CO R since they could stabilize a positive charge at
the bridgehead carbon, thereby influencing the regio-
chemical outcome, and they would also provide a handle
for functional and structural modifications. Additional
experiments and computations are underway to achieve
these objectives and to gain a deeper understanding of
the mechanisms and origins of selectivity for these
fascinating rearrangements.
2
that could
observed: irradiation at 7.0 ppm led to enhancements at 7.2,
6
%; 2.94, 60 6.7%; 1.74, 2.1%; 1.4, 4.6%; irradiation at 2.94
3
2
,
ppm, afforded enhancements at 7.2, 5.4%; 2.03, 3.3%; 1.5,
12.1%; irradiation at 2.03 ppm, led to enhancements at 2.94,
2
13
5
1
1
2
.18%; 1.74, 2.08%; 1.5, 5.4%; 1.4, 17.6%; C APT 180.7 (quat),
38.8 (quat), 129.0 (tertiary), 128.0 (tertiary), 126.8 (tertiary),
05.9 (quat), 47.6 (tertiary), 37.5 (secondary), 28.2 (secondary),
6.8 (secondary), 26.6 (secondary), 26.3 (secondary), 13.5
(
primary), 11.4 p; IR (neat, attenuated total internal reflec-
tance (ATR)) 3063, 3040, 2948, 2865, 1601, 1495, 1459, 1380,
759, 698 cm-1; MS (EI) 226, 211 (base peak), 183, 169, 141,
4
115, 91, 65, 41; HRMS (CI, CH ) 255.1854 obsd, 255.1861 calcd
+
for C17
,4-Dimethyl-anti-5-(4-methoxyphenyl)-2,3-diazaspiro-
bicyclo[2.2.1]heptane-7,1′-cyclopentan]-2-ene (p-Meth-
23 2
H N (M + H) .
1
Experimental Section
[
General Methods. Solvents used in reactions were pre-
pared via the following methods. Tetrahydrofuran was distilled
from Na/benzophenone, and methylene chloride and aceto-
phenone were distilled from calcium hydride. Chloroform was
either distilled from calcium hydride or purchased in predried,
oxyphenyldiazene 14c). Isopyrazole 13 (0.750 g, 5 mmol)
was dissolved in 5 mL of distilled chloroform. Trifluoroacetic
acid (0.34 mL, 4.5 mmol) and 5.0 g (37.5 mmol) of 4-methoxy-
styrene were added with stirring and heated to reflux. After
42 h, the mixture was allowed to cool to room temperature,
2 3
1.295 g (9.4 mmol) of powdered anhydrous K CO was added,
and the mixture was stirred for 35 min. The mixture was
filtered, solvents were removed under vacuum, and the crude
(
19) Tilley, L. J.; Shiner, V. J., Jr. J. Phys. Org. Chem. 1999, 12,
5
64-576.
J. Org. Chem, Vol. 70, No. 12, 2005 4603

Gerken et al.
product was chromatographed in 4:1 hexanes/ethyl acetate.
acetate/hexanes to yield 0.312 g (0.96 mmol, 26%) of p-
ethoxycarbonylphenyldiazene 14d.
p-Methoxyphenyldiazene 14c was isolated (0.590 g, 2.1 mmol,
4
1%) as a white powder.
The spectral data were: 1H NMR (400 MHz, CDCl
3
) δ 7.89
The spectral data were: 1H NMR (400 MHz, CDCl
) δ 6.90
3
(m, 2H), 7.04 (m, 2H), 4.35 (q, J ) 7.2, 2H), 3.00 (dd, J ) 9.2,
5.4, 1H), 2.05 (dd, J ) 13.6, 9.2, 1H), 1.77 (s, 3H), 1.60-1.43
(m, 5H), 1.53 (s, 3H), 1.41 (t, J ) 7.2, 3H), 1.39-1.20 (m, 4H);
(
m, 2H), 6.76 (m, 2H), 3.78 (s, 3H), 2.89 (dd, J ) 9.2, 5.4, 1H),
2
1
1
1
2
2
.02 (dd, J ) 13.7, 9.2, 1H), 1.72 (s, 3H), 1.7-1.55 (m, 2H),
1
3
.53 (s, 3H), 1.5-1.35 (m, 4H), 1.35 (dd, J ) 13.7, 5.2, 1H),
C NMR (100 MHz, CDCl ) 166.5, 144.2, 129.2, 129.0, 128.9
3
.28-1.16 (m, 2H); 13C NMR (100 MHz, CDCl
) 158.4, 130.7,
3
90.5, 90.0, 67.6, 60.8, 47.6, 37.3, 28.2, 26.7, 26.5, 26.2, 14.3,
29.8, 113.4, 90.5, 89.9, 67.4, 55.2, 46.7, 37.7, 28.2, 26.7, 26.5,
6.2, 13.4, 11.3; IR (neat) 2953, 1611, 1250, 1033; MS (CI, CH
85, 256, 241, 179, 151 (base peak); HRMS (CI, CH ) 285.1959
13.4, 11.3; IR (neat) 2961, 1714, 1275, 1108; MS (CI, CH ) 327,
4
4
)
298, 283, 151 (base peak), 131; HRMS (CI, CH ) 327.2068 obsd,
4
+
4
327.2073 calcd for C H N O (M + H) .
2
0
27
2
2
+
obsd, 285.1967 calcd for C18
,4-Dimethyl-anti-5-(4-methylphenyl)-2,3-diazaspiro-
bicyclo[2.2.1]heptane- 7,1′-cyclopentan]-2-ene (p-Methyl-
25 2
H N O (M + H) .
1
,4-Dimethyl-anti-5-(trimethylsilylmethylene)-2,3-di-
azaspiro[bicyclo[2.2.1] heptane-7,1′-cyclopentan]-2-ene
Trimethylsilyldiazene 17). Isopyrazole 13 (1.405 g, 9.4
1
[
(
phenyldiazene 14b). Isopyrazole 13 (0.722 g, 4.81 mmol) was
dissolved in 5.5 mL of chloroform in a round-bottomed flask.
