Factors affecting the selection of products from a photochemically generated singlet biradical

Article abstract of DOI:10.1039/b501422k

The chemistries of a monoradical of the ultrafast "radical-clock" type and a structurally related singlet biradical, generated by Norrish type II photochemistry, are compared. The monoradical is found to undergo the characteristic ring-opening reaction of its class at about 10¹⁰ s^-1 at room temperature. However, the singlet biradical shows no evidence of the analogous ring-opening reaction. The contrasting chemistry is traced not to a fundamental difference in electronic structure of the two intermediates, but rather to a steric interaction that the biradical alone would have to suffer during the ring opening. Although the magnitude of the steric hindrance is small (estimated 15-20 kJ mol^-1), it is enough to shut down the reaction, because the biradical has other facile product-forming reactions available. The Royal Society of Chemistry 2005.

Full text of DOI:10.1039/b501422k

View Article Online / Journal Homepage / Table of Contents for this issue
A R T I C L E
Factors affecting the selection of products from a photochemically
generated singlet biradical
David A. Broyles and Barry K. Carpenter*
Department of Chemistry and Chemical Biology, Baker Laboratory, Cornell University,
Ithaca, New York, 14853-1301, USA. E-mail: bkc1@cornell.edu; Fax: +1 607 255 4137;
Tel: +1 607 255 3826
Received 28th January 2005, Accepted 21st March 2005
First published as an Advance Article on the web 7th April 2005
The chemistries of a monoradical of the ultrafast “radical-clock” type and a structurally related singlet biradical,
generated by Norrish type II photochemistry, are compared. The monoradical is found to undergo the characteristic
ring-opening reaction of its class at about 1010 s−1 at room temperature. However, the singlet biradical shows no
evidence of the analogous ring-opening reaction. The contrasting chemistry is traced not to a fundamental difference
in electronic structure of the two intermediates, but rather to a steric interaction that the biradical alone would have
to suffer during the ring opening. Although the magnitude of the steric hindrance is small (estimated 15–20 kJ mol−1),
it is enough to shut down the reaction, because the biradical has other facile product-forming reactions available.
would provide information that could be used in combination
Introduction
with the observed product ratios from the Norrish type II
Singlet biradicals are among the most challenging organic
photochemistry to estimate the rate constants for the various
reactive intermediates to study. Contributing to the problems are
reactions of the biradical 4.
the typically short lifetimes of these species,1–3 the difficulties that
they present to electronic-structure theory,4–8 and the apparent
susceptibility of their chemistry to nonstatistical dynamical
effects.3,9–17 Elucidation of their reactivity patterns would be
facilitated if one could draw analogies between the chemistries
of biradicals and monoradicals, because the latter class of
intermediates has proven much more amenable to experimental
investigation.18–20
It is probably safe in most cases to draw parallels between the
chemistries of monoradicals and triplet biradicals,8,21–24 because
the Pauli exclusion principle ensures that the unpaired elec-
trons in triplets behave more-or-less independently.25 However,
for singlet biradicals an assumed analogy with monoradical
chemistry is potentially more hazardous. There certainly exist
examples where singlet biradicals do behave much like their
doublet-state cousins,26–29 but there are also cases where the
Scheme 1 The radical (2) and biradical (4) whose reactions are
compared.
chemistry is quite different.30,31 Theory can provide a guide to
the factors that are likely to make the analogy valid or not. For
example, singlet biradicals whose ground states involve large
Results and discussion
contributions from ionic electron configurations are unlikely to
behave like monoradicals.5,32
Synthesis of precursor 1
However, the chemistry of singlet biradicals has proven to be
The synthesis of PTOC ester 1 was accomplished as summarized
susceptible to so many, often subtle, factors that its complete
in Scheme 3.
and confident prediction by theory remains an unattained goal.
Perhaps the only thing worthy of comment in this synthesis
For that reason, it would appear useful to add to the list of
is the fact that the reduction of ester 12 to an alcohol, followed
experimental studies where a comparison of singlet biradicals
later by its reconversion to the carboxylic acid oxidation state,
with structurally related monoradicals has been made. This
was undertaken in order to exploit the known beneficial effect of
paper reports the results from one such investigation.
an alcohol on the Simmons–Smith and related cyclopropanation
The system chosen for study was one in which the mono-
reactions.36
radical would be of the ultrafast radical-clock type (2).33,34
The analogous biradical (4) was generated by Norrish type
Synthesis of ketone 3
II photochemistry.35 The reactive intermediates and their pre-
cursors (1 and 3, respectively) are shown in Scheme 1. Our
expectation was that radical 2 should undergo very rapid
ring opening to isomer 5, which could then be trapped in
an intermolecular reaction. By analogy, biradical 4 might be
expected to ring open to isomer 6, in addition to giving the
normal ring closure and fragmentation products (7 and 8,
respectively) expected from Norrish type II photochemistry.35
Biradical 6, in turn, could plausibly give products 9 and/or
10 (Scheme 2). If the chemistry occurred as anticipated, then
measurement of the rate constant for ring opening of radical 2
The obvious synthetic routes to ketone 3, involving some kind
of cyclopropanation of an indene derivative, were tried first,
but were universally unsuccessful. No cyclopropanation method
could be made to work on indenes 16a or 16b. Possibly a
Simmons–Smith reagent, or one of its variants, could have been
persuaded to cyclopropanate 16c, but the effect of the hydroxyl
ligand that leads to the desirable rate and yield effects on such
reactions, also leads to a syn stereoselectivity,36,37 which, in this
case, would have given the wrong diastereomer of the product
(Scheme 4), and so the reaction was not attempted.
T h i s j o u r n a l i s
T h e R o y a l S o c i e t y o f C h e m i s t r y 2 0 0 5
O r g . B i o m o l . C h e m . , 2 0 0 5 , 3 , 1 7 5 7 – 1 7 6 7
1 7 5 7
©

View Article Online
Scheme 4 Unsuccessful synthetic approaches to ketone 3. 16a: X =
OH; 16b: X = CH3.
Scheme 2 Anticipated reactions of radical 2 and biradical 4.
Scheme 5 Successful synthetic route to ketone 3.
The key to this approach was the photochemical ring con-
traction reported by Kakiuchi and coworkers.38 The yield of the
transformation with naphthol 17 as reactant was found to be
rather variable (30–60%), possibly because the work-up in the
presence of large amounts of AlCl3 was quite difficult. Various
methods for converting ketone 19 into the desired product were
explored, including Wittig reaction with 1-methoxyethylidene–
triphenylphosphorane,39,40 but none turned out to afford a yield
better than that from the sequence shown in Scheme 5. The
desired product, 3, and its diastereomer, 21, were formed in
about a 1 : 1 ratio, but could be separated by chromatography.
Independent synthesis of potential photochemical products 10
In an effort to aid in the detection of small amounts of
compounds 10 that may be formed from the photolysis of ketone
3, an independent synthesis was undertaken (Scheme 6). The
availability of compounds 10 also permitted a check of their
stability to the photochemical reaction conditions.
Again, the sequence depended for its first step on some
known photochemistry – this time converting the benzyne/
dimethylfulvene Diels–Alder adduct, 22, to polycyclic isomer
23 by a di-p-methane rearrangement.41
Determination of the activation parameters for ring opening of
radical 2
The activation parameters for ring opening of radical 2
were determined by a procedure pioneered by Newcomb and
coworkers.42 It involved generation of the radical in the presence
of varying concentrations of PhSeH over a range of temper-
atures, and measurement of the ratio of ring-closed and ring-
opened products, 26 and 27, respectively (Scheme 7).
Scheme 3 Synthesis of the radical precursor 1.
The failures of the cyclopropanation strategies prompted
the pursuit of an entirely different approach, which did lead
eventually to the desired product, 3. The synthesis is summarized
in Scheme 5.
1 7 5 8
O r g . B i o m o l . C h e m . , 2 0 0 5 , 3 , 1 7 5 7 – 1 7 6 7

View Article Online
trans-2-phenylcyclopropylcarbinyl radical has reported activa-
tion parameters33 for ring opening that translate to DH‡
=
10.9 kJ mol−1, DS‡ = +11.0 J mol−1 K−1 at 298 K. The differences
of the present results from these reference values seem somewhat
surprising. First, the r orbital of the cyclopropane C–C bond to
be broken in radical 2 appears to have a poor overlap with the
p orbitals of the benzene ring, which might be expected to raise
the enthalpic barrier with respect to that found for ring opening
of the trans-2-phenylcyclopropylcarbinyl radical. Second, it is
not obvious why the activation entropy would be positive for
the ring opening of trans-2-phenylcyclopropylcarbinyl radical
but negative for the ring opening of 2.
In an effort to understand these results, electronic structure
calculations were undertaken. Optimized geometries and har-
monic vibration frequencies for radical 2 and its ring-opening
transition structure were determined at the UB3LYP/cc-pVDZ
level.43 In addition, single-point CASPT2(9,9)/cc-pVDZ calcu-
lations were carried out,44 from a CASSCF(9,9) reference wave-
function involving the singly-occupied orbital of the radical, the
r and r* orbitals of the breaking cyclopropane bond, and the p
and p* orbitals of the benzene ring. The G3(MP2,B3) model45,46
was also investigated, but the UMP2 wavefunction for the ring-
opening transition state, which forms part of this composite
method, was so highly spin contaminated (ꢀS2ꢁ = 1.42) that the
activation parameters computed by this model were judged to
be unreliable.
