Synthesis of 2-silaoxane via 1,3- photocycloaddition

Article abstract of DOI:10.1002/jhet.5570380615

Irradiation of an aryl group with a silyl tethered alkene yields a tetracyclic 2-silaoxane with high regioselectivity. The multicyclic structure has been further modified to give an unstable tricyclic diol. Photocycloaddition between cyclopentene and phen

Full text of DOI:10.1002/jhet.5570380615

Nov-Dec 2001
Synthesis of 2-Silaoxane Via 1,3- Photocycloaddition
1341
Suhail Bahu, Steven A. Fleming*, Brad Nilsson, and Tim Turner
Department of Chemistry and Biochemistry, Brigham Young University, Provo, UT 84602
Received May 15, 2001
Irradiation of an aryl group with a silyl tethered alkene yields a tetracyclic 2-silaoxane with high regiose-
lectivity. The multicyclic structure has been further modified to give an unstable tricyclic diol.
Photocycloaddition between cyclopentene and phenylcyclopropane gave a single major cycloadduct without
cyclopropyl ring opening, indicating that the putative radical intermediate involved in cycloaddition appar-
ently has a very short lifetime if it exists at all.
J. Heterocyclic Chem., 38, 1341 (2001).
Introduction.
There is evidence that this intermediate is dipolar in nature.
Evidence of the diradical intermediate would be a signifi-
cant advance in our understanding of this reaction.
We have been interested in photochemical reactions
involving tethered systems for the development of syn-
thetic methodology that can be exploited in the synthesis
of complex molecules [1]. Recent literature is replete with
examples of using light-induced reactions for the ultimate
synthesis of unnatural and natural compounds [2]. There
is great potential for using controlled photochemical
cycloadditions in synthesis of complex compounds
because the cycloadditions typically introduce numerous
carbon-carbon bonds in a single step.
Analysis of photocycloadditions that are concerted
shows that control of the conformation and proximity of
the tethered chromophore and its addend can allow for sig-
nificant enhancement in the formation of a single product.
Even photoadditions that are stepwise (e.g., triplet photo-
chemical reactions) can be controlled by tethering. In this
case, the design of the reaction suggests that if formation
of the first bond is controlled then the subsequent bonds
can be predicted based on steric factors.
One of the most synthetically useful photochemical
reactions has been the 1,3 photocycloaddition (the arene-
olefin or meta photocycloaddition). This reaction has been
investigated by many groups [3]. The use of tethering has
contributed to the value of this reaction, for example, it is
known that the three-atom tether is preferred in the 1,3-
cycloaddition and the yields for these intramolecular reac-
tions are typically in the 80% range. This is presumably
due to the efficient overlap between the chromophore and
the alkene which facilitates formation of a necessary exci-
plex. A tether that is removable after the reaction has been
carried out would be a significant advance in the synthetic
methodology of this photocycloaddition.
Results.
In this paper we describe the use of silicon as a tempo-
rary tether in intramolecular 1,3 photocycloaddition which
allows for the synthesis of the 2-silaoxane ring system. We
also report our attempt to observe the diradical intermedi-
ate of this type of cycloaddition.
Our studies in the area of silicon tethered 2+2 photocy-
cloadditions led us to an attempt of the Paterno-Büchi pho-
tocycloaddition between a tethered alkene and o-hydroxy-
acetophenone. We observed that a minor product from this
reaction was a heterocyclic compound that appeared to be
the result of an intramolecular 1,3-photocycloaddition
between the alkene and the aromatic ring. This prompted
us to examine the photochemistry of just phenol tethered
to an alkene since the carbonyl portion of o-hydroxyace-
tophenone undoubtedly results in formation of the triplet
excited state and the standard 1,3-photocycloaddition is a
singlet process [3]. Indeed, we found that a good yield of
a 1,3-photocycloadduct is obtained from irradiation of
allylphenoxydimethylsilane (1).
Synthesis of the silyl compound 1 was accomplished in
one step by silylation of phenol with allyl-
dimethylchlorosilane under standard conditions.