Trifluoroacetic acid (0.35 mL, 4.5 mmol) was added with
stirring, followed by 10 mL (75 mmol) of 4-methylstyrene. This
mixture was then refluxed for 45 h and allowed to cool. Dry,
powdered potassium carbonate (1.322 g, 9.6 mmol) was added,
and the mixture was stirred for 1 h before being filtered. The
filtrate was concentrated in vacuo and purified by chromatog-
raphy on silica gel in 20% ethyl acetate/hexanes. p-Methyl-
phenyldiazene 14b (1.012 g, 3.78 mmol, 79%) was isolated as
a white powder.
mmol) was dissolved in 25 mL of dry, pentene-stabilized
chloroform. Allyltrimethylsilane (17 mL, 140.5 mmol) and 0.65
mL (8.43 mmol) of trifluoroacetic acid were added, and the
system was nitrogen purged. The mixture was heated to reflux
with stirring and allowed to reflux for 2 days before being
allowed to cool to room temperature. Powdered dry K
2 3
CO
(
2.44 g, 18 mmol) was added and the mixture allowed to stir
for 30 min. The mixture was filtered, solvents were removed
under vacuum, and the crude product was chromatographed
on silica gel in 4:1 hexanes/ethyl acetate. Trimethylsilyldiazene
1
7 was isolated (1.684 g, 6.38 mmol, 68%) as a white powder.
The spectral data were: 1H NMR (200 MHz, CDCl
d, 2H), 6.85 (d, 2H), 2.92 (dd, 1H), 2.30 (s, 3H), 2.02 (dd, 1H),
) 7.05
3
1
The spectral data were: H NMR (200 MHz) 1.70 (m, 2H),
(
1
2
.60 (s, 3H), 1.59 (s, 3H), 1.43 (m, 4H), 1.29 (m, 2H), 1.16 (m,
1
.75 (s, 3H), 1.6-1.4 (m, 6H), 1.55 (s, 3H), 1.42 (dd, 1H), 1.3-
.15 (m, 2H); C APT 136.4 (quat), 135.7 (quat), 128.9
tertiary), 128.8 (tertiary), 90.5 (quat), 89.98 (quat), 67.6 (quat),
7.3 (tertiary), 37.6 (secondary), 28.3 (secondary), 26.8 (second-
ary), 26.6 (secondary), 26.3 (secondary), 20.97 (primary), 13.5
) 269, 267, 240, 225,
51 (base peak), 105; HRMS (CI, CH ) 269.2007 obsd, 269.2018
13
H), 0.79 (dd, 1H), 0.60 (m, 1H), -0.08 (s, 9H); C APT 91.3
1
3
1
(
quat), 89.4 (quat), 66.5 (quat), 38.1 (tertiary), 37.5 (secondary),
(
2
8.2 (secondary), 27.1 (secondary), 26.8 (secondary), 26.5
secondary), 16.5 (secondary), 13.8 (primary), 11.3 (primary),
0.8 (primary); IR (neat ATR) 2952, 2871, 1458, 1381, 1245,
4
(
-
(
primary), 11.4 (primary); MS (CI, CH
4
-
1
8
39 cm ; MS (CI, CH
4
) 265, 249, 236, 223, 179, 151 (base
) 265.2089 obsd, 265.2100 calcd for
1
4
peak), 73; HRMS (CI, CH
4
+
calcd for C18
25 2
H N (M + H) .
+
C
15
H
29
N
2
Si (M + H) .
,4-Dimethyl-anti-2-phenylspiro[bicyclo[2.1.0]pentane-
,1′-cyclopentane] (Phenylhousane 10a). Diazene 14a
1.259 g, 4.95 mmol) was weighed into a 250 mL Pyrex round-
bottomed flask and dissolved in 50 mL of dry tetrahydrofuran
distilled from Na/benzophenone). The system was degassed
1
,4-Dimethyl-anti-5-(4-bromophenyl)-2,3-diazaspiro-
bicyclo[2.2.1]heptane-7,1′-cyclopentan]-2-ene (p-Bromo-
phenyldiazene 14e). Isopyrazole 13 (0.787 g, 5.24 mmol) was
dissolved in 14 mL of CHCl and purged with nitrogen.
1
[
5
(
3
Trifluoroacetic acid (0.4 mL, 5.2 mmol) was added with
stirring, followed by 10 g (55 mmol) of 4-bromostyrene. The
mixture was refluxed under nitrogen for 4 days and then
(
by bubbling argon through it for 10 mins. The solution was
then irradiated with 365 nm light from a UV hand lamp placed
adjacent to the flask, with periodic monitoring by TLC in 4:1
hexanes/ethyl acetate. After 20 h, no more diazene was
observed. The solvent was then removed under vacuum, and
the crude product was chromatographed on silica gel with
cyclohexane. Housane 10a (1.120 g) was isolated as a clear,
viscous liquid (4.95 mmol, >95%).
2 3
allowed to cool. Powdered dry K CO (1.046 g, 7.6 mmol) was
added and allowed stir for 1.5 h before the mixture was
filtered. The filtrate was concentrated in vacuo and chromato-
graphed on silica gel with 20% ethyl acetate/hexanes to yield
0
.982 g (2.95 mmol, 59%) of the diazene 14e.