Scheme 6 Synthetic route to potential photochemical products 10.
The UB3LYP optimized transition structure showed a 31◦
dihedral angle between the planes C3–C2–C7 and C6–C1–H14
(Fig. 1). Thus, the orbital overlap in the transition structure
(which is where it matters for the kinetics) made the problem
appear less severe than it did from inspection of the reactant
geometry. Nevertheless, this analysis did still predict that there
should be an overlap penalty on the barrier for ring opening,
even if its magnitude may be modest. That expectation was
born out in the directly computed activation enthalpies. The
UB3LYP calculations found DH‡ = 11.2 kJ mol−1, DS‡
=
−13.3 J mol−1 K−1, while the CASPT2 calculations (using
UB3LYP geometries and harmonic vibration frequencies) gave
DH‡ = 13.4 kJ mol−1. The activation enthalpy was therefore
higher than that calculated for the ring opening of trans-2-
phenylcyclopropylcarbinyl radical at the same levels of theory
(UB3LYP: DH‡ = 4.2 kJ mol−1, DS‡ = −8.8 J mol−1 K−1;
CASPT2: DH‡ = 7.2 kJ mol−1). There is consequently a
clear disagreement between theory and experiment for the
ring opening of radical 2, as revealed in Fig. 2. Neither of
the computational predictions fits the data points within the
estimates of random error.
Scheme 7 Reactions used in the determination of activation parameters
for ring opening of radical 2.
Since the reaction was run with a large stoichiometric excess
of PhSeH, its concentration could be considered effectively
constant. Measurement of the ratio of products 26 and 27
therefore gave the ratio k2[PhSeH]/k1. Since [PhSeH] was
measured, and k2 could be determined at each temperature from
the known activation parameters for reaction of PhSeH with
tertiary alkyl radicals, values for k1 could be deduced. They are
listed in Table 1.
A weighted, nonlinear least-squares fit of the Eyring equation
to these data gave the following activation parameters for ring
opening of radical 2: DH‡ = 3.6 0.5 kJ mol−1, DS‡ = −37.6
1.6 J mol−1 K−1.
Electronic structure calculations on the ring opening of radical 2
The activation enthalpy and entropy determined for the ring
opening of radical 2 are both significantly lower than those
found for other “ultrafast” radical clocks. For example, the
Table 1 Rate constants for ring opening of radical 2 as a function
of temperature. Uncertainty estimates were derived from 3–6 measure-
ments with different concentrations of PhSeH at each temperature
Fig. 1 UB3LYP/cc-pVDZ optimized transition structure for ring
opening of radical 2.
Temperature/K
k1/s−1
229
253
276
298
321
329
8.86 1.52 × 109
1.04 0.20 × 1010
1.40 0.29 × 1010
1.72 0.27 × 1010
1.92 0.20 × 1010
2.14 0.11 × 1010
Two sources of this discrepancy seem possible. The more
obvious is that the levels of electronic-structure theory used are
simply not up to the task. However, two arguments lead us not
to place all of the responsibility for the disagreement on the the-
oretical side. The first is that the models employed underestimate
O r g . B i o m o l . C h e m . , 2 0 0 5 , 3 , 1 7 5 7 – 1 7 6 7
1 7 5 9

View Article Online
Fig. 2 Comparison of theoretical (lines) and experimental (points)
estimates of the rate constant k1 as a function of temperature.
both DH‡ and DS‡ for ring opening of the trans-2-phenylcyclo-
propylcarbinyl radical, and so it seems unlikely that they would
overestimate those quantities for the apparently very similar re-
action of radical 2. The second is that the experimental activation
parameters found for the ring opening of radical 2 differ not
only quantitatively from the values directly calculated, but also
qualitatively from those expected on the basis of the physical
differences between 2 and trans-2-phenylcyclopropylcarbinyl, as
discussed above.
The other potential source of the discrepancy would obviously
be a systematic error in the experimental values. Unidentified
systematic errors are always possible, of course, but in this case
one potential problem does stand out. It is the adoption of the
published activation parameters for trapping of tertiary alkyl
radicals by phenylselenol. As the authors of that work took
pains to point out, the reaction is partly diffusion controlled,
and so the transferability of the activation parameters from one
tertiary radical to another is open to doubt, since structurally
disparate radicals may well have different diffusion constants.42
In principle, this issue could be explored by running the trapping
reactions in solvents of different viscosity, but an investigation
of that kind was deemed to be beyond the scope of the present
study.
In summary, while perfect agreement between experiment and
theory at the levels used here cannot be expected, there is reason
to believe that the agreement ought to be better than it turns out
to be. We suspect that a systematic error could have made both
the experimental activation enthalpy and activation entropy too
low. Since these effects will tend to cancel, the estimated rate
constant for ring opening of radical 2 of ∼1010 s−1 at room
temperature is probably still reasonably accurate. For example,
if one assumed the CASPT2//UB3LYP model to be accurate in
reproducing the difference in activation parameters between ring
openings of radical 2 and trans-2-phenylcyclopropylcarbinyl,
one would arrive at DH‡ = 17.1 kJ mol−1, DS‡ = +6.5 J
mol−1 K−1 for the former reaction. Although these values look
very different from the ones deduced experimentally, the two
sets lead to quite similar rate constants at 298 K: 1.37 × 1010 s−1
from the values just quoted and 1.72 0.27 × 1010 s−1 from the
experiment.
Scheme 8 Principal products from UV photolysis of ketone 3 with and
without the triplet quencher piperylene present.
reactions of that class often occur preferentially from the 3(n,p*)
states of aliphatic ketones,35 the effect of adding piperylene,
a triplet quencher, was examined. As expected, addition of
piperylene did reduce the yield of Norrish type I products,
although even at a concentration of 330 mM it could not
suppress them completely. This might indicate that some of the
Norrish type I chemistry occurred also in the singlet manifold.
With the piperylene present, a new product, identified as 8,
appeared. The failure to detect 8 in the absence of the triplet
quencher suggested that it might be unstable with respect to
direct photolysis. That was confirmed by experiment: photolysis
of 8 in the absence of piperylene led to its rapid disappearance,
apparently by Norrish type I photochemistry.
The increase in yield of the Norrish type II products 7 and 8
in the presence of piperylene indicated that they were formed
from a singlet excited state, as desired. Product 7 appeared
to be a single diastereomer. It was tentatively assigned the
structure in which the hydroxyl group was endo, by comparison
1
of its observed H NMR chemical shifts with those predicted
for the two epimers by GIAO calculations47 at the B3LYP/6-
311+G(2d,p)//B3LYP/6-31G(d) level, using the polarizable-
continuum model48 for CHCl3 as the solvent (Fig. 3).
The photochemical instability of compound 8 led to a search
for other primary photoproducts that might prove unstable
to the reaction conditions. Photolysis to low conversion did
reveal another product, which was found to disappear at longer
reaction times. It proved to be ketone 31. This product could
arise by Norrish type I chemistry, but it might also be a sec-
ondary photoproduct of the benzonorbornenols 10 (Scheme 9).
However, when the independently synthesized sample of 10 was
subjected to the photolysis conditions it was found to be stable,
and so the Norrish type II mechanism could be ruled out.
Electronic structure calculations on the reactions of singlet
biradical 4
Photochemistry of ketone 3
The photochemistry of ketone 3 revealed that the singlet birad-
ical 4 underwent only “normal” Norrish type II ring closure
and fragmentation. No products arising from the ring opening
of the cyclopropane could be detected, despite the high rate
constant for such a process deduced for the apparently related
monoradical, 2. In an effort to understand this observation,
electronic structure calculations were undertaken. Ideally for
a singlet biradical one would use a multireference method,
Irradiation of ketone 3, as a solution in degassed pentane,
was carried out with a 450 W medium-pressure mercury lamp
and a Pyrex filter. Reaction was essentially complete after 1 h.
Four products could be identified (Scheme 8). Minor products,
totaling about 4%, were present in individual quantities too
small to permit characterization. Since hydrocarbons 28–30 were
apparently products of a Norrish type I reaction, and since
1 7 6 0
O r g . B i o m o l . C h e m . , 2 0 0 5 , 3 , 1 7 5 7 – 1 7 6 7

View Article Online
Fig. 3 Comparison of experimental 1H NMR chemical shifts (ppm
downfield from TMS in CDCl3) for compound 7 with those calculated
for the two diastereomers. Calculations were of the GIAO type, con-
ducted at the B3LYP/6-311+G(2d,p)//B3LYP/6-31G(d) level, using a
polarizable-continuum model for CHCl3 solvent. Chemical shifts for the
individual protons of each methyl group were averaged to represent the
fast-exchange limit.
Scheme 9 Norrish type I and type II mechanisms for the formation of
ketone 31.