Irradiation of this compound resulted in a 52% conversion
to a major product, the tetracyclic 2-silaoxane 2 (see
Scheme 2). The relative stereochemistry of the hetero-
cyclic product was verified by 500 MHz 2-D nmr analysis
including COSY and HETCOR experiments. Comparison
with previously reported 1,3-photocycloadducts was very
helpful in the assignment of this structure [4].
The mechanism of this reaction is presumed to involve a
diradical intermediate that results from initial addition of the
alkene across the aromatic ring in a 1,3 (or 2,6 depending on
the substituents on the ring) fashion as shown in Scheme 1.
There are several potential cycloaddition products but
each of the alternative structures has characteristic nmr
signals which are not consistent with the observed com-
pound. For example, we initially thought that the alkene
addition occurred across the 1,3 position of the aromatic
ring which would produce the linear tetracyclic product
similar to that shown in Scheme 1. This structure was
ruled out by the signal for one of the C-11 hydrogens (see
Scheme 2) that is coupled to both cyclopropyl hydrogens;

1342
S. Bahu, S. A. Fleming, B. Nilsson, and T. Turner
Vol. 38
one by W-coupling and the other by vicinal coupling. The
linear molecule would not have shown coupling between
the methylene and the cyclopropyl hydrogens.
This relatively unstable compound has five contiguous
stereocenters, an alkene for further modification, a primary
alcohol, and a tertiary alcohol.
It is also interesting to note that the photocyclization
appears to occur via an alkoxy stabilized radical or zwitte-
rionic intermediate. The regiochemistry of the observed
product is consistent with a stepwise-like (nonsynchro-
nous) mechanism which has been proposed for this photo-
cycloaddition and favors oxygen substitution at the low
electron density center [5]. In addition, Gilbert has
recently reported 1,3-photocycloaddition on an organosi-
lane with regioselectivity that presumably arises from sili-
con stabilization of a β positive charge in the transition
state [4,6]. Thus, both the oxygen and the silicon groups
would favor the intermediate being zwitterionic with the
positive center at the ipso carbon.
This example with silicon-oxygen attachment demon-
strates the regioselectivity that can be obtained, it is also
the first example of using a silicon tether as a removable
group following the cycloaddition. This is a powerful tool
in synthetic applications because there are five contiguous
stereocenters that are formed with regio- and stereocontrol
in one step. These centers are maintained following
removal of the tethering group.
Based on the results of this siloxy tethered cycloaddi-
tion, we questioned if there is any diradical nature in the
intermediate of the 1,3-photocyclization [8]. A long-lived
diradical should be sensitive to ipso cyclopropyl substitu-
tion. So we reasoned that irradiation of phenylcyclo-
propane in the presence of an alkene would result in cyclo-
propyl ring-opened products if the diradical intermediate is
involved in the reaction as shown in Scheme 4.
We have explored the photocycloaddition in excess
cyclopentene as our test for ring opened photoproducts
which one would predict as a result of a radical intermedi-
-8
ate having a lifetime on the order of 10 s or more [9]. In
this system, closure of the ring-opened diradical would
give a bridgehead alkene with a trans substituted eight
member ring [10] and, significantly, no cyclopropyl rings
(see Scheme 4). Although the ring opened photoproduct
may be relatively unstable due to the bridgehead alkene,
we predicted that we would be able to observe it or an
alkene shifted by-product if the cyclopropyl ring opened.
Irradiation of phenylcyclopropane in the presence of a
10-fold excess of cyclopentene in photograde cyclohexane
was performed using an Ace Glass Thin-Film Reactor.
The estimated conversion for these conditions is 25%.
One major product (>80% of tractable material) was
obtained and identified as the non ring-opened photoprod-
uct 4 (see Scheme 5). Spectral analysis of this compound
is consistent with previously reported toluene + cyclopen-
tene photoadducts [11].
There are a number of potential manipulations that can
be considered for this complex tetracyclic silane. For
example, Tamao-Fleming oxidation [7] with hydrogen per-
oxide allowed conversion to the diol 3 (see Scheme 3).