The spectral data were: H NMR (200 MHz, CDCl
1
3
) 7.37
(
m, 2H), 6.85 (m, 2H), 2.90 (dd, 1H), 2.05 (dd, 1H), 1.75 (s,
The spectral data were: 1H NMR (400 MHz, CDCl
7.15 (m, 5H), 2.84 (dd, 1H), 2.09 (dd, 1H), 1.88 (dd, 1H), 1.85-
3
) 7.4-
3
1
H), 1.6 (m, 2H), 1.52 (s, 3H), 1.5-1.4 (m, 4H), 1.32 (dd, 1H),
13
.25-1.18 (m, 2H); C APT 137.9 (quat), 131.1 (tertiary), 130.7
1
.65 (m, 6H), 1.45-1.4 (m, 2H), 1.20 (s, 3H), 0.73 (s, 3H); NOE
(
(
tertiary), 90.1 (quat), 47.2 (tertiary), 37.6 (secondary), 28.2
secondary), 26.8 (secondary), 26.6 (secondary), 26.3 (second-
enhancements were observed: irradiated at 2.84 ppm, en-
hancements at 7.30, 8.3%, 2.08, 2.4%, 1.87, 2.1%, 1.82, 5.9%,
ary), 13.5 (primary), 11.3 (primary); MS (CI, CH
4
) 335, 333,
0.73, 0.7%, irradiation at 2.08, afforded enhancements at 2.84,
3
3
04, 302, 291, 289, 179, 151 (base peak); HRMS (CI, CH )
4
+
3.8%, 1.87, 23%, 1.82, 4.9%, 1.2, 0.5%, irradiation at 1.87,
afforded enhancements at 7.30, 5.9%, 2.84, 1.7%, 2.08, 21%,
31.0820 obsd, 331.0810 calcd for C17
,4-Dimethyl-anti-5-(4-ethoxycarbonylphenyl)-2,3-di-
azaspiro[bicyclo[2.2.1] heptane-7,1′-cyclopentan]-2-ene
20 2
H N Br (M - H) .
1
1
3
1
.2, 2.3%; C APT 145.5 (quat), 128.7 (tertiary), 128.3 (ter-
tiary), 125.98 (tertiary), 42.9 (tertiary), 36.7 (secondary), 29.0
(
p-Ethoxycarbonylphenyldiazene 14d). A round-bottomed
(
secondary), 27.1 (secondary), 25.7 (secondary), 24.9 (second-
flask equipped with a magnetic stirbar was oven-dried and
allowed to cool under nitrogen. The flask was charged with
ary), 12.9 (primary), 9.2 (primary); MS (EI) 226, 211 (base
peak), 169, 115, 91,41; HRMS (EI) 226.1713 obsd, 226.1722
100 mL of dry THF and 1.209 g (3.63 mmol) of p-bromophenyl
+
calcd for C17
H
22 M .
diazene 14e. The solution was stirred and chilled in a dry ice/
acetone bath. To the cold solution was added 8 mL of a 1 M
solution of tert-butyllithium and the mixture allowed to stir.
After 1 h, 7 mL (7.9 g, 73 mmol) of ethyl chloroformate was
added in one portion. The solution was then allowed to come
to room temperature. TLC on silica gel with 20% (v/v) ethyl
acetate/hexanes showed two spots with a positive indication
to the CuCl diazene stain. The solution was concentrated in
1,4-Dimethyl-anti-2-(4-methylphenyl)spiro[bicyclo[2.1.0]-
pentane-5,1′-cyclopentane] (p-Methylphenylhousane 10b).
Diazene 14b (0.9660 g, 3.60 mmol) was dissolved in 70 mL of
dry THF and stirred while being purged with nitrogen. The
solution was then irradiated with 365 nm light from a UV
handlamp placed against the flask with monitoring by TLC
in 4:1 hexanes/ethyl acetate. After 24 h, no more diazene was
observed. The solution was concentrated in vacuo and chro-
matographed on silica 66 gel with cyclohexane as eluent to
vacuo and then quenched with K
2 3
CO in ethanol. The products
were isolated via chromatography on silica gel in 20% ethyl
4604 J. Org. Chem., Vol. 70, No. 12, 2005

Rearrangements of Housane Cation Radicals
yield 0.762 g of 4-methylphenyl housane 10b (3.175 mmol,
bottomed flask and was stirred while being purged with
nitrogen. The solution was irradiated with 365 nm light from
a UV hand lamp. After 24 h, no diazene was detected via TLC,
and the mixture was concentrated in vacuo. The crude mate-
rial was chromatographed on silica gel in cyclohexane; the
desired fraction was concentrated in vacuo to obtain 0.375 g
of trimethylsilylmethyl housane 11 (1.59 mmol, 71%).
88%) as a clear liquid.
The spectral data were: 1H NMR (200 MHz, CDCl
) 7.25
3
(
m, 4H), 2.92 (t, 1H), 2.45 (s, 3H), 2.19 (dd, 1H), 1.95 (dd, 1H),
1
.9-1.7 (m, 6H), 1.6-1.45 (m, 2H), 1.30 (s, 3H), 0.86 (s, 3H);
1
3
C APT 142.3 (quat), 135.3 (quat), 128.9 (tertiary), 128.4
(
tertiary), 42.4 (tertiary), 36.7 (secondary), 28.98 (secondary),
1
2
7.1 (secondary), 25.6 (secondary), 24.8 (secondary), 21.2
primary), 12.9 (primary), 9.1 (primary); MS (EI) 240, 225 (base
peak), 197, 183, 105, 91; HRMS (EI) 240.1884 obsd, 240.1878
The spectral data were: H NMR (500 MHz, CDCl ) 1.82
3
(
(dd, 1H), 1.63 (m, 1H), 1.69-1.55 (m, 6H), 1.40-1.30 (m, 2H),
1.13 (dd, 1H), 1.01 (s, 3H), 0.94 (s, 3H), 0.80 (dd, 1H), 0.60 (m,
+
13
calcd for C18
,4-Dimethyl-anti-2-(4-methoxyphenyl)spiro[bicyclo-
2.1.0]pentane-5,1′-cyclopentane] (p-Methoxyphenyl-
H
24 M .
1H), -0.02 (s, 9H); C APT 37.6 (quat), 37.1 (secondary), 33.2
1
(quat), 33.2 (quat), 31.5 (tertiary), 28.7 (secondary), 26.9
(secondary), 25.5 (secondary), 24.4 (secondary), 21.9 (second-
ary), 13.6 (primary), 7.8 (primary), -0.8 (primary); MS (EI)
[
housane 10c). Diazene 14c (0.575 g, 2.02 mmol) was dissolved
in 40 mL of dry THF and stirred while purging with nitrogen.