The calculated activation enthalpies were found to reproduce,
at least qualitatively, most of the experimental observations on
the Norrish type II photochemistry of ketone 3. Specifically, the
closure to alcohol 7 was found to be favored, with the cleavage
leading eventually to product 8 competing as a minor channel,
since it faces a small barrier. Of obvious relevance to the purpose
of the study, was the discovery that the ring opening of biradical
4 to biradical 6 should face a much larger barrier than that
found for the formally analogous reaction of monoradical 2.
The barrier facing 4 was calculated to be large enough that
it could explain the virtual absence of products derived from
this pathway in the photochemistry of ketone 3. The question
then became whether the difference in ring-opening barriers
for the monoradical and biradical derived from fundamentally
distinct electronic structures of the two intermediates, or whether
some alternative explanation could be found. An answer became
apparent as soon as the geometries of the rotomeric transition
structures for ring opening of 4 were examined in detail. They
are shown in Fig. 4. One can see that in both cases, the
isopropylidene substituent is forced into a steric interaction with
the vicinal hydroxyethylidene substituent, because of the orbital
such as CASPT2, which can provide a reasonably balanced
treatment of dynamic and nondynamic electron correlation.
Unfortunately, biradical 4 proved to be too big for CASPT2
to be employed with a credible active space and basis set.
Consequently, a broken-symmetry UB3LYP/6-31 + G(d) model
was used. This approach for treating singlet biradicals has
problems, partly associated with the unphysical mixing of singlet
and triplet (plus higher multiplicity) spin states. However, there
exist classes of biradical for which the UB3LYP model is likely to
be appropriate; these have been described by Cremer.49 We judge
the biradical 4 to be within the group for which this model should
provide a reasonable description. The primary intent of the
calculations was to find out whether the reluctance of biradical
4 to undergo cyclopropane ring opening could be understood.
Two different rotomers of biradical 4 could be found as local
minima on the potential energy surface (PES). However, when
zero-point and thermal corrections were included, the enthalpic
barriers for their ring-closures to alcohols 7 disappeared. The
rotomer of 4 that would close to the epimer of 7 with the exo
hydroxyl was found to be lower in enthalpy by 2.5 kJ mol−1,
but this is well within the likely error of the calculations
and so cannot be taken as a reason to reverse the tentative
product assignment based on NMR chemical shifts (Fig. 3).
Furthermore, when the PES is as flat as it seems to be in the
vicinity of biradical 4, one must consider the possibility that
nonstatistical dynamical effects could perturb the product ratios
significantly. For example, the parent tetramethylene biradical in
its singlet state seems to be strongly influenced by such effects.50
The two rotomers of biradical 4 were found to be connected to
distinct transition states for cyclopropane ring opening and for
the cleavage that leads eventually to product 8. The reactions and
computed activation enthalpies are summarized in Scheme 10.
=
alignment required to form the new exocyclic C C bond. Simple
MM2 molecular-mechanics estimates suggest that these steric
interactions are worth 15–20 kJ mol−1, which is just about the
difference between the computed activation enthalpies for ring
opening of 2 and 4.
Experimental
General
All reagents were purchased from Aldrich Chemical Company,
Acros, or Fluka. All solvents were obtained from Fisher
O r g . B i o m o l . C h e m . , 2 0 0 5 , 3 , 1 7 5 7 – 1 7 6 7
1 7 6 1

View Article Online
in situ. Diisopropylamine and triethylamine were dried by
distillation from and stored over potassium hydroxide. Dimethyl
sulfoxide was distilled under reduced pressure from calcium
˚
hydride and was either stored over more calcium hydride or 4 A
˚
molecular sieves. Solvents were sometimes dried over 4 A molec-
ular sieves. Piperylene (90% cis–trans mixture) was distilled from
sodium borohydride before use and stored under Ar.
1H NMR spectra were obtained with a Varian XL-200
spectrometer, a Varian MERCURY-300 spectrometer Varian
INOVA-400 spectrometer, a Varian INOVA-500 spectrometer,
or a Varian INOVA-600. 13C NMR spectra were obtained with
a Varian MERCURY-300 spectrometer or a Varian INOVA-500
spectrometer operating at 75.4 or 125.7 MHz respectively.
Infrared spectra were obtained with a Nicolet Impact 410 FT-
IR spectrometer. Samples were either prepared as thin films on
salt plates or via a flattened potassium bromide mix (KBr pellet).
Mass spectrometry data were obtained with GC/MS using
a Hewlett Packard 5890 chromatograph equipped with a
0.25 mm × 30 m DB-5 capillary column and a Hewlett Packard
5970 series mass selective detector.
Analytical gas chromatography was carried out with either (1)
a Hewlett Packard 5880A gas chromatograph with a 0.25 m ×
15 m RTX-5 ((5%-phenyl)methylpolysiloxane) column, a flame
ionization detector, and a Hewlett Packard 5880 GC plotter-
integrator; or (2) a Hewlett-Packard 6890 gas chromatograph
having a 0.32 mm× 30 m HP-1 (crosslinked methyl polysiloxane)
column, a flame ionization detector, and using a computer with
HP ChemStation software (revision 6.03).
Hydrogenations were carried out in high-pressure bottles on
a Parr Model 3911 High Pressure Reactor.
Photolysis reactions were run using an Ace-Hanovia 450 W
medium-pressure mercury arc lamp, cooled with a Pyrex (or
quartz) jacket. The PTOC ester photolyses were run using a
standard sun lamp with a tungsten filament at 250 W.
Thin-layer chromatography was performed using glass plates
coated with a 0.25 mm layer of silica gel 60 F254. Plates
were visualized by visual inspection, UV light (254 nm) and
then either iodine or stain (most commonly anisaldehyde). The
anisaldehyde stain was prepared from 5 mL p-anisaldehyde,
90 mL 95% ethanol, 2 mL glacial acetic acid, and 5 mL
concentrated sulfuric acid. TLC plates were then dipped in the
stain and heated with a heat gun.
Scheme 10 UB3LYP/6-31+G(d) computed activation enthalpies
(kJ mol−1) for the ring opening of two rotomers of biradical 4. Rotomer
4x was found to be lower in enthalpy than 4n by 2.5 kJ mol−1
.
All column chromatography followed the flash chromatog-
raphy protocol,51 using silica gel 60 (0.04–0.063 mm, 230–400
mesh, E. M. Merck). Chromatographic solvents, as well as those
for workup procedures were used without purification.
Melting points were determined using a Thomas-Hoover cap-
illary melting point apparatus. Temperatures were uncorrected.
Purities of key compounds were confirmed by elemental
analysis. However, in the case of PTOC ester 1, its photochem-
ical lability made this impossible. Consequently the elemental
analysis was conducted on its immediate precursor, carboxylic
acid 15.
Synthesis of PTOC ester 1
2-(2-Hydroxyindan-2-yl)-2-methylpriopionic acid ethyl ester
(11). In a 250 mL round bottom flask, flame-dried under Ar,
diisopropylamine (5.161 g, 51.0 mmol) was placed in dry THF
(50 mL). The solution was cooled to −78 ◦C. nBuli (7.168 mL,
1.6 M, 51.0 mmol) was slowly added and the solution was stirred
for 15 min. Ethyl isobutyrate (5.808 g, 50.0 mmol) was added and
the mixture was stirred for 20 min, at which point 2-indanone
(6.608 g, 50.0 mmol) was added. The solution was stirred for
2 h at the low temperature. After warming the solution to 0 ◦C,
HCl (3 M) was carefully added to quench the reaction. The
organic layer was then washed twice with a saturated aqueous
sodium chloride solution and was dried with magnesium sulfate.
GC analysis showed forming of a new peak thought to be the
product, but also showed around 50% of the starting indanone
Fig. 4 UB3LYP/6-31+G(d) optimized transition structures for the
cyclopropane ring opening of exo and endo rotomers of biradical 4.
Scientific and were used as received except as noted below. When
needed dry, acetonitrile and dichloromethane were dried by
distillation from calcium hydride. THF was dried by distillation
from potassium benzophenone ketal, and diethyl ether was dried
by distillation from sodium benzophenone ketal. Methanol was
dried by distillation from magnesium methoxide, generated
1 7 6 2
O r g . B i o m o l . C h e m . , 2 0 0 5 , 3 , 1 7 5 7 – 1 7 6 7

View Article Online
remaining. The solvent was removed and the mixture was carried
onto the next step without purification.
1H), 2.26 (d of m, J = 8.5 Hz, 1H), 2.87 (d, J = 16.5 Hz, 1H),
3.24 (d, J = 16.5 Hz, 1H), 3.53 (s, 2H), 7.07 (m, 2H), 7.14 (m,
1H), 7.25 (m, 1H).
2-(1H-Inden-2-yl)-2-methylpropionic acid ethyl ester (12).
