The photo-nazarov cyclization of 1-cyclohexenyl(phenyl)methanone revisited - Trapping of the 2-oxyallyl intermediates by olefins

Article abstract of DOI:10.1002/1099-0690(200107)2001:14<2719::AID-EJOC2719>3.0.CO;2-Z

The photochemistry (> 300 nm) of 1-cyclohexenyl(phenyl)methanone (1) in the presence of various monoolefins and cyclopentadiene (9) has been investigated. While (Z)-cyclooctene and 1-pentene were unreactive, styrene, (E)-cyclooctene, and diene 9 co-reacted upon irradiation of 1. The reactions observed include [4 + 2] and [2 + 2] cycloadditions of the monoolefins to the primary photoproduct of 1 (which is the trans isomer 2), and trapping reactions of the 2-oxyallyl compound 4 formed thermally from 2 as a secondary intermediate. The trapping reactions of 4 are ene-type additions of the olefins, analogous to those observed with the prototypical 2-oxyallyl compound 18 in purely thermal reactions, and involving a hydrogen transfer from 4 to olefin to give adducts 10, 12, 13, 24, and 25. However, while 18 gives one single ene-type adduct with 9, intermediate 4 displays two regioisomeric ene-type addition modes with 9, both of which are concerted, albeit nonsynchronous. In the mode resulting in 10, it is suggested that hydrogen transfer lags behind C-C bond formation, whereas the reverse holds for the mode affording 12 and 13. For both major adducts, 10 and 12, concerted formation requires endo orientations of the two π systems.

Full text of DOI:10.1002/1099-0690(200107)2001:14<2719::AID-EJOC2719>3.0.CO;2-Z

FULL PAPER
The Photo-Nazarov Cyclization of 1-Cyclohexenyl(phenyl)methanone Revisited
؊ Trapping of the 2-Oxyallyl Intermediates by Olefins
Johannes Leitich,*[a] Ingeborg Heise,[a] Jürgen Rust,[b] and Kurt Schaffner[a]
Keywords: Ene reactions / Intermediates / Photochemistry / Reaction mechanisms
The photochemistry (Ͼ 300 nm) of 1-cyclohexenyl(phenyl)- ical 2-oxyallyl compound 18 in purely thermal reactions, and
methanone (1) in the presence of various monoolefins and involving a hydrogen transfer from 4 to olefin to give adducts
cyclopentadiene (9) has been investigated. While (Z)-cyclo- 10, 12, 13, 24, and 25. However, while 18 gives one single
octene and 1-pentene were unreactive, styrene, (E)-cyclooc- ene-type adduct with 9, intermediate 4 displays two re-
tene, and diene 9 co-reacted upon irradiation of 1. The reac- gioisomeric ene-type addition modes with 9, both of which
tions observed include [4 + 2] and [2 + 2] cycloadditions of are concerted, albeit nonsynchronous. In the mode resulting
the monoolefins to the primary photoproduct of 1 (which is
the trans isomer 2), and trapping reactions of the 2-oxyallyl
compound 4 formed thermally from 2 as a secondary inter-
mediate. The trapping reactions of 4 are ene-type additions
of the olefins, analogous to those observed with the prototyp-
in 10, it is suggested that hydrogen transfer lags behind C−C
bond formation, whereas the reverse holds for the mode af-
fording 12 and 13. For both major adducts, 10 and 12, concer-
ted formation requires endo orientations of the two π systems.
anism, the scope of these addition reactions remained un-
clear, but extension to olefins other than 9 might hold syn-
thetic promise. This situation prompted us to carry out this
investigation into the following points: (i) the definitive
structural assignment (‘‘which is which’’) of the two diaster-
eomeric adducts 12 and 13, (ii) the mechanisms of forma-
tion of 10, 12, and 13, and (iii) extension of the trapping of
4 by 9 to olefins other than 9.
Introduction
The title reaction in aprotic media proceeds essentially as
outlined in Scheme 1[1] (see Parts I[2] and II[3] of this series).
Moreover, when the reaction was conducted in the presence
of cyclopentadiene (9), the 2-oxyallyl[4] intermediate 4 was
trapped by 9 to give adducts 10, 12, and 13 (Scheme 2[5]),
in a ratio of 5:2.2:1 at room temperature.[2] Formation of
10 had precedents in the form of side reactions of the proto-
typical cyclopent-2-enylium-2-olate (18) with 9, to give 19
(Scheme 3),[6] and also with a substituted derivative of 9.[7]
The formation of 12 and 13, however, was novel. The mech-
anism of these addition reactions could not be elucidated
unequivocally in our previous work. Two alternatives were
discussed.[2] Firstly, in view of the fact that [4 ϩ 3] cycload-
dition was the dominant reaction of 18 with cisoid 1,3-di-
enes,[6,7] a primary formation of the [4 ϩ 3] cycloadduct 15
was considered as one possibility. Driven by aromatisation,
15 might subsequently have rearranged to 10, 12, and 13.
Secondly, direct formation of 10, 12, and 13 from 4 and 9
Ϫ either in a concerted fashion or in two steps by way of
the dipolar intermediate 16 or its corresponding diradical Ϫ
appeared another possibility, while still another conceivable
dipolar intermediate might be found in 17. Like the mech-
[a]
Max-Planck-Institut für Strahlenchemie,
Postfach 101365, 45413 Mülheim an der Ruhr, Germany
Fax: (internat.) ϩ 49-(0)208/306-3951
E-mail: leitich@mpi-muelheim.mpg.de
[b]
Laboratorium für Chemische Kristallographie, Max-Planck-
Institut für Kohlenforschung,
Kaiser-Wilhelm-Platz 1, 45470 Mülheim an der Ruhr, Germany
Fax: (internat.) ϩ 49-(0)208/306-2980
Scheme 1. The photo-Nazarov reaction of 1-cyclohexenyl phenyl
E-mail: lehmann@mpi-muelheim.mpg.de
ketone (1) in aprotic solution[2,3]
Eur. J. Org. Chem. 2001, 2719Ϫ2726
WILEY-VCH Verlag GmbH, 69451 Weinheim, 2001
1434Ϫ193X/01/0714Ϫ2719 $ 17.50ϩ.50/0
2719

J. Leitich, I. Heise, J. Rust, K. Schaffner
FULL PAPER
Figure 1. Molecular structure of 12 as found by X-ray crystallo-
graphy; hydrogen atoms are omitted for clarity; thermal ellipsoids
are drawn at the 50% probability level
ives of the products shown in Scheme 1 and Scheme 2. In
place of adducts 11 and 12, the deuterated products 11-d5
and 12-d5 were isolated (Scheme 4); since irradiation had
been carried to complete consumption of 1-d5, compounds
10-d5 and 13-d5 had undergone secondary photolysis
(Scheme 2). NMR spectroscopy unequivocally established
the positions of the non-aromatic D atoms in both. In ana-
logy to the formation of 11 from 10,[2] 11-d5 must have been
formed from 10-d5 under the irradiation conditions. Hence,
the position of the non-aromatic D atom in 10-d5 follows
from its position in 11-d5. At this point it is useful to re-
member that adducts 10, 12, and 13 are not interconverted
under the irradiation and workup conditions.
Scheme 2. Trapping of the 2-oxyallyl intermediate 4 with cyclo-
pentadiene (9)[2] and photoreactions of adducts 10[2] and 12; struc-
tures 15؊17 are conceivable (though rejected) intermediates en
route to 10, 12, and 13
Scheme 3. The prototypical addition of cyclopentadiene (9) to the
2-oxyallyl compound 18[3,9]
Results and Discussion
Addition of Cyclopentadiene 9 to 4 ؊ Structural
Assignment of the Diastereomeric Adducts 12 and 13
The adduct of m.p. 38Ϫ41 °C[2] was subjected to single-
crystal X-ray analysis, which found structure 12 (Figure 1).