The solution was irradiated with 365 nm light from a UV
handlamp placed against the flask with monitoring by TLC
in 4:1 hexanes/ethyl acetate. After 25 h, no more diazene was
observed. The solution was concentrated in vacuo, and the
crude product was chromatographed on silica in 4:1 hexanes/
ethyl acetate. Housane 10c was isolated as a clear, viscous
liquid (0.475 g, 1.86 mmol, 92%).
236, 221, 149, 122, 73 (base peak); HRMS (EI) 236.1962 obsd,
+
236.1960, calcd for C15
H
28Si M .
3,7a-Dimethyl-2-phenyl-2,4,5,6,7,7a-hexahydro-1H-
indene and 3,7a-Dimethyl-1-phenyl-2,4,5,67,7a-hexa-
hydro-1H-indene (Phenylbicyclononene 18 and 19).
Housane 10a (185 mg, 0.82 mmol) and 1 drop (∼8.5 mg, ∼0.04
mmol) of 2,6-di-tert-butylpyridine were dissolved in 5 mL of
dry methylene chloride. This solution was passed through a
short column of oven-dried basic alumina into an argon-
purged, flame-dried 25 mL flask. The column was then washed
The spectral data were: 1H NMR (400 MHz, CDCl
m, 2H), 6.89 (m, 2H), 3.82 (s, 3H), 2.79 (dd, J ) 5.0, 4.0, 1H),
) δ 7.19
3
(
2
1
5
9
.06 (dd, J ) 11.4, 5.5, 1H), 1.86-1.60 (m, 7H), 1.40 (m, 1H),
.18 (s, 3H), 0.72 (s, 3H); 13C APT 157.7, 137.4, 129.2, 113.5,
5.22, 41.8, 38.2, 36.7, 33.0, 28.8, 28.9, 26.9, 25.5, 24.5, 12.8,
with an additional 5 mL of CH
aminium hexachloroantimonate (60 mg, 0.07 mmol) was
weighed out and dissolved in 5 mL of dry CH Cl . This solution
2 2
Cl . Tris(4-bromphenyl)-
2
2
.0; MS (EI) 256, 241 (base peak), 213, 199, 121, 91, 41; HRMS
was then added dropwise, with stirring, via cannula to the
housane solution. The mixture was allowed to stir for 5 min
before being filtered through a short column of oven-dried basic
+
(
EI) 256.1834 obsd, 256.1827 calcd for C18
H
24O M .
1,4-Dimethyl-anti-2-(4-bromophenyl)spiro[bicyclo[2.1.0]-
pentane-5,1′-cyclopentane] (p-Bromophenylhousane 10e).
Diazene 14e (0.467 g, 1.40 mmol) was dissolved in 50 mL of
dry THF and purged with argon. The solution was then
irradiated with 365 nm light from a UV handlamp placed
against the flask with periodic monitoring by TLC in 4:1
hexanes/ethyl acetae. After 25 h, no more diazene was detected
and the mixture was concentrated in vacuo. The crude product
was chromatographed on silica with cyclohexane. The product
was then concentrated in vacuo to yield p-bromophenyl-
housane 10e (0.396 g, 1.30 mmol, 93%).
2 2
alumina that was then washed with 5 mL of CH Cl . Solvents
were removed in vacuo, and the crude mixture was chromato-
graphed over silica gel with cyclohexane as eluent. This was
followed by removal of the solvent to afford a clear, viscous oil
that was pure by HPLC on silica in cyclohexane (HPLC, 7.00
min retention time) in quantitative yield. Subsequent runs
were chromatographed in hexanes, and isomers 18 and 19
were separated.
The spectral data for isomer 18 were: 1H NMR (500 MHz,
CDCl ) 7.4-7.2 (m, 5H), 3.68 (dd, 1H), 2.42 (m, 1H), 2.20 (dd,
3
The spectral data were: 1H NMR (500 MHz, CDCl
m, 2H), 7.15 (m, 2H), 2.78 (dd, 1H), 2.05 (dd, 1H), 1.8-1.6
) 7.42
3
1H), 2.00 (m, 1H), 1.84 (m, 1H), 1.70-1.55 (m, 4H), 1.46 (s,
3H), 1.40-1.25 (m, 1H), 1.25-1.17 (m, 1H), 1.13 (s, 3H); NOE
enhancements were observed: irradiated at 7.2 ppm, enhance-
ments at 3.68, 2.5%, 1.46, 0.8%, 1.13, 1%, irradiation at 3.68,
afforded enhancements at 7.15, 4.5%, 2.20, 3.6%, 1.46, 1.4%,
irradiation at 2.46, afforded enhancements at 2.00, 13%, 1.84,
4.5%, 1.46, 3.5%, irradiation at 2.20, afforded enhancements
(
(
13
m, 8H), 1.36 (m, 2H), 1.15 (s, 3H), 0.69 (s, 3H); C NMR 144.3,
1
2
1
31.1, 130.1, 127.98, 119.4, 42.2, 38.3, 36.5, 32.7, 28.8, 26.9,
5.4, 24.6, 12.7, 8.95; MS (EI) 306, 304, 291, 289 (base peak),
68, 122, 93; HRMS (EI) 304.0837 obsd, 304.0827 calcd for
+
17
C H
21Br M .