The material from the previous reaction was dissolved in pyri-
dine (100 mL) and the solution was cooled to 0 ◦C. Phosphorus
oxytrichloride (15 g) was then added. The solution turned dark
and a precipitate formed. The solution was heated to reflux for
2 h, and the color darkened further. The mixture was poured
onto 300 g of ice and then placed into a separatory funnel with
diethyl ether. An emulsion ensued that did not clear until it had
been filtered through a Buchner funnel. Extractions with diethyl
ether and pentane, with plenty of washings with water eventually
led to a solution that was dried with magnesium sulfate. The
solvent was removed and the material was purified by column
chromatography (in a 100 mm column, 100% hexane for 200 mL,
10% diethyl ether for 200 mL, 20% diethyl ether for 200 mL and
then the remainder with 33% diethyl ether, the material with the
higher Rf was collected) to yield compound 12 (6.7 g, 58% over
two steps).
13C NMR (125 MHz, CDCl3): d 20.14, 22.16, 22.33, 25.95,
34.16, 35.94, 37.01, 70.86, 122.90, 125.10, 125.22, 125.81, 141.86,
147.25.
2-(1a,6-Dihydro-1H-cycloprop[a]inden-6a-yl)-2-methylpropionic
acid (15). The alcohol 14 (5.2 g, 25.7 mmol) was dissolved in
◦
acetone (150 mL) and the solution was cooled to 0 C. Jones’
reagent (20 mL) was added dropwise (prepared by dissolving
◦
5.44 CrO3 in 20 mL H2O, cooling to 0 C and adding 4.5 mL
of H2SO4). The mixture was stirred for 4 h total until the
starting material and the intermediate aldehyde had completely
vanished by TLC. The solvent was removed and water added,
and the mixture was extracted three times with EtOAc. The
organic layers were combined and washed twice with water and
once with saturated aqueous sodium chloride. After drying with
Na2SO4 and filtering, the solvent was removed to give 5.64 g
of an orange–white solid. The solid was subjected to column
chromatography (100 mm column, 1 : 1 hexanes–EtOAc, the
material was not very soluble and required a lot of solvent to
dissolve), enough to leave behind an orange band and give 5.05 g
of a light, yellow solid. The material was then recrystallized
from benzene–hexanes, gathering up to 5 crops of crystals and
a total recovery of 4.12 g (38% from 2-indanone) of 15 as an
off-white solid with a melting point of 126–128 ◦C.
1H NMR (500 MHz, CDCl3): d 1.23 (t, J = 7.0 Hz, 3H), 1.53
(s, 6H), 3.42 (s, 2H), 4.13 (q, J = 7.0 Hz, 2H), 6.68 (s, 1H), 7.14
(t, J = 7.5 Hz, 1H), 7.24 (t, J = 8.0 Hz, 1H), 7.32 (d, J = 7.0 Hz,
1H), 7.40 (d, J = 7.5 Hz, 1H).
13C NMR (125.7 MHz, CDCl3): d 14.13, 25.70, 38.67, 44.65,
60.87, 120.63, 123.48, 124.29, 126.32, 126.59, 143.13, 144.59,
152.51, 175.94.
1H NMR (500 MHz, CDCl3): d 0.22 (dd, J = 4, 5 Hz, 1H),
1.18 (s, 3H), 1.24 (s, 3H), 1.36 (dd, J = 5, 7.5 Hz, 1H), 2.44 (d
of m, J = 8.5 Hz, 1H), 2.98 (d, J = 17.5 Hz, 1H), 3.17 (d, J =
17.5 Hz, 1H), 7.07 (m, 3H), 7.24 (m, 1H), 11.5 (vbr, 1H).
13C NMR (125.7 MHz, CDCl3): d 20.26, 22.73, 23.12, 27.78,
34.95, 38.20, 43.46, 123.00, 125.19, 125.27, 125.89, 141.68,
146.86, 183.39.
2-(1H-Inden-2-yl)-2-methylpropan-1-ol (13). Ester 12 (6.7 g,
29.1 mmol) was dissolved in dry diethyl ether (150 mL) and
the solution was cooled to −78 ◦C. A solution of DIBALH
(1 M in toluene, 78 mL, 78 mmol) was slowly added and the
resulting mixture was stirred for 1 h. The solution was warmed
to room temperature and stirred overnight (20 h total), after
which it was carefully poured into aqueous NaOH (180 mL,
3 M). The mixture was extracted three times with diethyl ether.
The aqueous layer was then acidified using 3 M HCl and then it
was extracted once more with diethyl ether. These fractions were
dried with magnesium sulfate and the solvent removed to give a
white solid (5.34 g, 97%) that was only very slightly yellow. The
material was pure by GC and TLC and so was carried onto the
next step without purification.
IR: br –COOH, 1695.6 cm−1.
Calc. for C14H16O2: C, 77.75; H, 7.46. Found: C, 77.71; H,
7.55%.
2-(1a,6-Dihydro-1H-cycloprop[a]inden-6a-yl)-2-methylpropionic
acid 2-thioxo-2H-pyridin-1-yl ester (1). This material was
prepared immediately before use. The carboxylic acid 15
(1.000 g, 4.624 mmol) was dissolved in benzene (30 mL) along
with 10 drops of DMF. Under Ar, oxalyl chloride (0.994 mL,
11.574 mmol) was added via syringe. The solution bubbled
initially and was stirred for 1 h. The solvent was then removed
by vacuum to give a syrup. Benzene (20 mL) was added to
redissolve this material. The material from this point becomes
light sensitive, and all precautions must be taken to keep the
material from direct light. All flasks and beakers were coated
with aluminium foil. This solution of the acid chloride was
added to a mixture of benzene (20 mL), DMAP (0.059 g,
0.48 mmol) and 2-mercaptopyridine 1-oxide sodium salt
(0.716 g, 4.8 mmol). The reaction was stirred for 2 h in the
dark at which point it was washed with saturated aqueous
sodium bicarbonate, brine, dried with magnesium sulfate and
filtered. The solvent was removed and the material was purified
by column chromatography (1 : 1 methylene chloride–ether),
yielding 620 mg of a gummy, yellow foam after solvent removal.
1H NMR (500 MHz, CDCl3): d 0.36 (dd, J = 4.0, 5.5 Hz, 1H),
1.41 (s, 3H), 1.48 (s, 3H), 1.49 (m, 1H), 2.56 (d of m, J = 8.5 Hz,
1H), 3.09 (d, J = 17.0 Hz, 1H), 3.23 (d, J = 16.5 Hz, 1H), 6.51
(dt, J = 2.0, 7.0 Hz, 1H), 7.05–7.15 (m, 4H), 7.25 (m, 1H), 7.64
(m, 1H).
Mp 62–64 ◦C.
1H NMR (300 MHz, CDCl3): d 1.25 (s, 6H), 1.42 (br s, 1H),
3.38 (s, 2H), 3.54 (s, 2H), 6.65 (s, 1H), 7.10–7.26 (m, 2H), 7.31
(d, J = 6 Hz, 1H), 7.40 (d, J = 7.5 Hz, 1H).
13C NMR (75 MHz, CDCl3): d 24.78, 38.16, 39.12, 71.64,
120.35, 123.50, 124.11, 126.36, 126.99, 143.07, 144.93, 155.12.
2-(1a,6-Dihydro-1H-cycloprop[a]inden-6a-yl)-2-methylpropan-
1-ol (14). Methylene chloride (300 mL) was placed in a 500 mL
round bottom flask with a magnetic stir bar and the vessel
was cooled using an ice–NaCl bath. Diethyl zinc (14.9 mL,
145.5 mL) was slowly added dropwise to the solvent followed
by the addition of diiodomethane (11.74 mL, 145.5 mmol). The
solution was stirred for 30 min and a white precipitate formed.
The olefin from the previous reaction (13, 5.34 g, 28.4 mmol)
was added in methylene chloride (15 mL) and the mixture
was warmed to room tempera◦ture and stirred overnight. The
solution was then cooled to 0 C and very carefully quenched
by the addition of a solution of saturated aqueous ammonium
chloride. More water was added along with some dilute HCl
(3 M). The aqueous layer was extracted three times with
methylene chloride, and then the combined organic layers
were washed successively with saturated aqueous solutions of
sodium carbonate, sodium bicarbonate and sodium chloride.
The solution was dried with magnesium sulfate and filtered.
GC analysis showed only 3% of the starting olefin remained.
Removal of the solvent yielded 5.4 g of a slightly green oil.
1H NMR (500 MHz, CDCl3): d 0.12 (t, J = 4.5 Hz, 1H), 0.91
(s, 3H), 0.93 (s, 3H), 1.34 (dd, J = 4.5, 8.5 Hz, 1H), 1.48 (br s,
13C NMR (125.7 MHz, CDCl3): d 20.29, 22.51, 23.40, 28.27,
35.12, 38.36, 44.21, 123.23, 125.18, 125.58, 126.17, 133.15,
137.38, 137.63, 141.23, 146.17, 157.29.