This structure was confirmed independently by the photo-
cyclisation of the adduct to 14 (vide infra). It is thus neces-
sary to reverse the previous Ϫ explicitly tentative Ϫ assign-
ment of the two cyclopentadiene adducts to structures 12
and 13 in ref.[2]
Scheme 4. Trapping of the oxyallyl intermediate 4-d5 with cyclo-
pentadiene (9) and subsequent photocyclization of adduct 10-d5
In order to determine relative H/D isotope effects associ-
ated with the formation of the individual photoproducts, we
irradiated 1 and 1-d5 under identical conditions in parallel
experiments, in acetonitrile and in the presence of 9, and
1
compared the relative yields of products by 400-MHz H
Irradiation of the Pentadeuteriophenyl Methanone 1-d5 in
the Presence of 9
NMR analysis of the product mixtures. In order to
minimise the secondary photoreaction Ϫ (formation of 11
Ketone 1-d5 was irradiated in acetonitrile in the presence and 11-d5) Ϫ the degree of photoconversion was kept min-
of 9 in the same manner as 1,[2] to give deuterated derivat- imal. Not surprisingly, the relative yields of the three [4 ϩ 2]
2720
Eur. J. Org. Chem. 2001, 2719Ϫ2726

Photo-Nazarov Cyclization of 1-Cyclohexenyl(phenyl)methanone
FULL PAPER
adducts of 9 with 2[2] (not shown in the Schemes) did not 1,4-dipole 20, in contrast to the concerted formation of the
differ from those of 9 with 2-d5. Thus, the H/D isotope ef- [4 ϩ 2] adducts by way of a much less polar transition state.
fects associated with the formation of these adducts are There is a similarly dramatic increase in the rate of forma-
negligible. Ring closure 2-d5 Ǟ 4-d5, however, was slower tion of 12 and 13, relative to that of 10. When the pentadeu-
than formation of the three [4 ϩ 2] adducts by a factor of teriophenyl analogue 1-d5 was irradiated in the presence of
1.3 relative to the undeuterated case. Thus, there is a notice- 9 in TFE, we obtained a result corresponding to that ob-
able secondary isotope effect associated with the ring clos- tained with 1, as was to anticipated. However, the fact that
ure 2-d5 Ǟ 4-d5. The yield of 10-d5 (including a small 12-d5 was formed in place of a 12-d4 product (lacking the
amount of 11-d5) relative to 6-d5 ϩ 8-d10 was lower than non-aromatic D atom) is not trivial. If 12 were formed by
that of 10 (including a small amount of 11) relative to 6 ϩ a reaction between 5 and 9, then 5-d5 would quickly have
8 by a factor of 3.9 (3, 5 and 7, and deuterated analogues exchanged its OD group for OH by fast proton exchange
could be neglected). As a consequence, the primary H/D with the solvent TFE and hence would have produced
isotope effect for the formation of 10 from 4 ϩ 9 must be 12-d4 upon reaction with 9. Thus, formation of 12 from 5
larger than 3.9, since the conversion 4 Ǟ Ǟ (6 ϩ 8) can and 9 is ruled out.
also be expected to involve a significant primary H/D iso-
The position of the non-aromatic D atom in 10-d5 does
tope effect. Similarly, the primary isotope effect for the not preclude the possibility that 10 might have been formed
formation of 12 from 4 ϩ 9 was found to be larger than 6.2. by way of 15. When the irradiation of 1 was carried to par-
tial conversion only, so that 15 should be protected against
light by 1 (note that excited 1 is not quenched by 1,3-dienes
such as 9), we could neither detect any olefinic H NMR
Irradiation of 1 and 1-d5 in the Presence of 9 in 2,2,2-
Trifluoroethanol and of 1 in Toluene
1
signals attributable to 15 in the crude reaction mixture nor
isolate any 15 by chromatography. A thermal rearrange-
ment, 15 Ǟ 10, being symmetry-forbidden, would have to
proceed in two steps through a 1,5-dipole or diradical. The
activation energy expected for such a reaction[8] should am-
ply suffice to render 15 thermally stable at room temper-
ature. Thus, if 15 had been formed, we should have ob-
served it. Moreover, formation of 15 should not be subject
to an H/D isotope effect as high as Ͼ 3.9, which we observe
for the formation of 10. Since direct formation of 10 from
5 and 9 can be rejected on grounds of the expected reactiv-
ity of the reactants (which would favour formation of 12
and 13 rather than 10), 10 must be formed directly from 4
and 9, and we have to address the question of whether this
reaction is concerted or two-step by way of 16 (or the cor-
responding diradical). In an investigation of the reactions
of the prototypical 2-oxyallyl compound 18 with olefins
and 1,3-dienes in TFE,[9] we concluded that the symmetry-
allowed ene-type addition exemplified by the reaction 9 ϩ
18 Ǟ 19 and dubbed ‘‘ene-type 1’’ addition, should have a
transition state very close to a 1,5-dipole (as exemplified by
16) with a small but definite concertedness. Such concer-
tedness should increase dramatically when the second
‘‘step’’ in the nonsynchronous concerted addition Ϫ the hy-
drogen transfer Ϫ involved formation of an aromatic ring
(as in the formation of 10), since the more concerted the
reaction, the more it benefits from the high energy bonus
due to attaining aromaticity. Thus, we conclude that the
formation of 10 from 4 and 9 is quite strongly concerted.
The same conclusion also follows from the primary H/D
isotope effect of Ͼ 3.9. The alternative explanation for this
isotope effect, a rapidly established equilibrium between 4
ϩ 9 and 16 (or the corresponding diradical) followed by a
rearrangement 16 Ǟ 10 in the rate-determining step, can be
dismissed, since in the closely related case of 9 ϩ 18, the
In order to explore the influence of solvent polarity on
the relative rates of adduct formation, 1 was irradiated in
the presence of 9 in the nonpolar toluene and in 2,2,2-tri-
fluoroethanol (TFE), which is more powerfully ionising
than the acetonitrile previously used as solvent. In toluene
(upon 46% conversion), the following products were isol-
ated: 76.8% of the [4 ϩ 2] adducts of 9 with 2 (vide supra),
0.8% of 21, a hitherto overlooked [2 ϩ 2] adduct of 9 with
2 (Scheme 5), 1.6% of 3, 5% of 6, 2.5% of 8, 9.7% of 10,
and 1.1% of 12. Irradiation of 1 without added 9 in TFE
should lead exclusively to 5 ϩ 6, in analogy to the result
obtained in methanol,[2] due to catalysis of the reactions 2
Ǟ 4 and 4 Ǟ 5 by the hydroxy proton. In the presence of
9 in TFE under the same conditions as in toluene, com-
pound 1, after 43% conversion, gave 7.6% of the [4 ϩ 2]
adducts of 9 with 2 (vide supra), 10.2% of 21, 46.3% of the
anticipated main product 6, 22.8% of 12, and 6.6% of 13.
Remarkably, 10 and 11 were absent.
Scheme 5. [2 ϩ 2] addition of cyclopentadiene (9) to the primary
photo-Nazarov product 2
Mechanisms of Formation of 10, 12, 13, and 21
The increase by a factor of 129 in the rate of formation first step in the two-step diene addition is irreversible by a
of 21 relative to that of the [4 ϩ 2] adducts on going from very high energy margin.[9,10] As the position of the non-
toluene to TFE is in line with a formation of 21 via the aromatic D atom in 10-d5 shows, the concerted formation
Eur. J. Org. Chem. 2001, 2719Ϫ2726
2721

J. Leitich, I. Heise, J. Rust, K. Schaffner
FULL PAPER
of 10 involves endo orientations of the ‘‘nonreacting’’ CϭC oxyallyl CCO moiety of 4 are endo-oriented en route to 12
double bond in 9 and the ‘‘nonreacting’’ oxyallyl CCO moi- and exo-oriented in the case of 13 (Scheme 6).
ety in 4 (Scheme 6).