1
3
1
,4-Dimethyl-anti-2-(4-ethoxycarbonylphenyl)spiro-
at 3.68, 3.1%, 1.65, 13%, 1.35, 3.5%; C NMR 147.1, 143.7,
128.9, 128.5, 128.1, 127.96, 125.8, 54.5, 49.5, 47.2, 42.4, 30.3,
28.2, 25.5, 24.0, 23.1, 14.3, 12.6; MS 70 (EI) 226, 211 (base
[
bicyclo[2.1.0]pentane- 5,1′-cyclopentane] (p-Ethoxycar-
bonylphenylhousane 10d). A round-bottomed flask equipped
with a stirbar was charged with 40 mL of distilled THF and
peak), 183, 169, 91; HRMS (EI) 226.1726 obsd, 226.1722 calcd
+
0
.284 g (0.87 mmol) of p-ethoxycarbonyldiazene 14d. The
for C17H
22 M .
The spectral data for isomer 19 were: 1H NMR (500 MHz,
CDCl ) 7.35-7.20 (m, 5H), 3.10 (dd, 1H), 2.77 (dd, 1H), 2.42
solution was degassed by bubbling nitrogen through it for 10
min as it stirred. The flask was then irradiated with nominally
3
3
65 nm light from a UV hand lamp placed against the flask
(m, 1H), 2.30 (dd, 1H), 1.82 (m, 4H), 1.67 (s, 3H), 1.65-1.55
(m, 2H), 1.40 (dd, 1H), 0.58 (s, 3H); NOE enhancements were
observed: irradated at 3.10 ppm, enhancements at 7.3, 4.9%,
2.30, 3.8%, 1.40, 4.0%, irradiation at 2.77, afforded enhance-
ments at 7.3, 9.9%, 2.30, 20.5%, 1.67, 1.7%, 0.58, 1.8%,
irradiation at 2.42, afforded enhancements at 1.82, 19.6%, 1.67,
3.6%, irradiated at 2.30, afforded enhancements at 3.10, 4.0%,
2.77, 17.1%, 1.67, 2.4%, irradiation at 0.58, afforded enhance-
for 25 h. The solution was concentrated in vacuo and chro-
matographed on silica gel in 10% (v/v) ethyl acetate/hexanes
to yield 0.098 g (0.33 mmol, 38%) of p-ethoxycarbonylphenyl-
housane 10d.
The spectral data were: 1H NMR (400 MHz, CDCl
) δ 8.02
3
(
m, 2H), 7.31 (m, 2H), 4.37 (q, J ) 7.1, 2H), 2.89 (dd, J ) 5.1,
5
1
3
1
2
1
.1, 1H), 2.08 (dd, J ) 11.5, 5.5, 1H), 1.85 (dd, J ) 11.5, 3.8,
1
3
H), 1.82-1.60 (m, 6H), 1.42-1.34 (m, 2H), 1.4 (t, J ) 7.1,
ments at 7.3, 6.0%, 2.77, 3.9%, 1.82, 10.6%, 1.40, 7.63%;
C
H), 1.19 (s, 3H), 0.70 (s, 3H); 13C NMR (100 MHz, CDCl
)
3
APT 141.8 (quat), 139.9 (quat), 128.9 (tertiary), 127.96 (ter-
tiary), 126.2 (quat), 126.1 (tertiary), 57.8 (tertiary), 49.4 (quat),
41.1 (secondary), 39.9 (secondary), 26.9 (secondary), 23.8
(secondary), 22.6 (secondary), 18.5 (primary), 13.8 (primary);
IR (neat, ATR) 3026, 2923, 2844, 1601, 1495, 1453, 769, 699
66.7, 150.9, 129.4, 128.3, 126.1, 60.7, 42.8, 38.3, 36.3, 32.7,
8.8, 28.3, 26.9, 25.4, 24.7, 14.4, 12.7, 9.0 IR (neat) 2952, 1718,
273, 1105; MS (EI) 298, 283, 253, 122, 93 (base peak), 77;
+
HRMS (EI) 298.1936 obsd, 298.1933 calcd for C20
H
26
O
2
M .
-
1
cm ; MS (EI) 226, 211 (base peak), 183, 169, 91; HRMS (EI)
1
,4-Dimethyl-anti-2-(trimethylsilylmethylene)spiro-
+
2
26.1729 obsd, 226.1722 calcd for C17H22 M .
[
bicyclo[2.1.0]pentane- 5,1′-cyclopentane] (Trimethyl-
silylmethylhousane 11). Diazene 17 (0.593 g, 2.25 mmol)
was dissolved in 50 mL of dry THF in a 125 mL round-
Acid-Catalyzed Test Reaction. To ensure that the prod-
ucts were not formed via an acid-promoted rearrangement of
J. Org. Chem, Vol. 70, No. 12, 2005 4605

Gerken et al.
the housane, a control experiment was conducted. Phenyl-
housane 10a (216 mg, 0.95 mmol) was added totwo drops (∼58
mg, 0.4 mmol) of 70% aqueous perchloric acid in 11 mL of 10:1
3H); 13C APT 143.9 (quat), 143.3 (quat), 135.1 (quat), 129.0
(tertiary), 127.1 (tertiary), 53.9 (tertiary), 49.4 (secondary),
46.97 (quat), 42.2 (secondary), 28.0 (secondary), 25.3 (primary),
23.9 (secondary), 22.9 (secondary), 21.1 (primary), 12.5 (pri-
3 2 2
(v/v) CH CN/CH Cl and stirred for 20 h. It is noteworthy that
the time required for this reaction was markedly longer than
that associated with the electron-transfer processes. The
mary); MS (EI) 240, 225 (base peak), 183, 105, 91; HRMS (EI)