The results do not completely rule out the reversible ring
opening process shown in Scheme 4. However, a plausible
explanation based on our observation is that the 1,3-pho-
toaddition is dipolar in nature rather than radicaloid. This
is consistent with previous mechanistic analysis of this
cycloaddition [12]. The alkoxy substituent, in particular,

Nov-Dec 2001
Synthesis of 2-Silaoxane Via 1,3- Photocycloaddition
1343
Anal. Calcd. for C H OSi: C, 68.69; H, 8.39. Found: C,
68.50; H, 8.29.
is ideally suited for stabilizing a partial positive charge at
the ipso position that is presumably generated upon excita-
tion. We have previously reported the results of MOPAC
calculations that support such an intermediate for the irra-
diation of anisole [13]. More recently, Pincock has con-
vincing evidence for an analogous species from excitation
of substituted benzonitriles [14]. For the nitrile substituted
aryl group, the polarization favors a partial negative charge
at the ipso position.
The dipolar reactive intermediate explains the high
regioselectivity observed in the silyl tethered phenoxy
cyclization and the lack of cyclopropyl ring-opened prod-
uct. We are continuing our investigation in this area.
11 16
2,7 2,10
3-Oxa-4-sila-4,4-dimethyltetracyclo[4.4.1.0 .0 ]dec-8-ene (2).
Irradiation of allyldimethylphenoxysilane (107 mg) in 120 mL
-3
cyclohexane (4.6 x 10 M) for 12 hours was followed by concen-
tration. Chromatography (pentane:EtOAc 25:1) of the photomix-
ture gave 55.7 mg of tetracyclic silaoxane 2, b.p. 60 °C/1.0 Torr
1
(Kugelrohr). H nmr (200 MHz, deuteriochloroform): δ 5.59 (dd,
1H, H-9), 5.52 (ddd, 1H, H-8), 2.77 (bd, 1H, H-10), 2.63 (m, 1H,
H-7), 2.61 (m, 1H, H-6), 1.66 (m, 1H, H-11), 1.57 (dd, 1H, H-11'),
1.44 (ddt, 1H, H-1), 1.09 (ddd, 1H, H-5), 0.68 (dd, 1H, 1 H, H-5'),
13
0.30 (s, 3H, Si-CH ), 0.25 (s, 3H, Si-CH ); C nmr (50 MHz,
3
3
deuteriochloroform) 129, 127, 88, 70, 62, 50, 41, 29, 21, -1.
COSY and HETCOR at 500 MHz confirm these assignments.
hrms m/z calc. 192.0970, obs. 192.0978.
Anal. Calcd. for C H OSi: C, 68.69; H, 8.39. Found: C,
68.46; H, 8.50.
11 16
EXPERIMENTAL
2,8
6-Hydroxymethyltricyclo[3.3.0.0 ]oct-3-en-1-ol (3).
General.
In 11.6 mL of 1:1 THF and methanol was dissolved 280 mg (1.45
mmol) of 2-siloxane. To this was added 0.12 g of potassium bicar-
bonate, 0.296 mL of hydrogen peroxide, and 0.51 g of potassium
fluoride. The reaction stirred at room temperature for 18 hours then
was extracted with chloroform. The organic layer was concentrated
and the crude material was analyzed by nmr. Chromatographic
All solvents used were reagent grade and used as purchased
except for tetrahydrofuran (THF) and cyclohexane. THF was
distilled over Na metal and benzophenone. Reagent grade cyclo-
hexane (1.5 L) was washed with an equal volume of concentrated
H SO and HNO three times (200 mL each time). Then the
2
4
3
cyclohexane was washed with 200 mL of dilute aqueous KMnO
4
1
purification gave a poor yield (<5%) of the unstable diol. H nmr
(200 MHz, deuteriochloroform): δ 5.55 (bs, 2H, H-9), 3.95 (m, 2H,
and then 200 mL of water. The cyclohexane was dried over
CaCl before being passed through an activated basic alumina
2
CH O), 3.2 (bs, 2H, -OH), 2.1 (m, 1H), 1.9 (m, 5H).
column and then distilled over CaCl to obtain photograde cyclo-
2
2
hexane.