Irradiation of 10 and 12
Compound 10 on irradiation, even at high concentrations
of 9 and low concentrations of 1, is converted into 6 and
11 in a 1:1.5 ratio.[2] Compound 12 is photostable under
these conditions, but, when irradiated alone in acetonitrile,
it was converted into a complex mixture from which 6 and
14 were isolated in 30.6% and 6.5% yields, respectively. Ac-
cording to the structures of these products, both trans-
formations involve Norrish type-II 1,5-hydrogen abstrac-
tions: an expected result. The photoreaction of 12 is obvi-
ously quenched by the diene 9 and thus proceeds from the
excited triplet state, as is normally the case in Norrish type-
II reactions of phenyl ketones.[11] The photoreaction of 10,
however, is not quenched by 9. Since excited singlet states
of phenyl ketones are extremely short-lived, this result
points to a particularly rapid 1,5-hydrogen abstraction by
the excited carbonyl group of 10, whatever the spin multi-
plicity of the latter. Indeed, the energetically most favour-
able conformation of ground state 10 has the target hydro-
gen atom ideally placed for abstraction in the carbonyl
additions; the right-hand formulae, depicting open-chain dipolar plane and close to the oxygen atom. The allylic radical in
Scheme 6. Proposed transition states for the ene-type additions of
9 to 4; the left-hand formulae illustrate the concerted nature of the
resonance structures (cf. 16 and 17) significantly contributing to
the moment of formation by the hydrogen abstraction has
the transition states, illustrate the nonsynchronous nature of the
its other end placed above this plane, corresponding to an
elongated CϪC single bond between it and the other rad-
ical centre; collapse of this elongated bond gives 11.
additions
The position of the non-aromatic D atom in 12-d5 pre-
Irradiation of 1 in the Presence of Styrene ؊ Addition of
cludes the possibilities that 12 be formed through 16 (or the
corresponding diradical) or via 15. Formation of 12 from 5
and 9 has been ruled out above, so three possibilities for
the formation of 12 remain: directly from 4 and 9, either in
a concerted fashion or via the ion pair 17 (or the corres-
ponding radical pair). The very strong acceleration of the
formation of 12 and 13 relative to 10 on increasing the solv-
ent ionising power from acetonitrile to TFE points to the
route through 17 in TFE. However, the regiospecificity,
documented by the position of the non-aromatic D atom in
12-d5, rules out equivalence of the two ends of the allylic
cation in 17. It can only be reconciled with a symmetry-
allowed concerted reaction, with 17 merely approximating
the transition states of concerted but non-synchronous
Styrene to 4
Irradiation of 1 to 84.7% conversion, in acetonitrile in
the presence of 1.8  styrene, gave three [2 ϩ 2] adducts
of constitution 22 in 2.3, 8.7, and 2.1% yields (based on
1
unrecovered 1), 40.5% (according to H NMR analysis of
the reaction mixture; isolated yield 25.7%) of the adduct
24, 1.1% of an open-chain adduct [2-(2Ј-styryl)cyclohexyl
phenyl ketone], 13.7% of 6, and 18.7% of 8 (Scheme 7). The
three adducts 22 are analogous to 21; 24 is analogous to 10.
Irradiation of 1 in the Presence of (E)-Cyclooctene ؊
Addition of (E)-Cyclooctene to 4
In the presence of 1.0  (E)-cyclooctene in acetonitrile,
pathways to 12 and 13. The much stronger increase in the compound 1 gave the two DielsϪAlder [4 ϩ 2] adducts of
rates of formation of 12 and 13, relative to that of 10, with (E)-cyclooctene to 2 (6.7 and 1.3%), 10.9% of the [2 ϩ 2]
solvent-ionising power means that the degree of concer- adduct 23, 29.3% of the adduct 25, 3.3% of a cyclobutanol
tedness for the former is less than for the latter, with the formed from 25 by a Norrish type-II photocyclization, 8.1%
dipolar resonance structure contributing more to the trans- of 6, and two diastereomeric oxetanes formed by photoad-
ition state in the former case than in the latter. Inspection dition of (E)-cyclooctene to 6 (2.0 and 0.2%) (yields based
of the dipolar resonance structures (17 and 16) suggests a on unrecovered 1 at 66.1% conversion) (Scheme 7). Com-
possible cause for this: a much lower energy for 17 than for pound 23 is analogous to 21, and 25 analogous to 10, 12,
16, due to the presence of a benzene ring in the former but and 13. The mode of formation of secondary photoprod-
not in the latter. Since the 12/13 ratio, in contrast to the 12/ ucts was verified by irradiation of 25 alone, and of 6 in the
10 ratio, does not vary strongly with solvent polarity, the presence of (E)-cyclooctene. When the photoconversion of
mechanisms of formation of 12 and 13 ought to be similar, 1 was driven to 100%, the proportion of secondary
the difference being that in the transition state the ‘‘nonre- photoproducts among the products increased strongly. Ac-
acting’’ CϭC double bond site of 9 and the ‘‘nonreacting’’ cording to analytical experiments (see Table 1), the com-
2722
Eur. J. Org. Chem. 2001, 2719Ϫ2726

Photo-Nazarov Cyclization of 1-Cyclohexenyl(phenyl)methanone
FULL PAPER
cyclooctene is cleanly suprafacial, in line with a concerted
addition of (E)-cyclooctene to 4 in analogy to the formation
of 10.
Irradiation of 1 in the Presence of Other Olefins
In contrast to 9, styrene, and (E)-cyclooctene, the olefins
(Z)-cyclooctene and 1-pentene did not co-react when pre-
sent during irradiation of 1 in acetonitrile. The irradiation
of 1 in the presence of the highly reactive, electron-rich ol-
efin ethyl vinyl ether has been described by us previously.[12]
Conclusion
Similarly to what was found for the prototypical cyclo-
pent-2-enylium-2-olate (18, Scheme 3),[9] the 2-oxyallyl
compound 4, which is formed as an intermediate in the
course of the irradiation of 1, gives ene-type addition reac-
tions with 1,3-cyclopentadiene (9), styrene, and (E)-cyclooc-
tene, but not with (Z)-cyclooctene, nor with 1-pentene.
Thanks to the readier hydrogen abstraction from 4, as com-
pared to 18, in the course of this ene-type addition, the
addition is more strongly concerted with 4 than it is with
18. While the addition of 9 to 18 highly regioselectively af-
fords only 19,[9] the reaction between 4 and 9 unselectively
produces both 10 (which is analogous to 19) and 12 ϩ 13,
in what appear in each case to be concerted reactions. The
Scheme 7. Trapping of the primary photoproduct, (E)-cyclohexenyl
phenyl ketone 2, and of the oxyallyl intermediate 4 with styrene
and (E)-cyclooctene
bined yield of the [4 ϩ 2] adducts ϩ 23, relative to the com- addition producing 12 ϩ 13 differs from the one producing
bined yield of 6 ϩ 25 ϩ secondary products, was roughly 10 in that it has a much more strongly polar transition state
proportional to the concentration of (E)-cyclooctene, as in which hydrogen abstraction precedes CϪC bond forma-
was to be expected. The yield of 23 relative to that of the tion, whereas the reverse sequence holds for the formation
two [4 ϩ 2] adducts, however, did not change within error of 10, analogously to the formation of 19.[9] For the two
limits upon a fourfold increase of (E)-cyclooctene concen- major adducts 10 and 12, the addition reactions involve
tration, indicating that 23, like the [4 ϩ 2] adducts, origin- endo orientations of the two π systems.
ates from an addition of (E)-cyclooctene to 2.