+
240.1871 obsd, 240.1878 calcd for C18
H
24 M .
reaction was then quenched with NaHCO
3
, filtered, and
The spectral data for isomer 26 (X ) Me) were: 1H NMR
500 MHz, CDCl ) 7.38 (m, 2H), 6.93 (m, 2H), 3.04 (dd, 1H),
concentrated in vacuo. The crude reaction mixture was chro-
matographed on silica gel with cyclohexane as the eluent. GC/
MS analysis indicated the presence of six products, two of
which corresponded to 18 and 19, the products of the electron-
transfer induced process. The product ratios, listed in the order
in which they eluted, were 55 (adduct 18): 12:3:2:4:24 (adduct
(
3
2
1
2
9
.75 (dd, 1H), 2.42 (m, 1H), 2.35 (s, 3H), 2.25 (dd, 1H), 1.90-
.79 (m, 3H), 1.66 (s, 3H), 1.65-1.45 (m, 2H), 1.40-1.15 (m,
H), 0.55 (s, 3H); MS 73 (EI) 240, 225 (base peak), 183, 105,
+
1; HRMS (EI) 240.1879 obsd, 240.1878 calcd for C18
H
24 M .
3
,7a-Dimethyl-2-(4-ethoxycarbonylphenyl)-2,4,5,6,7,7a-
hexahydro-1H-indene (p-Ethoxycarbonylphenyl bicyclo-
nonenes 25 (X ) CO Et) and 26 (X ) CO Et)). A two-necked
1
9). Neither the number of components, nor the ratio of 18 to
9, matched the results of the aminium ion promoted process.
1
2
2
3
,7a-Dimethyl-2-(4-methoxyphenyl)-2,4,5,6,7,7a-hexa-
flask equipped with a magnetic stir bar was oven-dried and
allowed to cool under nitrogen. The flask was charged via a
short column of oven-dried basic alumina with a solution of
hydro-1H-indene (p-Methoxyphenylbicyclononenes 25
and 26 (X ) OMe)). p-Methoxyphenylhousane 10c (0.393 g,
1
.54 mmol) and 1 drop (∼8.5 mg, ∼0.04 mmol) of 2,6-di-tert-
0.087 g (0.29 mmol) of p-ethoxycarbonylphenylhousane 10d
butylpyridine were dissolved in 5 mL of dry methylene
chloride. This solution was passed through a short column of
oven-dried basic alumina into an argon-purged, flame-dried
and 1 drop (∼8.5 mg, ∼0.04 mmol) of 2,6-di-tert-butylpyridine
in 5 mL of distilled methylene chloride. The column was then
rinsed with 5 mL of distilled methylene chloride into the
reaction flask. The reagent tris(p-bromophenyl)aminium
hexachloroantimonate was weighed (24 mg, 0.029 mmol) and
2 2
25 mL flask. An additional 5 mL of CH Cl was then passed
through the column into the flask. Tris(4-bromophenyl)-
aminium hexachloroantimonate (0.129 g, 0.16 mmol) was
dissolved in 5 mL of CH Cl , added to the reaction flask
2 2
dropwise via cannula, and allowed to stir for 1 h. The reaction
mixture was then passed through a short column of oven-dried
basic alumina and concentrated in vacuo. In 100% petroleum
ether, the isomers displayed a nearly identical R of 0.2.
f
Nevertheless, they could be separated by painstaking chro-
matography on a 2 cm diameter column that was loaded with
2 2
dissolved in 5 mL of CH Cl . The reagent solution was added
to the reaction flask and allowed to stir at room temperature
for 4 h. The reaction mixture was then filtered through a short
column of oven-dried basic alumina and concentrated in vacuo.
The crude product was chromatographed on silica in 10% (v/
v) ethyl acetate/hexanes to yield 0.075 g (0.25 mmol, 87%) of
an inseparable mixture of bicyclononenes 25 (X ) CO
2
Et) and
2
6 (X ) CO Et) which were characterized together. Neverthe-
2
2
0 cm of silica gel and eluted with 100% petroleum ether; the
less, it was easy to differentiate between the isomers because
the benzylic proton that is also allylic resonates at lower field.
products were collected in ∼8 mL fractions.
The spectral data for the major product, 25 (X ) OMe),
The spectral data for the major product 25 (X ) CO
2
Et)
) δ 7.95 (m, 2H), 7.20 (m,
H), 4.37 (q, J ) 7.1, 2H), 3.70 (dd, J ) 10.4, 3.7, 1H), 2.41
1
were: H NMR (400 MHz, CDCl
3
) δ 7.07 (m, 2H), 6.83 (m,
1
were: H NMR (400 MHz, CDCl
3
2
2
1
H), 3.80 (s, 3H), 3.62 (dd, J ) 10.3, 4.0, 1H), 2.41 (m, 1H),
.17 (dd, J ) 13.4, 10.3, 1H), 2.01 (m, 1H), 1.83 (m, 1H), 1.70-
.56 (m, 3H), 1.44 (s, 3H), 1.32-1.15 (m, 2H), 1.13 (s, 3H); 13
2
(
m, 1H), 2.21 (dd, J ) 13.4, 10.4, 1H), 2.02 (m, 1H), 1.85 (m,
C
1
7
H), 1.70 (m, 1H), 1.62-1.55 (m, 1H), 1.42 (s, 3H), 1.38 (t, J )
NMR (100 MHz, CDCl
3
) 157.6, 143.1, 139.0, 128.5, 128.1,
13
.1, 3H), 1.13-1.35 (m, 3H), 1.10 (s, 3H); C NMR (100 MHz,
113.3, 55.2, 53.4, 49.3, 46.8, 42.1, 27.9, 25.3, 23.8, 22.8, 12.3.
3
CDCl ) 166.8, 152.5, 144.2, 129.7, 127.9, 127.9, 127.2, 60.7,
The spectral data for the minor product, 26 (X ) OMe),
54.3, 49.1, 47.1, 42.1, 27.9, 25.2, 23.8, 22.8, 14.4, 12.3; IR (neat)
2929, 2857, 1716, 1275; MS (EI) 298, 283 (base peak), 253,
122, 108; HRMS (EI) 298.1927 obsd, 298.1933 calcd for
1
were: H NMR (400 MHz, CDCl
3
) δ 7.18 (m, 2H), 6.85 (m,
2
1
1
3
1
1
H), 3.81 (s, 3H), 3.03 (dd, J ) 11.1, 7.9, 1H), 2.63-2.74 (m,
+
H), 2.41-2.44 (m, 1H), 2.30 (dd, J ) 14.6, 7.5, 1H), 1.77-
.83 (m, 1H), 1.56-1.69 (m, 2H), 1.18-1.48 (m, 3H), 1.44 (s,
C H O M .