Synthesis of 1-(6a-isopropyl-1,1a,6,6a-tetrahydrocycloprop[a]-
inden-6-yl)ethanone (3)
1-Hydroxynaphthalene-2-carboxylic acid methyl ester. In
dry THF (60 mL) was placed 1-hydroxy-2-naphthoic acid
O r g . B i o m o l . C h e m . , 2 0 0 5 , 3 , 1 7 5 7 – 1 7 6 7
1 7 6 3

View Article Online
(20.01 g, 106.3 mmol), and to this solution was added lithium
hydroxide monohydrate (4.46 g, 106.3 mmol). This mixture
was allowed to stir for 30 min under an argon atmosphere, at
which point dimethyl sulfate (10 mL, 106 mmol) was added
slowly via syringe. The solution was stirred for 1 h and then
heated at reflux for 3 h. Some of the THF was removed by
distillation (about 35 mL) and then a saturated aqueous sodium
bicarbonate solution (40 mL) was added. The mixture was
placed in a separatory funnel with more diethyl ether and H2O
and the aqueous layer was extracted four times with diethyl ether.
The diethyl ether layers were combined, dried with magnesium
sulfate, filtered, and the solvent removed to give 21.5 g of a peach
solid. This material was dissolved in 25 mL benzene, applying
heat to dissolve all of the material. White crystals formed upon
cooling, and these were removed by filtration (they were mostly
starting material). The solvent was removed on the solution
to give 20.62 g of the product (95.9% yield, 99.4% yield with
recovery of starting material).
agitated for 1 h, the flask recharged to 50 psi, and then allowed
to shake 30 min to 1 h more. The solution was then filtered
through Celite to remove the Pd/C and give a yellow solution.
The solvent was removed on the combined batches to give 22.55 g
of a red-colored oil of 2-isopropylnaphthalen-1-ol (typical yield
from three steps, 75–85%).
1H NMR (200 MHz, CDCl3): d 1.35 (d, J = 6.8 Hz, 6H), 3.32
(sept., 1H), 5.2 (br s, 1H), 7.32–8.08 (m, 6H).
3-Chloromethyl-2-isopropylindan-1-one (18). In a typical se-
quence, the reaction was run in 24/40 joint-size traps (roughly
100 mL volume). Each tube was filled with methylene chloride
(ca. 30 mL) and aluminium trichloride (21.5 g) and a stir bar.
While on ice, a solution of 2-isopropylnaphthalen-1-ol (22.55 g,
121.1 mmol) in methylene chloride (45 mL) was slowly added to
the vials (in this case, split in three, 15 mL of the solution added
to each). The solutions often started out black–green, but ended
up a black–red by the end of the photolysis. The solutions were
swished around to help mix in the aluminium trichloride, but
often a large amount of solid was left on the bottom. Each tube
was then subjected to three cycles of freeze–pump–thaw. Finally,
the tubes were photolyzed under vacuum for 11 days with a Hg
medium-pressure lamp. After that time, the tubes were cooled
on ice and the solutions were very carefully quenched with H2O,
adding diethyl ether when necessary. The quenched mixtures
were combined and the aqueous layer was extracted five times
with diethyl ether. The combined organic layers were dried with
magnesium sulfate, filtered and the solvent removed to give 24 g
of a dark, maroon sludge. This residue was purified by Kugelrohr
distillation under vacuum to give 14.5 g of a translucent yellow
oil. This oil was then subjected to column chromatography to
give 10.8 g of 3-chloromethyl-2-isopropylindan-1-one (18) as a
slightly yellow oil (90% pure by GC, 36%). Typical yields ranged
from 30–60%.
Mp 72–73.5 ◦C, (lit.52 76–77 ◦C).
1H NMR (400 MHz, CDCl3): d 1.54 (br s, 1H), 3.98 (s, 3H),
7.27 (d, J = 8.8 Hz, 1H), 7.51 (t, J = 7.2 Hz, 1H), 7.59 (t, J =
7.2 Hz, 1H), 7.75 (d, J = 8.8 Hz, 2H), 8.40 (d, J = 8 Hz, 1H).
2-(1-Hydroxy-1-methylethyl)naphthalen-1-ol. A 3-L three-
neck flask was equipped with a reflux condenser, a mechanical
stirring rod and an addition funnel, and the entire apparatus was
flame-dried under an argon atmosphere. The flask was charged
with methylmagnesium bromide (3 M in diethyl ether, 540 mL,
1.62 mol). Stirring was started while a solution of 1-hydroxy-2-
naphthoic acid methyl ester (30.0 g, 148 mmol) in 500 mL dry
diethyl ether was added dropwise (in two batches). The mixture
stirred for 6 h and then was quenched by the addition of dilute
HCl (3 M, 500 mL) and finally concentrated HCl until the salts
dissolved. The mixture was placed in a 2 L separatory funnel
and the aqueous layer was washed three times with diethyl ether.
The organic layers were combined, dried with magnesium sulfate
and filtered. The mixture was carried onto the next step without
purification or removal of solvent.
1H NMR (400 MHz, CDCl3): d 0.83 (d, J = 6.8 Hz, 3H), 1.05
(d, J = 7.2 Hz, 3H), 2.37 (dsept, J = 4, 6.8 Hz, 1H), 2.55 (dd,
J = 3.6, 4 Hz, 1H), 3.52 (m, J = 2.8 Hz, 1H), 3.80 (m, 2H), 7.42
(t, 1H), 7.60 (m, 2H), 7.73 (d, 1H).
1H NMR (200 MHz, CDCl3): d 1.74 (br s, 1H), 1.74 (s, 6H),
7.1–7.7 (m, 6H).
13C NMR (100 MHz, CDCl3): d 18.0, 20.0, 29.6, 42.6, 54.9,
56.7, 123.1, 125.3, 128.1, 134.6, 137.3, 153.2, 206.4.
2-Isopropenylnaphthalen-1-ol. A solution containing the
starting material (usually not purified from the previous
step, in this case, approximately 148 mmol of 2-(1-hydroxy-1-
methylethyl)naphthalen-1-ol) had the solvent volume adjusted
to about 400 mL diethyl ether. Sulfur trioxide pyridine complex
(100. g, 628 mmol) was added and the solution was stirred for
8 h, monitoring the reaction by TLC. Concentrated HCl (about
500 mL) was added to quench the reaction and the mixture was
poured into a separatory funnel with diethyl ether and H2O. The
aqueous layer was extracted three times with diethyl ether, and
the combined organic layers were then washed with dilute HCl
(1 M), dried with magnesium sulfate, and filtered to give a
dark yellow solution. This solution was usually carried onto the
next step (see below) without further purification, though some
solvent was often removed to reduce the volume. The solvent
was removed during a smaller-scale reaction to give a yellow
oil (98% for that small-scale reaction). The alcohol could be
dehydrated with phosphorus pentoxide, but the solution had to
be thoroughly washed to avoid slowing the hydrogenation step.
1H NMR (400 MHz, CDCl3): d 2.09 (s, 3H), 5.17 (m, 1H),
5.43 (m, 1H), 6.31 (s, 1H, –OH), 7.15–8.19 (m, 6H).
6a-Isopropyl-1a,6a-dihydro-1H-cycloprop[a]inden-6-one (19).
A 250 mL round-bottomed flask with a stir bar was flame
dried under an argon atmosphere. It was then charged with dry
THF (70 mL) and 1,1,1,3,3,3-hexamethyldisilazane (15.5 mL,
73.5 mmol) and cooled to −78 ◦C. A solution of nBuLi (1.6 M in
hexanes, 45.9 mL, 73.5 mmol) was added slowly and the solution
was stirred for 40 min at the low temperature. A solution of the
chloride (18, 10.8 g, 48.5 mmol) in 40 mL dry THF was then
added by syringe, whereupon the solution turned bright yellow
and then darkened. The reaction was stirred at −78 ◦C for 1.5 h,
then it was warmed to room temperature and stirred for 1.2 h.
The reaction was then quenched with H2O and 3 M HCl, and
the mixture was poured into a separatory funnel where it was
extracted three times with diethyl ether. The combined diethyl
ether layers were washed with 1 M HCl, then a saturated aqueous
sodium bicarbonate solution, dried with magnesium sulfate and
filtered. The solvent was removed to give a yellow–orange liquid
which was then subjected to column chromatography to give
8.27 g of 19 (90% pure by GC, 82% yield).
1H NMR (200 MHz, CDCl3): d 1.03 (d, J = 7.0 Hz, 3H), 1.10
(d, J = 7.0 Hz, 3H), 1.26 (dd/t, J = 3.6, 4.2 Hz, 1H), 1.48 (dd,
J = 4.2, 7.2 Hz, 1H), 2.25 (septet, J = 6.8 Hz, 1H), 2.68 (dd,
J = 3.4, 7.2 Hz, 1H), 7.21–7.62 (m, 4H).