In the light of the mechanism of formation of 12 as estab-
When 1-d5 was used in place of 1, a 25-d5 derivative was lished in the present work, it now appears that the adduct
obtained. This product was a ca. 1:1 mixture of two diaster- 27 obtained upon irradiation of 1 in the presence of ethyl
[12]
eomers, which differed only in the positions of the non- vinyl ether
is probably formed from 4 and ethyl vinyl
aromatic D atom in the cyclooctane ring. Both positions ether via a transition state resembling 26 (Scheme 8) ana-
are seq-trans vicinal to the exocyclic CϪC bond (for details logously to the formation of 12, rather than from 5 and
see Exp. Sect., last but one paragraph). Thus, the formal ethyl vinyl ether as was suggested in our previous paper[12]
addition of hexahydrofluorenone to the CϭC bond of (E)- on grounds of precedents.
Table 1. Analytical irradiation of 1 in the presence of (E)-cyclooctene
Run
(E)-Cyclooctene [][a]
[23]/[A[b]
]
[A]/[25]
[23]/[25]
[25]/[6 ϩ B[c]]
[23 ϩ A]/[6 ϩ 25 ϩ B]
1
2
3
4
5
6
0.7
0.7
1.4
1.4
2.8
2.8
1.52
1.21
1.81
1.79
1.69
1.87
0.105
0.116
0.138
0.140
0.295
0.267
0.16
0.14
0.25
0.25
0.50
0.50
3.24
3.74
5.44
4.78
5.93
4.55
0.203
0.202
0.328
0.323
0.680
0.627
[a]
[b]
0.1  1 in acetonitrile; six parallel runs in a merry-go-round apparatus at λ Ͼ300 nm for 4 h. Ϫ
A ϵ two [4 ϩ 2] DielsϪAlder
[c]
adducts of 9 to 2. Ϫ B ϵ two oxetanes (secondary photoproducts from 6).
Eur. J. Org. Chem. 2001, 2719Ϫ2726
2723

J. Leitich, I. Heise, J. Rust, K. Schaffner
FULL PAPER
placed 125-W high pressure mercury lamp as described previ-
ously.[2] After 1 h, conversion of 1 had reached 43% and irradiation
was discontinued. After evaporation of TFE and 9, the residue was
chromatographed with pentane ϩ 2% ether. Sequence of elution:
[4 ϩ 2] adducts, 21, 12, 1, 6.
1
Compound 21: H NMR (CDCl3): δ ϭ 0.99 (qt; J ϭ 3 ϫ 13.1 and
2 ϫ 3.0 Hz; 1 H), 1.39 and 1.49 (2 m; 1 H each), 1.61 (m; 3 H),
1.97 (td; J ϭ 2 ϫ 14.0 and 3.4 Hz; one 3-H), 2.12 (dddd; J ϭ 16.7,
2.7, 2.4, and 2.0 Hz; one 9-H), 2.39 (m; one 3-H), 2.42 (ddtd; J ϭ
16.7, 7.8, 2 ϫ 2.4, and 2.0 Hz; one 9-H), 2.60 (ddd; J ϭ 7.8, 7, and
2 Hz; 7-H), 2.77 (td; J ϭ 2 ϫ 7.8 and 6.2 Hz; 8-H), 3.50 (ddtd; J ϭ
6.2, 2.7, 2 ϫ 2.4, and 2.0 Hz; 1-H), 5.38 (dq; J ϭ 5.4 and 3 ϫ
2.4 Hz; 11-H), 5.70 (dq; J ϭ 5.4 and 3 ϫ 2.0 Hz; 10-H), 7.41 (m;
two m-phenyl H), 7.49 (m; p-phenyl H), 7.91 (d; J ϭ 8 Hz; two o-
phenyl H). Ϫ 1H,1H NOE enhancements (CDCl3): o-phenyl-H/1-
H, o-phenyl-H/11-H, o-phenyl-H/two 3-H, 1-H/8-H, 1-H/3-H at
δ ϭ 1.97, 3-H at δ ϭ 1.97/8-H, 7-H/9-H at δ ϭ 2.12. Ϫ 13C NMR
(CDCl3): δ ϭ 21.9 and 22.1 (C-4, -5), 24.9, 34.4, 35.5, 38.4, 40.4,
53.7 and 59.3 (C-6, -3, -8, -9, -7, -2, and -1, respectively), 128.4 (2
m-phenyl-C), 129.1 (2 o-phenyl-C), 129.9 (C-11), 132.3 (p-phenyl-
C), 134.1 (C-10), 134.9 (ipso-phenyl-C), 203.9 (CϭO).
Scheme 8. Proposed ion-pair-like (26) transition state for the addi-
tion of ethyl vinyl ether to 2-oxyallyl compound 4
Experimental Section
General Methods: See our earlier publications.[2,3] As described in
these, the molecular structures of new compounds encountered in
the present work could be fully elucidated by the use of 400-MHz
1
1H NMR techniques, including H,1H spin decoupling and NOE
(Nuclear Overhauser Effect) experiments where suitable, 13C BB
(Broad Band decoupling), 13C DEPT (Distortionless Enhancement
by Polarisation Transfer), and CH correlation spectroscopy.
(1R*,2S*,7S*,10S*,11S*)-8,9-Benzotetracyclo[9.3.0.02,7.02,10]-
tetradec-13-en-10-ol (14): A deaerated solution of 12 (103 mg,
0.4 mmol) in acetonitrile (40 mL) was irradiated with light of λ Ͼ
300 nm under argon at room temp. until full conversion (9 h). After
evaporation of the solvent, the residue was chromatographed with
pentane ϩ 3% ether. Sequence of elution: 10.1 mg of unidentified
mixtures, 23.3 mg (30.6%) of 6, 8.4 mg (6.5%) of 14, 61.1 mg of
unidentified mixtures.
Crystal Data Collection and Refinement of the Structure of 12: Crys-
˚
tal system: monoclinic; space group: Cc (no. 9); a ϭ 7.3572(6) A,
˚
˚
b ϭ 15.2850(12) A, c ϭ 12.4239(10) A; β ϭ 100.696(3)°; V ϭ
1372.85(19) A ; Z ϭ 4; ρcalcd. ϭ 1.221 g cmϪ3; F(000) ϭ 544 e; λ
(Mo-Kα) ϭ 0.71073 A; T ϭ 100 K; 2θmax ϭ 66.16°; reflections col-
3
˚
˚
Compound 14: 1H NMR (CDCl3): δ ϭ 1.19 (dddd; J ϭ 13, 11, 10.3,
and 4 Hz; one 6-H), 1.3 (m; 3 H), 1.42 (ddd; J ϭ 14, 9.5, and 4 Hz;
one 3-H), 1.51 (m; 1 H), 1.68 (ddd; J ϭ 14, 7, and 4 Hz; one 3-H),
1.94 (variable; bs; OH), 2.00 (ddddd; J ϭ 13, 6, 4, 3.5, and 1.5 Hz;
one 6-H), 2.46 (ddt; J ϭ 17.5, 10.5, and 2 ϫ 2.3 Hz; one 12-H),
2.57 (dtd; J ϭ 7.5, 2 ϫ 3.0, and 1.0 Hz; 1-H), 2.80 (ddtd; J ϭ 17.5,
3.8, 2 ϫ 3.0, and 2.3 Hz; one 12-H), 2.88 (ddd; J ϭ 10.5, 7.5, and
3.8 Hz; 11-H), 2.91 (dd; J ϭ 10.3 and 6.0 Hz; 7-H), 5.70 (dtd; J ϭ
5.6, 2 ϫ 3.0, and 2.3 Hz; 14-H), 5.95 (dtd; J ϭ 5.6, 2 ϫ 2.3, and
1.0 Hz; 13-H), 7.24 (m; 3 H), 7.37 (m; 1 H). Ϫ 1H,1H NOE en-
hancements (CDCl3): 1-H/7-H, 3-H at δ ϭ 1.42/14-H. Ϫ 13C NMR
(CDCl3): δ ϭ 21.3 and 21.8 (C-4, -5), 23.4 (C-3), 31.5 (C-6), 32.9
(C-12), 48.1 (C-11), 50.7 (C-7), 51.7 (C-1), 54.4 (C-2), 83.8 (C-10),
123.4, 124.3, 127.8, and 128.2 (4 CH), 130.8 (C-14), 134.6 (C-13),
147.7 and 147.9 (C-8, -9).