2
0
26
2
3
,7a-Dimethyl-1-(trimethylsilylmethyl)-2,4,5,67,7a-
H), 0.55 (s, 3H); 13C NMR (100 MHz, CDCl
3
) 157.81, 139.7,
hexahydro-1H-indene (Trimethylsilylmethylbicyclo-
nonenes 27 and 28). A round-bottomed flask equipped with
a magnetic stir bar was oven-dried and allowed to cool under
nitrogen. The flask was fitted with a short column of oven-
dried basic alumina that was rinsed with 10 mL of dichloro-
methane into the flask. The flask was then charged via the
alumina column with a solution of 2.360 g of trimethylsilyl-
methyl housane 11 (10 mmol) and 5 drops (∼42.5 mg, ∼0.02
mmol) of 2,6-di-tert-butylpyridine in 10 mL of distilled meth-
ylene chloride. The column was then rinsed with 2 × 5 mL of
33.6, 129.5, 125.9, 113.1, 56.8, 49.0, 40.9, 39.9, 26.6, 23.6, 22.4,
8.3, 13.6; IR (neat) 2926, 2856, 1509, 1244; MS (EI) 256, 241
(
base peak), 213, 199, 121; HRMS (EI) 256.1831 obsd, 256.1827
+
calcd for C18
,7a-Dimethyl-2-(4-methylphenyl)-2,4,5,6,7,7a-hexa-
hydro-1H-indene and 3,7a-Dimethyl-1-(4-methylphenyl)-
,4,5,67,7a-hexahydro-1H-indene (p-Methylphenylbicy-
clononenes 25 and 26 (X ) Me)). p-Methylphenylhousane
0b (0.480 g, 1.80 mmol) and 1 drop (∼8.5 mg, ∼0.04 mmol)
of 2,6-di-tert-butylpyridine were dissolved in 5 mL of CH Cl
H24O M .
3
2
1
2
2
CH
Cl
followed by 2 × 10 mL of CH
2 2
Cl into the flask. In an
2
2
and passed through a short oven-dried basic alumina column
into an argon-purged, oven-dried 25 mL round-bottomed flask.
An additional 5 mL of CH Cl was then passed through the
2 2
column into the flask. Tris(4-bromophenyl)aminium hexa-
chloroantimonate (161 mg, 0.20 mmol) was dissolved in 5 mL
2 2
of CH Cl , added to the reaction flask dropwise via cannula,
and allowed to stir for 1.5 h. The reaction mixture was then
passed through a short column of oven-dried basic alumina
and concentrated in vacuo. The crude product was chromato-
graphed in cyclohexane on silica to yield two products in 0.457
g (1.71 mmol) total yield (95% mass balance).
addition funnel, 0.795 g (0.97 mmol) of tris(p-bromophenyl)-
aminium hexachloroantimonate was dissolved in 30 mL of
CH Cl and added dropwise. A slight exotherm was observed.
2 2
The mixture was allowed to react for 90 min before being
filtered through oven-dried basic alumina. The reaction mix-
ture was concentrated in vacuo to afford 1.94 g (8.2 mmol; 82%)
obtained as an inseparable mixture of the two isomers 27 and
2
8, this despite numerous attempts to effect a separation (silica
gel, alumina, C18 reversed-phase HPLC). Consequently, the
data reported below corresponds to that of the mixture. The
presence of diastereotopic protons renders the assignment of
proton NMR resonances particularly difficult. Only in cases
where our confidence level is high have specific assignments
been made. The product distribution was determined by
capillary GC, and by GC mass spectral analyses. Of particular
The spectral data for isomer 25 (X ) Me) were: 1H NMR
(
1
1
3
500 MHz, CDCl ) 7.15-7.05 (m, 4H), 3.65 (dd, 1H), 2.43 (dm,
H), 2.38 (s, 3H), 2.21 (dd, 1H), 2.05 (td, 1H), 1.84 (m, 1H),
.71-1.59 (m, 4H), 1.46 (s, 3H), 1.39-1.19 (m, 2H), 1.18 (s,
4606 J. Org. Chem., Vol. 70, No. 12, 2005

Rearrangements of Housane Cation Radicals
FIGURE 1. Rearrangement of methylhousane. Relative energies are in kcal/mol. QCISD/6-31G(d)//MP2/6-31G(d) are from ref
23, while the other values were computed by us.
2
note is the near identity of the mass spectral fragmentation
patterns, thereby confirming the isomeric nature of products.
The GCMS data can be found in the Supporting Information.
0.9806,22 are included in all reported energies. <S > values
for all structures were between 0.75 and 0.78. Despite a bias
1
,23,24
toward delocalized structures,
the B3LYP method has
The spectral data for the mixture of compounds 27 and 28
been used successfully to study the structures and reactions
1
25
were: H NMR (400 MHz, CDCl
2
3
3
) δ 2.58-2.51 (m, 1H), 2.46-
of several different types of cation radicals. In addition, to
.28 (m, 2H), 2.05-1.86 (m, 4H), 1.79-1.69 (m, 3H), 1.64 (s,
H; allylic methyl of one isomer), 1.62 (s, 3H, allylic methyl of
assess the proficiency of this method for describing rearrange-
ment reactions of housane cation radicals in particular, we
examined the rearrangement of the simple methyl-substituted
housane shown in Figure 1. Reactions of this species were
examined previously using MP2 calculations for geometry
optimizations and QCISD single point calculations to assess
relative energies of minima and transition state struc-
the other isomer), 1.60-1.52 (m, 5H), 1.36-1.05 (m, 9H), 1.22
s, 3H, bridgehead methyl of one isomer), 0.91 (s, 3H, bridge-
head methyl of the other isomer), 0.76-0.46 (m, 2H), 0.13 (s,
H, trimethylsilyl unit of one isomer), 0.12 (s, 9H, trimethyl-
(
9
1
3
silyl unit of the other isomer); C NMR (100 MHz, CDCl
3
) δ
42.2, 141.3, 139.7, 132.5, 48.6, 48.5, 45.3, 44.4, 44.3, 42.7, 41.0,
1
,23,26
1
3
1
7
tures.
