Intrazeolite photooxidations of electron-poor alkenes.

Article abstract of DOI:10.1021/jo025657c

The inability of the intrazeolite environment to influence the regiochemistry of addition of singlet oxygen to electron-poor olefins is reported. This divergent behavior with respect to electron-rich alkenes is rationalized with a multicomplexation model that emphasizes the importance of intrazeolite substrate-cation binding.

Full text of DOI:10.1021/jo025657c

J . Org. Chem. 2002, 67, 3975-3978
3975
Sch em e 1. Com p a r ison betw een th e In tr a zeolite
a n d Solu tion P h otooxid a tion s of
In tr a zeolite P h otooxid a tion s of
Electr on -P oor Alk en es
(E)-1,1,1-Tr id eu ter o-2-m eth yl-2-p en ten e
Edward L. Clennan,* J akub P. Sram, Andrea Pace,†
Katie Vincer, and Sophia White
Department of Chemistry, University of Wyoming,
Laramie Wyoming 82071
clennane@uwyo.edu.
Received February 26, 2002
Abstr a ct: The inability of the intrazeolite environment to
influence the regiochemistry of addition of singlet oxygen
to electron-poor olefins is reported. This divergent behavior
with respect to electron-rich alkenes is rationalized with a
multicomplexation model that emphasizes the importance
of intrazeolite substrate-cation binding.
Sch em e 2. Mech a n istic Mod el for th e In tr a zeolite
P h otooxid a tion s of Alk en es
The plethora of natural products containing the per-
oxide linkage and their unique biological activity imparts
a sense of urgency to the development of new stereo- and
regiospecific routes to these unusual materials.1 Singlet
oxygen provides an elegant approach for the introduction
of the peroxide linkage. However, despite the promise of
this protocol, the small size of singlet oxygen is a
significant obstacle to its implementation since many of
the traditional techniques used to control stereo- and
regioselectivity do not work.2 Nevertheless, the recent
introduction of zeolites as reaction media for photooxi-
dations provides a potential solution for this enigma.3
These supramolecular systems have the very attractive
feature of being able to impart an “enzyme-like” organi-
zation to the encapsulated activated complexes.
Previous studies in our laboratory and in other labo-
ratories have demonstrated that intrazeolite singlet
oxygen ene reactions exhibit a unique regioselectivity for
hydrogen abstraction from the least substituted side and
the most substituted end of trialkyl-substituted alkenes.4-7
This phenomenon is illustrated for (E)-1,1,1-trideutero-
2-methyl-2-pentene in Scheme 1.8 In this case, the
selectivity for hydrogen abstraction from the most sub-
stituted end increases from 56 to 94% and that for the
least substituted side increases from 8 to 33% as the
photooxidation is moved from CD3CN to the interior of
the zeolite NaMBY.
large extent by the relative stabilities of the diastereo-
meric sodium complexed perepoxides A and B. Other
factors, however, such as the charge distribution on the
carbon framework and the alignment of allylic substit-
uents induced by π-complexation of the alkene also play
regiochemical-defining roles. For example, the preference
for perepoxide intermediate B to abstract a proton from
the gem-dimethyl end of the alkene reflects the prefer-
ential stabilization of the positive charge on the more
substituted end of the alkene and the inaccessibility of
the methylene hydrogens from the face of the olefin
approached by singlet oxygen.
We report here the first examples of the intrazeolite
photooxidations of electron-poor alkenes 1, 4, and 5
(Scheme 3). These alkenes do not exhibit the same
dramatic regiochemical enhancements observed with
electron-rich systems8 but do provide new insight into
the unique environment of the zeolite supercage.
The photooxidations of alkenes 1, 4, and 5, and for
comparison carbonyl-substituted but electron-rich alk-
enes, 8 and 11, were conducted in NaMBY/hexane
slurries as described previously.8 Gas chromatographic
To provide an explanation for the unique regiochemical
influence of the zeolite supercage, the model depicted in
Scheme 2 was proposed.8 This model suggests that the
zeolite-induced regiochemical change is determined to a
† Current address: Dipartimento di Chimica Organica “E. Paterno`”
Universita` degli Studi di Palermo, Viale delle Scienze, Parco d’Orleans
II, 90128, Palermo, Italy.
(1) Dussault, P. H.; Liu, X. Org. Lett. 1999, 1, 1391-1393.
(2) Clennan, E. L. Tetrahedron 2000, 56, 9151-9179.
(3) Li, X.; Ramamurthy, V. J . Am. Chem. Soc. 1996, 118, 10666-
10667.
(4) Clennan, E. L.; Sram, J . P. Tetrahedron Lett. 1999, 40, 5275-
5278.
(5) Robbins, R. J .; Ramamurthy, V. J . Chem. Soc., Chem. Commun.
1997, 1071-1072.
(6) Ramamurthy, V.; Lakshminarasimhan, P.; Grey, C. P.; J ohnston,
L. J . J . Chem. Soc., Chem. Commun. 1998, 2411-2424.
(7) Stratakis, M.; Froudakis, G. Org. Lett. 2000, 2, 1369-1372.
(8) Clennan, E. L.; Sram, J . P. Tetrahedron 2000, 56, 6945-6950.
10.1021/jo025657c CCC: $22.00 © 2002 American Chemical Society
Published on Web 05/01/2002