lected/unique: 6780/2867 [R(int) ϭ 0.1012]. The structure was
solved by direct methods and refined by full-matrix, least-squares
methods on F2 with 2 restraints, 172 parameters (SHELX-97). Hy-
drogen atoms were placed in calculated positions. All non-hydrogen
atoms were refined anisotropically. R1 ϭ 0.0622 [I Ͼ 2σ(I)], wR2 ϭ
0.1555, Goodness of fit on F2 ϭ 1.021; max./min. residual den-
sity ϭ 0.9/Ϫ0.3 e·AϪ3. Crystallographic data (excluding structure
˚
factors) for 12 have been deposited with the Cambridge Crystallo-
graphic Data Centre as supplementary publication no. CCDC-
155324. Copies of the data can be obtained free of charge on ap-
plication to CCDC, 12 Union Road, Cambridge CB2 1EZ, UK
[Fax: (internat.) ϩ 44-1223/336-033; E-mail: deposit@ccdc.cam.a-
c.uk].
(1ЈS*,4aS*,5ЈS*,9aR*)-9a-(5Ј-Deuteriocyclopent-2Ј-enyl)-
1,2,3,4,4a,9a-hexahydro-5,6,7,8-tetradeuteriofluoren-9-one (12-d5)
and (1S*,1ЈЈR*,4R*,4aS*,9S*,9aR*)-9,9a-Butano-1,4-(monodeut-
eriomethano)-1,4,4a,9a-tetrahydro-5,6,7,8-tetradeuterio-9H-fluoren-
4a-ol (11-d5): These compounds were prepared from the pentadeut-
Irradiation of 1 in Acetonitrile in the Presence of Styrene: Com-
pound 1 (1.0 g, 5.37 mmol), dissolved in a deaerated mixture of
acetonitrile (40 mL) and freshly distilled styrene (10 g), were irradi-
ated as described in the preceding sections for 70 min, to 84.7%
conversion. Still under argon, p-toluenesulfonic acid (10 mg) was
added, immediately bleaching the yellow solution (7 Ǟ 8). After
evaporation of the acetonitrile and most of the styrene, the residue
was chromatographed with pentane ϩ 2% ether. Sequence of elu-
tion: styrene, 22 (unknown stereochemistry, 2.3%), (1S*,6S*,8S*)-
22 (8.7%), (1R*,6S*,8S*)-22 (2.1%), [2-(2Ј-styryl)cyclohexyl](phen-
yl)methanone, 1, 24, 8, 6.
[2]
eriophenyl ketone 1-d5 analogously to the preparation of 12 and
1
11 from 1.[2] Comparison of the H and 13C NMR spectra of 12[2]
and 12-d5 shows that the H atom with signal at δ ϭ 1.68 on C-5Ј
(δ ϭ 25.3) of 12 is replaced by D in 12-d5 (the atom numbering is
different from that employed in Figure 1!). Similarly, the H atom
with signal at δ ϭ 1.25 on C-1ЈЈ (δ ϭ 44.0) in 11, which shows a
strong mutual 1H,1H NOE enhancement with 9-H,[2] is replaced by
D in 11-d5. No further differences in the nonaromatic part are ob-
served in the spectra of either compound.
Phenyl{8-phenylbicyclo[4.2.0]oct-1-yl}methanones (22)
Phenyl{(1R*,2R*,7S*,8S*)-tricyclo[6.3.0.02,7]undec-10-en-2-yl}-
methanone (21): Compound 1 (2.0 g, 10.75 mmol), dissolved in a
(1S*,6S*,8S*)-22: 1H NMR (CDCl3): δ ϭ 0.90 (qdd; J ϭ 3 ϫ 13.3,
deaerated mixture of 2,2,2-trifluoroethanol (TFE, 150 mL) and 5.6, and 3.6 Hz; 1 H), 1.40 (m; 1 H), 1.45 and 1.76 (2 m; 1 CH2),
freshly distilled cyclopentadiene (9) (23 g), was irradiated at room
temp. under argon with light of λ Ͼ 300 nm from a concentrically
1.60 (dm; J ϭ 12.4 Hz; one 5-H), 1.85 (ddd; J ϭ 13.4, 12.4, and
3.4 Hz; one 2-H), 2.12 (tdd; J ϭ 2 ϫ 12, 7.3, and 3.1 Hz; 6-H),
2724
Eur. J. Org. Chem. 2001, 2719Ϫ2726

Photo-Nazarov Cyclization of 1-Cyclohexenyl(phenyl)methanone
FULL PAPER
1
2.23 (ddd; J ϭ 9.4, 7.3, and 6.4 Hz; one 7-H), 2.24 (dtd; J ϭ 12.4,
duct): H NMR (CDCl3): δ ϭ 0.96 (qd; J ϭ 3 ϫ 13.3 and 3.6 Hz;
2 ϫ 12, and 4.1 Hz; one 5-H), 2.33 (dt; J ϭ 12.4 and 2 ϫ 3.0 Hz; 1 H), 1.33 (m; 2 H), 1.4Ϫ1.94 (m; 14 H), 1.46 and 1.75 (2 m; 12a-
one 2-H), 2.99 (ddd; J ϭ 12, 10.8, and 9.4 Hz; one 7-H), 3.70 (dd; H and 12b-H), 2.00 (m; 1 H), 2.12 (m; 1 H), 2.57 (dq; J ϭ 13.8
J ϭ 10.7 and 6.4 Hz; 8-H), 7.03 (m; 3 phenyl-H), 7.10 (m; 2 phenyl- and 3 ϫ 2.5 Hz; one 4-H), 3.65 (ddd; J ϭ 9.8, 6.5, and 2.7 Hz; 6a-
H), 7.17 (m; 5 phenyl-H). Ϫ 13C NMR (CDCl3): δ ϭ 23.2 and 26.9 H), 7.24 (tt; 2 ϫ 7.2 and 2 ϫ 1.6 Hz; p-phenyl-H), 7.31 (t; 2 ϫ
(C-3, -4), 23.7 (C-5), 32.2 (C-7), 39.1 (C-2), 45.7 (C-6), 55.1 (C-8), 7.2 Hz; 2 m-phenyl-H), 7.36 (dd; J ϭ 7.2 and 1.6 Hz; 2 o-phenyl-
64.3 (C-1), 127.0 (p-phenyl-C), 127.1, 128.2, 128.6, and 128.7 (4 ϫ
2 phenyl-C), 130.7 (p-phenyl-C), 138.2 and 139.7 (2 ipso-phenyl-C), and 28.0 (7 CH2), 29.3 (C-4), 31.0 (CH2), 33.3 (CH2), 42.8 and 44.6
H). Ϫ 13C NMR (CDCl3): δ ϭ 23.9, 25.6, 26.5, 26.8, 27.0, 27.4,
205.4 (CϭO).