2,11 3,7
1-Cyclopropyltetracyclo[6.3.0.0 .0 ]undec-9-ene (4).
All reactions were run under nitrogen. Chromatography was
performed using Silica Gel 60 (230-400 mesh) on a gravity col-
umn or a chromatotron made by Harrison Research. The chro-
matotron is a rotating chromatography disk that allows detection
and separation of 25-200 mg quantities.
Phenylcyclopropane and cyclopentene (1:10 molar equivalent)
were added to 200 mL of cyclohexane (0.1 M phenylcyclo-
propane in cyclohexane) and circulated through a Thin-Film
Reactor. After 5 hours of irradiation with a 450 W Hg lamp, the
mixture was evaporated and chromatographed. The percent con-
version based on unreacted phenylcyclopropane was typically
25%. The only product that is obtained in sufficient quantity to
be identified is the cyclic compound 4 which is formed in >80%
based on the converted phenylcyclopropane.
Nuclear magnetic resonance (nmr) spectra were obtained
either on a Varian Gemini 200 MHz, a Varian Inova 300 MHz, or
a Varian Unity 500 MHz spectrometer. Chemical shifts reported
are ppm downfield from tetramethylsilane (TMS). High resolu-
tion mass spectra (hrms) were determined on a JEOL-SX102A
mass spectrometer using direct probe sample introduction.
Irradiations were performed using a Hanovia 450 W medium
pressure mercury lamp in a water-cooled quartz well. All irradia-
tions were deoxygenated by bubbling purified nitrogen gas through
the reaction solution for 10 minutes before irradiation and continu-
ing though the time of the reaction. The nitrogen was purified by
passing it through an Ace-Burlitch inert atmosphere system con-
taining a column packed with a BASF R3-11 catalyst followed by
another column packed with Aquasorb® drying agent.
1
The data for 4 are: H nmr (500 MHz, deuteriochloroform): δ
5.70 (d, 1H, H-10), 5.63 (d, 1H, H-9), 3.20 (m, 1H, H-7), 2.86 (dt,
1H, H-3), 2.81 (d, 1H, H-8), 1.82 (m, 1H, H-6), 1.65 (m, 2H, H 4,
H-6'), 1.52 (d, 1H, H-11), 1.43 (m, 2H, H-2, H-4'), 1.40 (t, 2H, H-
5), 1.21 (dt, 1H, H-12), 0.40 (m, 2H, H-13, H-14), 0.10 (m, 2 H,
13
H-13', H-14'); C nmr (50 MHz, deuteriochloroform) 135.5,
130.6, 62.1, 58.4, 57.6, 49.5, 35.8, 34.8, 30.5, 29.8, 26.2, 12.0,
2.7, 2.3. COSY, HETCOR, and NOESY on a 500 MHz confirm
the assignments.
Anal. Calcd. for C
H
C, 90.26; H, 9.74. Found: C, 90.12;
14 18
Allyldimethylphenoxysilane (1).
H, 9.40.
The reaction of allylchlorodimethylsilane with phenol under
Acknowledgement.
standard conditions (NEt , CH Cl , -78 °C to room temperature,
3
2
2
We thank Dr. Li Du for running the 500 MHz 2-D nmr experi-
ments and Andrew Diether for setting up the Thin-Film Reactor.
We acknowledge Research Corporation and the BYU
Development Fund for financial support. Instrumental (nmr)
support from the National Science Foundation, Grant No. CHE
8712101 is also acknowledged.
8 hours) gave an 80% yield of allyldimethylphenoxysilane fol-
lowing distillation, b.p. 60 °C/1.0 Torr (Kugelrohr). H nmr (200
MHz, deuteriochloroform): δ 6.8-7.4 (m, 5H, aromatic H), 5.80
1
(m, 1H, vinylic H), 4.95 (m, 2 H, vinylic H), 1.80 (d, J = 8 Hz,
13
2H, allylic H), 0.30 (s, 6H, 2 x Si-CH ); C nmr (50 MHz,
3
CDCl ) 154, 133, 129, 121, 120, 120, 115, 114, 24, -2; ir (neat)
3
Professor Bradshaw has been a terrific colleague and an excel-
lent example. I (SAF) will personally miss his teaching and
2960, 1630, 1596, 1491, 1253, 1164, 1070 cm-1; hrms m/z calc.