2-Isopropylnaphthalen-1-ol. A solution containing the start-
ing material (not purified from the previous reaction, but
containing roughly 148 mmol of 2-isopropenylnaphthalen-1-ol
in about 800 mL of diethyl ether and THF) was split into three
batches for hydrogenation (each about 250–300 mL). Each batch
was prepared by placing 0.450 g Pd/C (10%) in the reaction
bomb, adding the solution of 2-isopropenylnaphthalen-1-ol and
then charging the H2 reactor at about 55 psi. The bomb was
6-Ethyl-6a-isopropyl-1,1a,6,6a-tetrahydrocycloprop[a]inden-6-
ol. A flask flame-dried under argon atmosphere was charged
with ethylmagnesium bromide (3 M, 45 mL, 130 mmol) and
compound 19 (8.27 g, 44.4 mmol) in dry diethyl ether (25 mL)
was added slowly via syringe. The solution was stirred for
2 h and then it was carefully quenched on ice with H2O and
1 7 6 4
O r g . B i o m o l . C h e m . , 2 0 0 5 , 3 , 1 7 5 7 – 1 7 6 7

View Article Online
1 M HCl. The aqueous layer was extracted five times with
diethyl ether, and the combined organic layers were washed
with a saturated aqueous sodium bicarbonate solution, dried
with magnesium sulfate, and filtered. The solvent was removed
via rotary evaporation and the yellow oil was subjected to
column chromatography to give 6-ethyl-6a-isopropyl-1,1a,6,6a-
tetrahydrocycloprop[a]inden-6-ol (7.78 g, 73%).
syringe and the solution was stirred for 5 min, warmed to
room temperature and stirred for 15 min. The solution had
a cloudy, yellow color. Water was added (20 mL) to dissolve
the solid, and the aqueous layer was washed three times with
methylene chloride. The combined organic layers was washed
with a saturated aqueous sodium chloride solution, dried with
magnesium sulfate, and filtered. The solvent was removed
to give a dark yellow liquid, which was subjected to column
chromatography (30 : 1 pentane–ether) to give 630 mg of a
mixture of 3 and 21 (64%), which could be separated by a
second column chromatography.
1H NMR (400 MHz, CDCl3): d 0.16 (t, J = 3.2 Hz, 1H), 0.53
(t, J = 7.6 Hz, 3H), 0.70 (d, J = 6.8 Hz, 3H), 0.93 (dd, J =
4 Hz, 1H), 1.13 (d, J = 6.8 Hz, 3H), 1.5 (br s, 1H), 2.0 (diast m,
J ∼7 Hz, 2H), 2.40 (m, 1H), 2.41 (m, J ∼7 Hz, 1H), 7.1–7.2 (m,
4H).
1H NMR of 3 (400 MHz, CDCl3): d 0.00 (t, J = 3.8 Hz, 1H),
0.73 (d, J = 6.8 Hz, 3H), 1.07 (d, J = 6.4 Hz, 3H), 1.14 (dd, J =
4.8, 8.2 Hz, 1H), 1.78 (s, 3H), 2.01 (septet, J = 6.6 Hz, 1H), 2.49
(m, J = 3.0 Hz, 1H), 3.98 (s, 1H), 7.01–7.07 (m, 2H), 7.15 (t, J =
7.4 Hz, 1H), 7.30 (d, J = 7.2 Hz, 1H).
6-Ethylidene-6a-isopropyl-1,1a,6,6a-tetrahydrocycloprop-[a]-
indene (20). 6-Ethyl-6a-isopropyl-1,1a,6,6a-tetrahydrocyclo-
prop[a]inden-6-ol (7.87 g, 36.4 mmol) was dissolved in methylene
chloride (175 mL), and to this was added distilled triethylamine
(36.9 g, 363.8 mmol), and 4-dimethylaminopyridine (0.178 g,
1.5 mmol). The reaction was stirred under argon for 10 min.
Then a solution of methanesulfonyl chloride (11.26 mL,
146 mmol) in methylene chloride (30 mL) was very slowly added
to the stirring mixture; the solution quickly turned yellow. After
the addition was complete, the reaction was stirred for 30 min,
and then quenched by the addition of water. The aqueous layer
was extracted three times with pentane, the combined organic
layers washed with very dilute HCl, dried with magnesium
sulfate and filtered. Removal of the solvent gave a yellow oil
that was subjected to column chromatography (100% pentane),
collecting the bright-yellow band to give 6.95 g of 20, both E
and Z isomers, as a yellow oil (90% pure by GC, 87% yield).
13C NMR of 3 (125.7 MHz, CDCl3): d 17.7, 18.2, 21.7, 26.7,
27.0, 28.1, 35.8, 65.6, 123.2, 125.5, 125.7, 127.4, 139.6, 149.3,
196.4.
1H NMR of 21 (400 MHz, CDCl3): d 0.70 (dd, J = 3.2, 4.8 Hz,
1H), 0.94 (d, J = 6.8 Hz, 3H), 0.99 (d, J = 6.8 Hz, 3H), 1.12
(dd, J = 4.8, 8.2 Hz, 1H), 1.80 (septet, J = 6.8 Hz, 1H), 2.21
(dd, J = 3.6, 8.2 Hz, 1H), 2.24 (s, 3H), 4.32 (s, 1H), 7.03 (d, J =
7.2 Hz, 1H), 7.08 (t, J = 7.4 Hz, 1H), 7.14 (t, J = 7.4 Hz, 1H),
7.23 (d, J = 7.2 Hz, 1H).
13C NMR of 21 (100 MHz, CDCl3): d 19.7, 20.1, 20.7, 27.6,
29.6, 32.2, 36.3, 61.1, 123.2, 125.3, 125.6, 127.1, 139.9, 147.2,
209.8.
Photolysis of ketone 3
1
Major isomer: H NMR (400 MHz, CDCl3): d 0.45 (t, 1H),
0.96 (d, 6H), 2.10 (d, 3H), 2.82 (septet, 1H), 6.13 (q, 1H).
Minor isomer: H NMR (400 MHz, CDCl3): d 0.63 (t, 1H),
0.90 (d, 6H), 2.01 (d, 3H), 3.15 (septet, 1H), 5.84 (q, 1H).
The photolysis studies on the Norrish-II radical clock precursors
were first accomplished by creating a solution of 3 in pentane
around 7 mM. The solution was divided into Pyrex tubes (0.5–
1.0 mL) that already had one end sealed; the tubes had a 6 mm
outer diameter and a 4 mm inner diameter. The tubes were then
subjected to three freeze–pump–thaw cycles and finally sealed
under vacuum. Using tape, the sealed tubes were attached to
the cooling jacket of a Hg-arc lamp and allowed to irradiate for
various lengths of time.
As the studies progressed, piperylene was added as a triplet
quencher to reduce the amount of Norrish-I products forming.
Piperylene came as a 90% tech grade (remainder was cyclopen-
tene), and was distilled from sodium borohydride before use.
The concentration of piperylene was varied in the range 30–
330 mM. The principal new compounds formed were as follows:
1
1-(6a-Isopropyl-1,1a,6,6a-tetrahydrocycloprop[a]inden-6-yl)-
ethanol. The alkene mixture, 20, (6.95 g, 35.0 mmol) was
dissolved in dry THF (300 mL) under an argon atmosphere,
and a solution of BH3·THF (1 M, 36 mL, 36 mmol) was slowly
added by syringe. The solution turned clear and colorless
(from bright yellow) after 1 h. The reaction was allowed to
stir overnight (18 h total, since early quenching of an aliquot
revealed 27% starting material around after 3 h). A solution
was prepared of sodium hydroxide (3 M, 50 mL) and hydrogen
peroxide (30% in water, 50 mL), and this solution was slowly
added to the cooled (by ice) reaction vessel. The mixture was
then warmed to room temperature and stirred for 3 h, put in a
separatory funnel, and dilute HCl was added to dissolve some
of the solid that had formed. The aqueous layer was extracted
four times with diethyl ether (checking the extracts by TLC
for any remaining spots), and the combined organic layers
were dried with magnesium sulfate and filtered. The solvent
was removed to give a slightly yellow, very thick oil, which
was then subjected to column chromatography to give 1-(6a-
isopropyl-1,1a,6,6a-tetrahydrocycloprop[a]inden-6-yl)ethanol
(5 g, ∼70%) a clear, colorless, thick oil. It was difficult to
determine exact purity and yield by GC owing to the presence
of four diastereomers. NMR spectra could not be assigned, for
the same reason.
1-(3-Isopropyl-1,2-dihydronaphthalen-1-yl)ethanone
(31).
1H NMR (400 MHz, CDCl3): d 1.09 (d, J = 7.0 Hz, 3H), 1.10
(d, J = 7.0 Hz, 3H), 1.98 (s, 3H), 2.42 (septet, J = 7.0 Hz, 1H),
2.46 (ddd, J = 1.5, 9.5 Hz, 1H), 2.71 (dd, J = 3.5, 16.5 Hz, 1H),
3.59 (dd, J = 3.5, 7.5 Hz, 1H), 6.16 (s, 1H), 7.02–7.23 (m, 4H).
1-[2-(Isopropylidenecyclopropyl)phenyl]propan-2-one (8). 1H
NMR (400 MHz, CDCl3): d 1.63 (t, 1H), 1.85 (s, 3H), 1.89 (s,
3H), 2.15 (s, 3H), 2.46 (m, 1H), 3.87 (s, 2H), 6.97–7.2 (m, 4H).
MS: m/z 214 (<1), 199 (22), 171 (12), 157 (27), 129 (34), 115
(29), 43 (100%).