(C-12a, -12b), 80.1 (C-6a), 113.7 (C-4a), 127.6 (p-phenyl-C), 127.8
(2 m-phenyl-C), 129.0 (2 o-phenyl-C), 137.0 (ipso-phenyl-C), 146.5
(C-5).
(1R*,6S*,8S*)-22: 1H NMR (CDCl3): δ ϭ 0.98 (qdd; J ϭ 3 ϫ 13.3,
5.6, and 3.6 Hz; 1 H), 1.45 (m; 2 H), 1.62 (m; 1 H), 1.62 and 1.71
(2 m; two 5-H), 1.87 (m; two 2-H), 2.04 (ddd; J ϭ 10.2, 8, and
7.8 Hz; one 7-H), 2.31 (td; J ϭ 2 ϫ 10.8 and 10.2 Hz, one 7-H),
2.95 (dddd; J ϭ 10.8, 8, 6.7, and 1.4 Hz; 6-H), 3.62 (dd; J ϭ 10.8
and 7.8 Hz; 8-H), 7.21 (d; J ϭ 7.3 Hz; 2 o-phenyl-H), 7.29 (m; p-
phenyl-H), 7.31 (m; 4 m-phenyl-H), 7.43 (t; J ϭ 2ϫ 7.3 Hz; p-
benzoyl-H), 7.51 (d; J ϭ 7.8 Hz; 2 o-benzoyl-H). Ϫ 1H,1H NOE
enhancements (CDCl3): o-benzoyl-H/2-H, o-benzoyl-H/5-H at δ ϭ
1.71, o-benzoyl-H/6-H, o-benzoyl-H/8-H, 6-H/8-H. Ϫ 13C NMR
(CDCl3): δ ϭ 21.3 and 21.4 (C-3 and -4), 25.2 (C-7), 25.4 (C-5),
25.8 (C-2), 32.3 (C-6), 45.3 (C-8), 57.0 (C-1), 126.7 (p-phenyl-C),
127.9 and 128.0 (each 2 m-phenyl-C), 128.6 (2 o-benzoyl-C), 129.5
(2 o-phenyl-C), 131.4 (p-benzoyl-C), 136.8 and 139.7 (2 ipso-
phenyl-C), 206.8 (CϭO).
Phenyl{(1R*,2S*,7S*,8R*)-tricyclo[6.6.0.02,7]tetradec-2-yl}-
1
methanone (23): H NMR (CDCl3): δ ϭ 1.03 (m; J8,9α ϭ 10.0 Hz;
9α-H), 1.03 (m; J3β,4α ϭ 13.5 Hz; 4α-H), 1.1Ϫ1.8 (m; 10 H), 1.50
(m; J3β,4β ϭ 3.5 Hz; 4β-H), 1.52 and 1.64 (2 m; 6-CH2), 1.70 (m;
J1,14α ϭ 5.7 Hz; 14α-H), 1.77 (m; J8,9β ϭ 3.5 Hz; 9β-H), 1.80 (m;
J1,14β ϭ 10.0 Hz; 14β-H), 1.83 (ddd; J ϭ 14.2, 13.5, and 3.5 Hz;
3β-H), 1.91 (tdd; J ϭ 2 ϫ 10.0, 9.0, and 3.5 Hz; 8-H), 2.02 (dtd;
J ϭ 14.2, 2 ϫ 3.6, and 1 Hz; 3α-H), 2.13 (ddd; J ϭ 10.0, 9.0, and
5.7 Hz; 1-H), 2.19 (dddd; J ϭ 10.0, 5.7, 2, and 1 Hz; 7-H), 7.38 (t;
J ϭ 2 ϫ 7.2 Hz; 2 m-phenyl-H), 7.45 (t; J ϭ 2 ϫ 7.2 Hz; p-phenyl-
H), 7.71 (d; J ϭ 7.2 Hz; 2 o-phenyl-H). Ϫ 1H,1H NOE enhance-
ments (CDCl3): o-phenyl-H/1-H, o-phenyl-H/3α-H, o-phenyl-H/4α-
H, o-phenyl-H/7-H, o-phenyl-H/both 14-H, 1-H/9α-H, 3α-H/14α-
H, 7-H/9α-H. Ϫ 13C NMR (CDCl3): δ ϭ 21.6 (C-4), 22.2 (CH2),
23.6 (C-6), 24.6 (CH2), 26.0 (CH2), 26.9 (C-3), 27.1 (CH2), 28.3
(CH2), 28.4 (C-14), 35.7 (C-9), 38.0 (C-8), 40.2 (C-7), 47.5 (C-1),
53.0 (C-2), 128.1 (2 m- phenyl-C), 128.3 (2 o-phenyl-C), 131.5 (p-
phenyl-C), 136.9 (ipso-phenyl-C), 207.2 (CϭO).
(4aR*,9aS*)-9a-(2-Phenylethyl)-1,2,3,4,4a,9a-hexahydrofluoren-9-
one (24): M.p. 65 °C. Ϫ MS (EI and CI): m/z ϭ 290 [Mϩ]. Ϫ H
1
NMR (CDCl3): δ ϭ 1.27 (m; 2 H), 1.48 and 1.57 (2 m; 1 H each),
1.58 and 1.76 (2 m; 1 CH2), 1.83 and 1.99 (2 m; 1 CH2), 1.89 and
2.00 (each a d-quasi-t; J ϭ 13.8 and 2 ϫ 8.6 Hz; two 1Ј-H), 2.49
(quasi-t; J ϭ 2 ϫ 8.6 Hz; two 2Ј-H), 3.29 (t; J ϭ 2 ϫ 5.5 Hz; 4a-
H), 7.14 (m; 3 phenyl-H), 7.23 (m; 2 phenyl-H), 7.37 (t; J ϭ 2 ϫ
7.4 Hz; 7-H), 7.47 (d; J ϭ 7.6 Hz; 5-H), 7.59 (td; J ϭ 2 ϫ 7.4 and
1.1 Hz; 6-H), 7.77 (d; J ϭ 7.6 Hz; 8-H). Ϫ 13C NMR (CDCl3): δ ϭ
20.3, 20.5, 26.7, and 31.0 (4 CH2), 31.1 (C-2Ј), 38.9 (C-1Ј), 43.2 (C-
4a), 53.3 (C-9a), 124.2 (C-8), 124.8 (C-5), 125.8 (p-phenyl-C), 127.4
(C-7), 128.3 (2 m-phenyl-C), 128.4 (2 o-phenyl-C), 134.5 (C-6),
136.0 (ipso-phenyl-C), 142.4 (C-8a), 156.7 (C-4b), 210.4 (C-9).