Figure 1 shows the relative energies of various
4.6, 28.6, 27.6, 26.6, 24.5, 24.3, 23.6, 23.6, 23.2, 17.6, 16.8,
4.1, 12.2, 0.21, 0.32; MS (EI) 236 (molecular ion), 221, 147,
6 73 (base peak); HRMS (EI) 236.1969 obsd, 236.1960 calcd
minima and transition structures involved in the reactions of
methyl-substituted housane computed at three different levels
of theory. While differing from each other in magnitude, all
three methods reproduce the experimentally observed prefer-
+
for C15
H
28Si M .
2
3,27
Computational Methods. All calculations were performed
ence for hydrogen migration toward the methyl group.
2
0
with GAUSSIAN03. Geometries were optimized using un-
21
restricted B3LYP/6-31G(d). All stationary points were char-
acterized by analysis of their vibrational frequencies. Zero-
point energy corrections from these calculations, scaled by
Acknowledgment. We are grateful to the National
Science Foundation under Grant No. CHE-0200855, for
its support of this research. We are also grateful to
Professors Barry Carpenter (Cornell) and Jeehiun Lee
(20) Frisch, M. J.; Trucks, G. W.; Schlegel, H. B.; Scuseria, G. E.;
(
Rutgers) for helpful comments. A.B.P. is particularly
Robb, M. A.; Cheeseman, J. R.; Montgomery, J. A., Jr.; Vreven, T.;
Kudin, K. N.; Burant, J. C.; Millam, J. M.; Iyengar, S. S.; Tomasi, J.;
Barone, V.; Mennucci, B.; Cossi, M.; Scalmani, G.; Rega, N.; Petersson,
G. A.; Nakatsuji, H.; Hada, M.; Ehara, M.; Toyota, K.; Fukuda, R.;
Hasegawa, J.; Ishida, M.; Nakajima, T.; Honda, Y.; Kitao, O.; Nakai,
H.; Klene, M.; Li, X.; Knox, J. E.; Hratchian, H. P.; Cross, J. B.; Bakken,
V.; Adamo, C.; Jaramillo, J.; Gomperts, R.; Stratmann, R. E.; Yazyev,
O.; Austin, A. J.; Cammi, R.; Pomelli, C.; Ochterski, J. W.; Ayala, P.
Y.; Morokuma, K.; Voth, G. A.; Salvador, P.; Dannenberg, J. J.;
Zakrzewski, V. G.; Dapprich, S.; Daniels, A. D.; Strain, M. C.; Farkas,
O.; Malick, D. K.; Rabuck, A. D.; Raghavachari, K.; Foresman, J. B.;
Ortiz, J. V.; Cui, Q.; Baboul, A. G.; Clifford, S.; Cioslowski, J.; Stefanov,
B. B.; Liu, G.; Liashenko, A.; Piskorz, P.; Komaromi, I.; Martin, R. L.;
Fox, D. J.; Keith, T.; Al-Laham, M. A.; Peng, C. Y.; Nanayakkara, A.;
Challacombe, M.; Gill, P. M. W.; Johnson, B.; Chen, W.; Wong, M. W.;
Gonzalez, C.; Pople, J. A. Gaussian 03, Revision C.02; Gaussian, Inc:
Wallingford, CT, 2004.
pleased to acknowledge the ARC Program (Academic
Research Consortium), the UC Leads (University of
California Leadership Excellence through Advanced
(22) Scott, A. P.; Radom, L. J. Phys. Chem. 1996, 100, 16502-16513.
(23) Blancafort, L.; Adam, W.; Gonzalez, D.; Olivucci, M.; Vreven,
T.; Robb, M. A. J. Am. Chem. Soc. 1999, 121, 10583-10590.
(24) The Supporting Information for ref 23 points out a specific
example of this bias.
(25) (a) Saettel, N.; Oxgaard, J.; Wiest, O. Eur. J. Org. Chem. 2001,
1429-1439. (b) Wiest, O.; Oxgaard, J.; Saettel, N. J. Adv. Phys. Org.
Chem. 2003, 38, 87-110. (c) Fokin, A. A.; Schreiner, P. R. Chem. Rev.
2002, 102, 1551-1593.
(26) Several recent reviews point to the bias of MP2 calculations
towards localized structures as well as its tendency to suffer from spin
2
5
(
21) (a) Becke, A. D. J. Chem. Phys. 1993, 98, 5648-5652. (b) Becke,
contamination.
A. D. J. Chem. Phys. 1993, 98, 1372-1377. (c) Lee, C.; Yang, W.; Parr,
R. G. Phys. Rev. B 1988, 37, 785-789. (d) Stephens, P. J.; Devlin, F.
J.; Chabalowski, C. F.; Frisch, M. J. J. Phys. Chem. 1994, 98, 11623-
(27) The first B3LYP activation barrier is negative due to the zero
point energy corrections; in short, B3LYP predicts that the potential
energy surface in the vicinity of the housane cation radical and the
1,3-diyl cation radical is very flat.
1
1627.
J. Org. Chem, Vol. 70, No. 12, 2005 4607

Gerken et al.
Supporting Information Available: 1H and 13C NMR
spectral data, along with coordinates and energies of computed
structures. This material is available free of charge via the
Internet at http://pubs.acs.org.
Degrees) Program, and the CAMP Program (California
Alliance for Minority Participation in Science, Engi-
neering and Mathematics) for supporting his under-
graduate research participation in this effort. D.J.T. and
S.C.W. are grateful to UC Davis and the National
Computational Science Alliance for support.
JO050087C
4608 J. Org. Chem., Vol. 70, No. 12, 2005