Alcohol 7. 1H NMR (500 MHz, CDCl3): d 0.48 (dd, J = 4.0,
5.3 Hz, 1H), 0.85 (s, 1H), 0.97 (dd, J = 5.5, 8.5 Hz, 1H), 1.30 (s,
3H), 1.33 (s, 3H), 1.50 (br s, 1H), 2.30 (d of m, 1H), 3.57 (s, 1H),
7.09–7.32 (m, 4H).
MS: m/z 214 (<1), 199 (1), 171 (17), 155 (12), 141 (17), 129
(100), 115 (20), 86 (57%).
1-(6a-Isopropyl-1,1a,6,6a-tetrahydrocycloprop[a]inden-6-yl)-
ethanone (3 + 21). A 100 mL flask with a stir bar was
flame-dried under an argon atmosphere, charged with oxalyl
chloride (2 M in methylene chloride, 4.62 mL, 9.25 mmol), and
cooled to −78 ◦C. Dimethyl sulfoxide (1.13 mL, 18.49 mmol)
was added via syringe, taking care not to let it freeze in the
needle-tip, and this mixture was stirred for 2 min. A solution
3-Isopropyl-1,2-dihydronaphthalene
(30). 1H
NMR
of
1-(6a-isopropyl-1,1a,6,6a-tetrahydrocycloprop[a]inden-6-
(400 MHz, CDCl3): d 1.09 (d, J = 6.8 Hz, 6H), 2.23 (t,
J = 8.0 Hz, 2H), 2.41 (septet, J = 6.8 Hz, 1H), 2.77 (t, J ∼8 Hz,
2H), 6.20 (s, 1H), 7.01–7.14 (m, 4H).
yl)ethanol (1.00 g, 4.62 mmol) in CH2Cl2 (10 mL) was slowly
added via syringe, and the reaction was stirred for 15 min.
Distilled triethylamine (5.15 mL, 36.98 mmol) was added via
MS: m/z 172 (43), 157 (89), 142 (19), 129 (100), 115 (39%).
O r g . B i o m o l . C h e m . , 2 0 0 5 , 3 , 1 7 5 7 – 1 7 6 7
1 7 6 5

View Article Online
1
The other principal product, 2-isopropylnaphthalene, is a
known compound.53
Major isomer: H NMR (400 MHz, CDCl3): d 1.06 (s, 3H),
1.57 (s, 3H), 1.78 (s, very fine splitting, 3H), 1.8 (d, blends in
with 1.78, 1H), 2.74 (d of m, J = 13.6 Hz, 1H), 3.01 (dm, J =
4.2 Hz, 1H), 3.56 (br s, 1H), 7.05–7.16 (m, 4H).
Synthesis of alcohols 10
MS: m/z 214 (7), 199 (2), 171 (47), 129 (100), 115 (12).
Minor isomer: H NMR (400 MHz, CDCl3): d 1.34 (s, 3H),
1.47 (s, 3H), 1.74 (s, very fine splitting, 3H), 2.53 (dm, J ∼20 Hz,
1H), 3.06 (d, J = 3.2 Hz, 1H), 3.61 (br s, 1H), 7.20–7.27 (m, 4H);
one proton hidden.
Hydrocarbon 23. Hydrocarbon 2241 (0.250 g, 1.4 mmol) was
dissolved in acetone (18 mL), and Ar was bubbled through
the solution in a Pyrex test tube for 1 min. The solution was
irradiated on a medium-pressure Hg lamp (with a quartz jacket)
for 6 h, at which point GC analysis showed the reaction to be
complete. The solvent was removed and the residue was purified
via pipet column chromatography (2 : 1 pentane–methylene
chloride) giving 23 (201 mg, 80%) as a slightly milky liquid.
1H NMR (400 MHz, CDCl3): d 1.49 (s, 3H), 1.52 (s, 3H), 2.58
(m, 1H), 2.77 (t, J = 5.2 Hz, 1H), 3.56 (m, 1H), 3.74 (br t, 1H),
7.02–7.1 (m, 3H), 7.34 (d, J = 7.2 Hz, 1H).
1
Conclusions
The studies reported here suggest that the radical 2 ring opens
at about 1010 s−1 at 298 K. Nevertheless, the apparently similar
singlet biradical, 4, generated by Norrish type II photochemistry,
showed no evidence of ring opening at all. The theoretical work
that has accompanied this experimental investigation suggests
that the distinction arises not from a fundamental difference in
electronic structure between the monoradical and biradical, but
rather from the fact that the transition state for ring opening of
biradical 4 faces a steric interaction that has no counterpart
in the ring opening of 2. Although the magnitude of this
steric interaction is modest (estimated 15–20 kJ mol−1), it is
enough to shut the reaction down because the biradical has
two other product-forming channels that experience little or
no barrier. Thus one learns that even when theory leads to
an expectation of similar physical properties of monoradicals
and singlet biradicals, the chemistry can nonetheless be quite
different because singlet biradicals will often decay by very fast
unimolecular reactions that are unavailable to monoradicals.
2-Isopropylidene-1,2,3,4-tetrahydro-1,4-methanonaphthalen-
9-ol (24). The starting hydrocarbon 23 (0.025 g, 0.14 mmol)
was dissolved in 0.5 mL THF and 0.5 mL water. A small amount
of hydrochloric acid (0.5 mL, 3 M) was added and this mixture
was stirred vigorously. The reaction was monitored by TLC,
indicating the formation of two products with Rfs low on the
plate in 8 : 1 : 1 hexanes–benzene–EtOAc. The reaction was then
extracted four times with diethyl ether, dried with magnesium
sulfate and the solvent removed to give 36 mg of a residue. This
material was purified by pipet column chromatography using a
3:1 pentane-methylene chloride mixture, affording the product,
24 (26 mg total).
1H NMR (400 MHz, CDCl3): d 1.46 (s, 3H), 1.59 (br s, 1H),
1.73 (m, very fine splitting, 3H), 1.81 (d, J = 15.2 Hz, 1H), 2.47
(d of m, J = 14.2 Hz, 1H), 3.32 (s, 1H), 3.82 (m, 1H), 4.18 (m,
1H), 7.13–7.17 (m, 2H), 7.23–7.25 (m, 1H), 7.29–7.31 (m, 1H).
Acknowledgements
2-Isopropylidene-1,2,3,4-tetrahydro-1,4-methanonaphthalen-
9-one (25). A round bottom flask with stir bar was flame-dried
under Ar, cooled to −78 ◦C and then charged with oxalyl
chloride (0.130 mL, 3 M, 0.260 mmol). DMSO was added
slowly via syringe (0.037 mL, 0.519 mmol) and the solution
was stirred for 2 min. A solution of the alcohol 24 (0.026 g,
0.130 mmol) in 1 mL dry methylene chloride was added, and
the solution was stirred for 15 min. Triethylamine (0.145 mL,
1.039 mmol) was added and the solution stirred for 5 min at
low temperature, then 15 min at room temperature; the solution
was cloudy white. Water was added (3 mL) along with five
drops of HCl (3 M) to dissolve the salts. The mixture was the
extracted four times with methylene chloride and the combined
organic layers were washed with saturated aqueous ammonium
chloride and dried with magnesium sulfate. The solvent was
removed and the residue was purified using pipet column
chromatography with 1 : 1 pentane–methylene chloride as the
eluent (Rf ∼ 0.5, purple spot with anisaldehyde stain) to give
15 mg of the product 25 as a clear, colorless oil (60%).
Support of this research by the U.S. National Science Founda-
tion (grant CHE-0212149) is gratefully acknowledged.
References
1 C. P. Casey and R. A. Boggs, J. Am. Chem. Soc., 1972, 94, 6457.
2 S. De Feyter, E. W. G. Diau and A. H. Zewail, Angew. Chem., Int.
Ed., 2000, 39, 260.
3 M. B. Reyes and B. K. Carpenter, J. Am. Chem. Soc., 2000, 122,
10163.
4 L. Salem, Pure Appl. Chem., 1973, 33, 317.
5 L. Salem and C. Rowland, Angew. Chem., Int. Ed. Engl., 1972, 11,
92.
6 P. J. Hay, W. J. Hunt and W. A. Goddard, III, J. Am. Chem. Soc.,
1972, 94, 638.
7 W. T. Borden, React. Intermed. (Wiley), 1985, 3, 151.
8 A. D. Dutoi, Y. Jung and M. Head-Gordon, J. Phys. Chem. A, 2004,
108, 10270.
9 S. L. Debbert, B. K. Carpenter, D. A. Hrovat and W. T. Borden,
J. Am. Chem. Soc., 2002, 124, 7896.
10 C. Doubleday, G. Li and W. L. Hase, Phys. Chem. Chem. Phys., 2002,
4, 304.
11 M. B. Reyes, E. B. Lobkovsky and B. K. Carpenter, J. Am. Chem.
Soc., 2002, 124, 641.
12 J. E. Baldwin and E. J. Keliher, J. Am. Chem. Soc., 2002, 124, 380.
13 C. Doubleday, Jr., K. Bolton and W. L. Hase, J. Phys. Chem. A, 1998,
102, 3648.