Spiro[1,2,3,4,4a,9a-hexahydro-9H-fluorene-9,10Ј-9Ј-oxabicyclo-
[6.2.0]decane] (Major Oxetane Photoproduct from 6): 1H NMR
(CDCl3): δ ϭ 1.0Ϫ1.7 (m; 15 H), 1.73 (dddd; J ϭ 12.9, 12.0, 10.8,
and 3.4 Hz; one 7Ј-H), 1.83 (m; 9a-H and 2 H), 1.91 (m; 1 H), 2.03
(ddt; J ϭ 12.9, 5.5, and 2 ϫ 3.5 Hz; one 7Ј-H), 3.04 (bt; 2 ϫ 6 Hz;
4a-H), 3.18 (ddd; J ϭ 11.5, 8.2, and 3.6 Hz; 1Ј-H), 4.83 (ddd; J ϭ
10.8, 8.2, and 3.5 Hz; 8Ј-H), 7.10 (m; 1 H), 7.22 (m; 2 H), 7.63 (m;
1 H). Ϫ 13C NMR (CDCl3): δ ϭ 25.6, 26.5, 27.7, 27.8, 28.2, 28.3,
28.6, 29.8, and 31.0, (9 CH2), 38.1 (CϪ7Ј), 41.5 (C-9a), 48.6 (C-
4a, -1Ј), 83.5 (C-8Ј), 99.5 (C-9), 122.7, 125.8, 126.0, 128.0 (C-5
through C-8), 143.9 and 147.3 (C-4b, -8a).
Irradiation of 1 in Acetonitrile in the Presence of (E)-Cyclooctene:
Compound 1 (1.0 g, 5.37 mmol), dissolved in a deaerated mixture
of acetonitrile (45 mL) and freshly distilled (E)-cyclooctene (5.5 g),
was irradiated as described[2] for 1 h and to 66.1% conversion.
After evaporation of the acetonitrile and most of the (E)-cyclooc-
tene, the residue was chromatographed with pentane ϩ 1% ether.
Sequence of elution: (E)-cyclooctene, major [4 ϩ 2] adduct (6.7%),
minor [4 ϩ 2] adduct (1.3%), 23, major oxetane (secondary
photoproduct from 6), 25, 1, minor oxetane (secondary photoprod-
uct from 6), 6, cyclobutanol isomer of 25 (secondary photoproduct
from 25).
(4aS*,9aR*)-9a-Cyclooctyl-1,2,3,4,4a,9a-hexahydrofluorenone
(25): M.p. 80 °C. Ϫ MS (EI): m/z ϭ 296 [Mϩ]. Ϫ 1H NMR
(CDCl3): δ ϭ 1.02 (m; 2 H), 1.22 (dtd; J ϭ 13.5, 2 ϫ 9.0, and
3.5 Hz; one 8Ј-H), 1.27 (ddt; J ϭ 13.5, 7.5, and 2 ϫ 3.5 Hz; one
8Ј-H), 1.3Ϫ1.6 (m; 11 H), 1.33 (m; one 2Ј-H), 1.67 (m; 3 H; one 1-,
one 2Ј-, and 1 H), 1.76 (ddd; J ϭ 13.8, 10.4, and 4.5 Hz; one 1-
H), 1.83 (dq; J ϭ 13.9 and 3 ϫ 4.5 Hz; one 4-H), 1.89 (ddt; J ϭ
9.0, 8.0, and 2 ϫ 3.5 Hz; 1Ј-H), 1.95 (ddt; J ϭ 13.9, 11.5, and 2 ϫ
5.3 Hz; one 4-H), 3.26 (dd; J ϭ 5.3 and 4.5 Hz; 4a-H), 7.32 (t; J ϭ
2 ϫ 7.4 Hz; 7-H), 7.43 (d; J ϭ 7.7 Hz; 5-H), 7.55 (t; J ϭ 2 ϫ 7.5 Hz;
(6aS*,12aR*,12bS*)-5-Phenyl-1,3,4,6a,7,8,9,10,11,12,12a,12b-
dodecahydro-2H-6-oxacycloocta[a]naphthalene (Major [4 ؉ 2] Ad-
1
duct): H NMR (CDCl3): δ ϭ 1.1Ϫ1.95 (m; 18 H), 1.85 (td; 2 ϫ
1
6-H), 7.69 (d; J ϭ 7.7 Hz; 8-H). Ϫ H,1H NOE enhancements: 4a-
13.1 and 4.1 Hz; one 4-H), 2.05 (m; 12a-H and 12b-H), 2.50 (ddt;
J ϭ 13.1, 4.0, and 2 ϫ 2.0 Hz; one 4-H), 3.87 (ddd; J ϭ 10.2, 5.6,
and 3.1 Hz; 6a-H), 7.23 (tt; J ϭ 2 ϫ 7.2 and 2 ϫ 1.6 Hz; p-phenyl
H), 7.30 (t; J ϭ 2 ϫ 7.2 Hz; 2 m-phenyl-H), 7.36 (dd; J ϭ 7.2 and
1.6 Hz; 2 o-phenyl-H). Ϫ 13C NMR (CDCl3): δ ϭ 21.9, 25.7, 26.7,
27.0, 27.3, 28.2, 28.9, 29.1, and 29.9 (9 CH2) 30.7 (C-4), 39.6 (C-
12b), 42.3 (C-12a), 76.0 (C-6a), 114.1 (C-4a), 127.4 (p-phenyl-C),
127.7 (2 m-phenyl-C), 129.7 (2 o-phenyl-C), 136.7 (ipso-phenyl-C),
145.6 (C-5).
H/5-H, 4a-H/1Ј-H, 4a-H/2Ј-H at δ ϭ 1.67, 4a-H/8Ј-H at δ ϭ 1.22.
Ϫ
13C NMR (CDCl3): δ ϭ 17.9, 18.9, 26.1, and 26.2 (4 CH2), 26.3
(C-4), 26.6, 26.9, 27.3 (3 CH2), 27.8 (C-1), 29.0 (C-2Ј), 29.5 (C-8Ј),
41.3 (C-4a), 43.3 (C-1Ј), 57.5 (C-9a), 123.3 (C-8), 124.9 (C-5), 127.2
(C-7), 134.4 (C-6), 137.7 (C-4b), 158.1 (C-8a), 212.8 (C-9).
Spiro[1,2,3,4,4a,9a-hexahydro-9H-fluorene-9,10Ј-9Ј-oxabicyclo-
[6.2.0]decane] (Minor Oxetane Photoproduct from 6): 1H NMR
(CDCl3): δ ϭ 0.9Ϫ1.4 (m; 7 H), 1.46 (m; 1 H), 1.60Ϫ1.95 (m; 10
(6aS*,12aR*,12bR*)-5-Phenyl-1,3,4,6a,7,8,9,10,11,12,12a,12b- H), 2.06 (dq; J ϭ 14.0 and 3 ϫ 4.0 Hz; one 4-H), 2.11 (ddd; J ϭ
dodecahydro-2H-6-oxacycloocta[a]naphthalene (Minor [4 ؉ 2] Ad-
12.8, 5.2, and 3.4 Hz; one 7Ј-H), 2.75 (ddd; J ϭ 12.1, 7.8, and
Eur. J. Org. Chem. 2001, 2719Ϫ2726
2725

J. Leitich, I. Heise, J. Rust, K. Schaffner
2.9 Hz; 1Ј-H), 2.85 (dt; J ϭ 10.6 and 2 ϫ 5.9 Hz; 9a-H), 3.04 (dddt; reduced in intensity but unchanged in shape; the 8Ј-H signal at δ ϭ
FULL PAPER
J ϭ 5.9, 5.7, 4.0, and 2 ϫ 1.0 Hz; 4a-H), 4.54 (td; J ϭ 2 ϫ 7.8 and
3.4 Hz; 8Ј-H), 7.13 (m; 5-H), 7.25 (m; 6-H and 7-H), 7.55 (quasi-
d; 8-H). Ϫ 13C NMR (CDCl3): δ ϭ 21.9, 23.4, 24.5, 26.4 (4 CH2),
26.5 (C-4), 28.2, 28.3, 28.7, and 29.7 (4 CH2), 38.1 (C-7Ј), 40.8 (C-
4a), 43.6 (C-9a), 54.0 (C-1Ј), 83.4 (C-8Ј), 94.0 (C-9), 122.6 (C-5),
124.2 (C-8), 126.7 and 128.1 (C-6, Ϫ7), 143.8 and 147.9 (C-4b, -8a).