192.0970, obs. 192.0966.

1344
S. Bahu, S. A. Fleming, B. Nilsson, and T. Turner
Vol. 38
Marcel Dekker, New York, 1989.
research prowess in the Brigham Young University Department
of Chemistry and Biochemistry. But I look forward to many
more years of interaction.
[4a] A. Gilbert and G. N. Taylor, J. Chem. Soc., Perkin Trans.
1, 1761 (1980); [b] D. M. Amey, D. C. Blakemore, M. G. B. Drew, A.
Gilbert, and P. Heath, J. Photochem. Photobiol., A: Chem., 102, 173
(1997).
[5] D. De Keukeleire, S.-L. He, D. Blakemore, and A. Gilbert,
J. Photochem. Photobiol. A: Chem., 80, 233 (1994).
[6] D. C. Blakemore and A. Gilbert, Tetrahedron Lett., 36,
2307 (1995).
[7a] M. Kumada, K. Tamao, J. I. Yoshida J. Organomet.
Chem., 239, 115 (1982); [b] I. Fleming, R. Henning, H. Plaut, J.
Chem. Soc., Chem. Commun., 29 (1984).
[8] J. Cornelisse, Chemical Reviews, 93, 615 (1993).
[9] M. Newcomb, Tetrahedron, 49, 1151 (1993).
[10] For an example of stable eight-member ring bridgehead
olefins see: M. Toda, K. Okada, and M. Oda, Tetrahedron Lett., 29,
2329 (1988).
REFERENCES AND NOTES
[1a] S. A. Fleming and R. W. Jones, J. Heterocyclic Chem., 27,
1167 (1990); [b] S. A. Fleming and V. D. Bachelder, J. Heterocyclic
Chem., 28, 2061 (1991); [c] S. A. Fleming and S. C. Ward,
Tetrahedron Lett., 33, 1013 (1992); [d] S. C. Ward and S. A. Fleming,
J. Org. Chem., 59, 6476 (1994); [e] C. L. Bradford, S. A. Fleming, and
S. C. Ward, Tetrahedron Lett., 36, 4189 (1995); [f] S. A. Fleming and
J. J. Gao, Tetrahedron Lett., 38, 5407 (1997). For a review see: S. A.
Fleming, C. L. Bradford, and J. J. Gao, Organic Photochemistry;
Molecular and Supramolecular Photochemistry, 1, 187 (1997).
[2a] D. L. Oldroyd and A. Weedon, J. Org. Chem., 59, 1333
(1994); [b] P. A. Wender and T. M. Dore, Tetrahedron Lett., 39, 8589
(1998); [c] N. Hoffmann and J.-P. Pete, Tetrahedron Lett., 38, 5027
(1998); [d] D. L. Comins, Y.-M. Zhang, and X. Zheng, J. Chem. Soc.,
Chem. Commun., 2509 (1998); [e] J. D. Winkler and J. M. Axten, J.
Am. Chem. Soc., 120, 6425 (1998); [f] T. Bach and J. Schröder, J.
Org. Chem., 64, 1265 (1999).
[11] P. De Vaal, E. M. Osselton, E. S. Krijnen, G. Lodder, and
J. Cornelisse, Recl. Trav. Chim. Pays-Bas, 107, 407 (1988).
[12] A. W. H. Jans, B. van Arkel, J. J. van Dijk-Knepper, H.
Mioch, and J. Cornelisse, Tetrahedron, 40, 5071 (1984).
[13] S. A. Fleming and A. W. Jensen, J. Org. Chem., 61, 7040
(1996).
[14] J. Foster, A. L. Pincock, J. A. Pincock, and K. A.
Thompson, J. Am. Chem. Soc., 120, 13354 (1998).
[3] For an excellent review see: P. A. Wender, L. Siggel, and
J. M. Nuss, in Organic Photochemistry, A. Padwa, Ed., Vol. 10,