14 D. A. Hrovat, S. Fang, W. T. Borden and B. K. Carpenter, J. Am.
Chem. Soc., 1997, 119, 5253.
15 C. Doubleday, Jr., K. Bolton, G. H. Peslherbe and W. L. Hase, J. Am.
Chem. Soc., 1996, 118, 9922.
16 B. K. Carpenter, J. Am. Chem. Soc., 1995, 117, 6336.
17 D. C. Sorescu, D. L. Thompson and L. M. Raff, J. Chem. Phys.,
1995, 102, 7910.
18 B. A. Thrush, J. Chem. Soc., Faraday Trans. 2, 1986, 82, 2125.
19 K. A. McLauchlan, Chem. Soc. Rev., 1993, 22, 325.
20 D. Beckert, Electron Paramag. Res., 2002, 18, 74.
21 M. V. Encinas, P. J. Wagner and J. C. Scaiano, J. Am. Chem. Soc.,
1980, 102, 1357.
22 P. J. Wagner, K.-C. Liu and Y. Noguchi, J. Am. Chem. Soc., 1981,
103, 3837.
1H NMR (400 MHz, CDCl3): d 1.52 (s, 3H), 1.79 (s, 3H), 2.06
(d, J = 15 Hz, 1H), 3.26 (br d, J = 15.5 Hz, 1H), 3.53 (d, J =
3.9 Hz, 1H), 4.02 (br s, 1H), 7.19–7.32 (m, 4H).
13C NMR (75.4 MHz, CDCl3): d 20.0, 20.9, 32.2, 50.1, 54.2,
121.0, 121.9, 123.6, 125.3, 126.7, 127.0, 139.5, 140.5, 185.4.
2-Isopropylidene-9-methyl-1,2,3,4-tetrahydro-1,4-methanona-
phthalen-9-ol (10). To a solution of ketone 25 in 1 mL THF
(approx. 12 mg, 0.061 mmol) was added MeMgBr (0.200 mL,
3 M). This solution was stirred overnight, whereupon it was
carefully quenched with water and a small amount of HCl
(3 M). The mixture was extracted with diethyl ether and the
combined organic layers were dried with magnesium sulfate.
Upon removal of the solvent, the residue was purified via pipet
column chromatography (1 : 1 pentane–methylene chloride, Rf
∼ 0.2 in 8 : 1 : 1 hexanes–benzene–EtOAc, greenish-blue spot
with anisaldehyde stain). Purification gave 9 mg of a colorless
oil-residue (70%), with GC and NMR analysis showing roughly
a 5 : 1 ratio of the diastereomers.
1 7 6 6
O r g . B i o m o l . C h e m . , 2 0 0 5 , 3 , 1 7 5 7 – 1 7 6 7

View Article Online
23 W. Adam, C. van Barneveld, O. Emmert, H. M. Harrer, F. Kita, A. S.
Kumar, W. Maas, W. M. Nau, S. H. K. Reddy and J. Wirz, Pure Appl.
Chem., 1997, 69, 735.
24 C. J. Cramer and J. Thompson, J. Phys. Chem. A, 2001, 105,
2091.
25 For a caution against literal acceptance of this simple notion, see:
C. F. Logan and P. Chen, J. Am. Chem. Soc., 1996, 118, 2113.
26 T. H. Peterson and B. K. Carpenter, J. Am. Chem. Soc., 1992, 114,
1496.
Schlegel, G. E. Scuseria, M. A. Robb, J. R. Cheeseman, J. A.
Montgomery, Jr., T. Vreven, K. N. Kudin, J. C. Burant, J. M. Millam,
S. S. Iyengar, J. Tomasi, V. Barone, B. Mennucci, M. Cossi, G.
Scalmani, N. Rega, G. A. Petersson, H. Nakatsuji, M. Hada, M.
Ehara, K. Toyota, R. Fukuda, J. Hasegawa, M. Ishida, T. Nakajima,
Y. Honda, O. Kitao, H. Nakai, M. Klene, X. Li, J. E. Knox,
H. P. Hratchian, J. B. Cross, V. Bakken, C. Adamo, J. Jaramillo,
R. Gomperts, R. E. Stratmann, O. Yazyev, A. J. Austin, R. Cammi,
C. Pomelli, J. W. Ochterski, P. Y. Ayala, K. Morokuma, G. A. Voth,
P. Salvador, J. J. Dannenberg, V. G. Zakrzewski, S. Dapprich, A. D.
Daniels, M. C. Strain, O. Farkas, D. K. Malick, A. D. Rabuck, K.
Raghavachari, J. B. Foresman, J. V. Ortiz, Q. Cui, A. G. Baboul,
S. Clifford, J. Cioslowski, B. B. Stefanov, G. Liu, A. Liashenko, P.
Piskorz, I. Komaromi, R. L. Martin, D. J. Fox, T. Keith, M. A. Al-
Laham, C. Y. Peng, A. Nanayakkara, M. Challacombe, P. M. W. Gill,
B. Johnson, W. Chen, M. W. Wong, C. Gonzalez, and J. A. Pople,
Gaussian, Inc., Wallingford, CT, 2004.
27 T. H. Peterson and B. K. Carpenter, J. Am. Chem. Soc., 1993, 115,
5466.
28 K. Redmond and B. K. Carpenter, J. Org. Chem., 1997, 62, 5668.
29 A. Maiti, J. B. Gerken, M. R. Masjedizadeh, Y. S. Mimieux and R. D.
Little, J. Org. Chem., 2004, 69, 8574.
30 E. D. Nelson, A. Artau, J. M. Price, S. E. Tichy, L. Jing and H. I.
Kenttaemaa, J. Phys. Chem. A, 2001, 105, 10155.
31 M. Abe, W. Adam, M. Hara, M. Hattori, T. Majima, M. No-
jima, K. Tachibana and S. Tojo, J. Am. Chem. Soc., 2002, 124,
6540.
32 K. Jug and C. Koelle, J. Phys. Chem. B, 1998, 102, 6605.
33 M. Newcomb, Tetrahedron, 1993, 49, 1151.
34 A. A. Martin-Esker, C. C. Johnson, J. H. Horner and M. Newcomb,
J. Am. Chem. Soc., 1994, 116, 9174.
44 CASPT2 calculations were carried out with the MOLCAS program
package: MOLCAS 5.4: K. Andersson, M. Barysz, A. Bernhardsson,
M. R. A. Blomberg, D. L. Cooper, T. Fleig, M. P. Fulscher, C.
˚
de Graaf, B. A. Hess, G. Karlstrom, R. Lindh, P.-A. Malmqvist,
P. Neogrady, J. Olsen, B. O. Roos, A. J. Sadlej, M. Schutz, B.
Schimmelpfennig, L. Seijo, L. Serrano-Andres, P. E. M. Siegbahn,
J. Stalring, T. Thorsteinsson, V. V. Pidmark, and P.-O. Widmark,
University of Lund, Sweden, 2002.
35 N. J. Turro, Modern Molecular Photochemistry, Ben-
jamin/Cummings, Menlo Park, California, 1978.
36 A. B. Charette and A. Beauchemin, Org. React., 2001, 58, 1.
37 M. Nakamura, A. Hirai and E. Nakamura, J. Am. Chem. Soc., 2003,
125, 2341.
45 A. G. Baboul, L. A. Curtiss, P. C. Redfern and K. Raghavachari,
J. Chem. Phys., 1999, 110, 7650.
46 L. A. Curtiss, P. C. Redfern, K. Raghavachari, V. Rassolov and J. A.
Pople, J. Chem. Phys., 1999, 110, 4703.
47 K. Wolinski, J. F. Hinton and P. Pulay, J. Am. Chem. Soc., 1990, 112,
8251.
48 B. Mennucci and J. Tomasi, J. Chem. Phys., 1997, 106, 5151.
49 D. Cremer, Mol. Phys., 2001, 99, 1899.
50 C. Doubleday, Jr., Chem. Phys. Lett., 1995, 233, 509.
51 W. C. Still, M. Kahn and A. Mitra, J. Org. Chem., 1978, 43, 2923.
52 N. J. P. Broom and P. G. Sammes, J. Chem. Soc., Perkin Trans. 1,
1981, 465.
38 K. Kakiuchi, B. Yamaguchi and Y. Tobe, J. Org. Chem., 1991, 56,
5745.
39 D. R. Coulson, Tetrahedron Lett., 1964, 3323.
40 U. Schubert and E. O. Fischer, Chem. Ber., 1973, 106, 1062.
41 W. Eberbach, P. Wuersch and H. Prinzbach, Helv. Chim. Acta, 1970,
53, 1235.
42 S.-Y. Choi, P. H. Toy and M. Newcomb, J. Org. Chem., 1998, 63,
8609.
43 UB3LYP calculations were carried out with the Gaussian 03 program
package: Gaussian 03; Rev. B.05: M. J. Frisch, G. W. Trucks, H. B.
53 L. Ernst and P. Schulz, Magn. Res. Chem., 1992, 30, 73.
O r g . B i o m o l . C h e m . , 2 0 0 5 , 3 , 1 7 5 7 – 1 7 6 7
1 7 6 7