1.22 is unchanged in intensity but changed in shape.
Analytical Irradiation of 1 in the Presence of (E)-Cyclooctene: Six
equal Solidex glass tubes, each containing 13.1-ml samples of
250 mg of 1 (0.1 ) and varying concentrations of (E)-cyclooctene
(0.5 , 1.0 , 1.94 ; 2 samples of each) in acetonitrile, were placed
in a merry-go-round apparatus with a central 125-W high pressure
mercury lamp (Philips HPK 125) and irradiated for 4 h, after which
time about half of 1 had been converted into products. [Concentra-
tions of (E)-cyclooctene higher than the ones above resulted in in-
homogeneous solutions and were discarded.] A 1-ml aliquot from
each sample was concentrated at room temp. and 0.1 mbar, which
did not remove all of the (E)-cyclooctene, and the residue was ana-
lysed by 400-MHz 1H NMR spectroscopy in the regions δ ϭ 2.5Ϫ5
and 7Ϫ8, making use of the known absorption of the individual
components. For the results see Table 1.
8,9-Benzotetracyclo[9.6.0.02,7.02,10]heptadecan-10-ol
(Secondary
Photocyclisation Product from 25): 1H NMR (CDCl3): δ ϭ 1.07 (m;
1 H), 1.15Ϫ1.95 (m; 20 H), 2.00 (ddd; J ϭ 10, 9, and 2.5 Hz; 11-
H), 2.06 (ddd; J ϭ 11, 9, and 2.5 Hz; 1-H), 3.25 (t; 2 ϫ 6.3 Hz; 7-
1
H), 7.20 (m; 2 H), 7.23 (m; 2 H). Ϫ H,1H-NOE enhancement: 1-
H/7-H. Ϫ 13C NMR (CDCl3): δ ϭ 19.6, 20.0, 26.3, 27.3, 27.5, 28.0,
28.0, 28.3, 28.5, and 30.6 (10 CH2) 42.3 (C-7), 47.4 (C-11), 49.3 (C-
1), 53.3 (C-2), 85.4 (C-10), 122.0, 123.5, 126.9, 127.6 (4 CH), 147.5
and 148.3 (C-8, Ϫ9).
Acknowledgments
The authors thank Mr. Jörg Bitter and Mrs. Kerstin Sand for carry-
ing out the NMR experiments and Mr. Wulf Schmöller (Max-
Planck-Institut für Kohlenforschung) for mass spectra.
(4aS*,9aR*)-9a-(1ЈR*,2ЈR*- and 1ЈS*,2ЈS*-2Ј-Deuteriocyclo-
octyl)-1,2,3,4,4a,9a-hexahydro-5,6,7,8-tetradeuteriofluoren-9-one
(25-d5): The ca. 1:1 mixture of the two 2Ј-deuterio compounds 25-
d5 was prepared by irradiation of the pentadeuteriophenylme-
thanone 1-d5 and (E)-cyclooctene as described above for nondeu-
terated 25. The NMR spectra (500 MHz 1H and 125 MHz 13C)
showed the following differences relative to nondeuterated 25:
(i) The four aromatic CϪH moieties were replaced by CϪD accord-
[1]
Unbracketed compounds in Scheme 1 were stable enough to
be isolated or detected in solution even after prolonged times
at room temperature. In all schemes in this work, arrows bear-
ing ‘‘hν’’ refer to photochemical reactions, while arrows with-
out ‘‘hν’’ refer to dark reactions at room temperature. In all
schemes, stereoformulae are intended to designate relative, not
absolute, configurations.
1
ing to both the H and 13C BB spectra. (ii) The 1Ј-H signal at δ ϭ
1.89 was changed from a ddt (J ϭ 9.0, 8.0, and 2 ϫ 3.5 Hz) to a
clean ddd (J ϭ 9.0, 8.0, and 3.5 Hz), indicating that the fifth D
atom was in a vicinal position. Since the two larger couplings (J ϭ
9.0 and 8.0 Hz) correspond to antiperiplanar H,H arrangements
and the two smaller ones (J ϭ 2 ϫ 3.5 Hz) to synclinal H,H ar-
rangements, the cyclooctane 1Ј-H is seq-trans-configured relative to
each of the two H atoms coupled to it by the larger coupling con-
stants. This and the disappearance of exactly one 3.5 Hz coupling
means that the exocyclic CϪC bond on the cyclooctane and the
vicinal D atom are also seq-trans-configured to each other. This
demonstrates a clean suprafacial addition of D and fluorenonyl to
(E)-cyclooctene, irrespective of the distribution of the D atom be-
tween C-2Ј and C-8Ј. (Note that after the D atom has been placed
on C-8Ј, the cyclooctyl ring positions have to be renumbered so as
to change C-8Ј into -2Ј, whereas no renumbering is carried out if
the D atom has been placed on C-2Ј.) Further differences in the
NMR spectra show that this distribution is ca. 1:1, with the D
atom replacing 2Ј-H (signal at δ ϭ 1.67) and 8Ј-H (signal at δ ϭ
1.27). (iii) In the 13C BB spectra, about half of the C-8Ј singlet at
δ ϭ 29.5 is replaced by a small triplet (J ϭ 18.5 Hz) at δ ϭ 29.1;
similarly, about half of the C-2Ј singlet at δ ϭ 29.0 is replaced by
a small triplet (J ϭ 19.1 Hz) at δ ϭ 28.6. (iv) The proton integrals
in the ranges δ ϭ 1.15Ϫ1.60 and 1.60Ϫ1.80 are reduced by approx-
imately one half H atom each. (v) The 8Ј-H signal at δ ϭ 1.27 is
[2]
J. Leitich, I. Heise, S. Werner, C. Krüger, K. Schaffner, J.
Photochem. Photobiol. A: Chem. 1991, 57, 127Ϫ151.
J. Leitich, I. Heise, K. Schaffner, J. Photochem. Photobiol. A:
[3]
Chem., accepted for publication.
Reviews on 2-oxyallyl compounds: J. Mann, Tetrahedron 1986,
[4]
42, 4611Ϫ4659. H. M. R. Hoffmann, Angew. Chem. 1984, 96,
29Ϫ48; Angew. Chem. Int. Ed. Engl. 1984, 23, 1.
Compounds 10, 12, and 13 were isolated after irradiation of 1
[5]
to partial conversion. On more extended irradiation, 10, 12,
and 13 were converted into the products shown.
[6]
R. K. Siemionko, J. A. Berson, J. Am. Chem. Soc. 1980, 102,
3870Ϫ3882.
[7]
B. Föhlisch, R. Joachimi, Chem. Ber. 1987, 120, 1951Ϫ1960.
W. v. E. Doering, M. Franck-Neumann, D. Hasselmann, R. L.
[8]
Kaye, J. Am. Chem. Soc. 1972, 94, 3833Ϫ3844.
[9]
J. Leitich, I. Heise, Eur. J. Org. Chem., 2001, 2707Ϫ2718, pre-
ceding paper.
[10]
C. J. Cramer, S. E. Barrows, J. Org. Chem. 1998, 63,
5523Ϫ5532.
J. C. Scaiano, Acc. Chem. Res. 1982, 15, 252Ϫ258.
A. E.-A. M. Gaber, I. Heise, J. Leitich, K. Schaffner, J.
[11]
[12]
Photochem. Photobiol. A: Chem. 1999, 123, 31Ϫ37.
Received January 26, 2001
[O01035]
2726
Eur. J. Org. Chem. 2001, 2719Ϫ2726