3976 J . Org. Chem., Vol. 67, No. 11, 2002
Notes
Sch em e 4. In tr a zeolite La cton e F or m a tion
Sch em e 3. P r od u cts fr om In tr a zeolite
P h otooxid a tion s of Electr on -P oor Alk en es
the formation of 14 in the zeolite even at very short
contact times, and as a result, the 71% of 6 reported in
Scheme 3 really represents the sum of 6 and 14. In
contrast, the peroxy-lactone 14 was not formed during
photooxidations of 5 in solution. This provided us with
the opportunity to verify the ability of NaMBY to catalyze
cyclization. An 87/13 mixture of 6 and 7 formed in
solution was added to a hexane slurry of NaMBY ( S
MB
) 0.01) and allowed to stir for 1 h. The peroxide mixture
was then recovered with a high-mass balance by continu-
ous extraction with diethyl ether to give 60% of 6, 31%
of 14, and 9% of 7.
The peroxy-lactones analogous to 14 were not formed
at all in the intrazeolite reaction of 1 and only as a trace
product during intrazeolite photooxidation of 4. We
attribute the unique ability of 6 to form substantial
amounts of the peroxy-lactone to the gem-dimethyl
(Thorpe-Ingold) effect, which promotes ring formation.10
The acid-catalyzed cyclization of a â-hydroperoxy acid
similar to 6 that formed during photooxidation of tiglic
acid has previously been observed.11 This suggests that
residual acid sites in the zeolite might be responsible for
the cyclization.12
Alkenes 1, 4, 5, 8, and 11 exhibited a wide spectrum
of behavior. The carbonyl compounds, 8 and 11, react
with a dramatic change in regiochemistry as the photo-
oxidation is moved from solution into the zeolite remi-
niscent of the behavior exhibited by the electron-rich
trialkyl-substituted alkenes previously investigated.
(Scheme 1) At the other end of the spectrum, the product
distributions during photooxidations of 1 and 4 were
nearly identical in solution and in the zeolite. Finally,
alkene 5 exhibited intermediate behavior producing an
easily measurable shift in ene regiochemistry to give an
enhanced intrazeolite yield of the allylic hydroperoxide
formed by hydrogen abstraction from the gem-dimethyl
group.
To account for these observations we suggest the
multicomplexation mechanism depicted in Scheme 5. In
this model, we postulate that several species are in
dynamic equilibrium in the zeolite supercage. The un-
complexed species, C, is in a solutionlike homogeneous
environment floating free in the supercage void. The
cation-complexed species can either be bound through
their olefinic linkages, D, or through their carbonyl
linkages, E.13 We do not include in this mechanism any
substrate bound to the external surface or outside the
zeolite in the hexane slurry because we have previously
demonstrated that singlet oxygen does not diffuse out,
and 1, 4, and 5 exist exclusively in the zeolite supercage.14
analysis of the hexane as a function of time prior to
irradiation demonstrated that 100% of the alkenes 1, 4,
and 5 intercalated into the zeolite within 15 min.
2-Methyl-2-heptene, for comparison, achieved an equi-
librium distribution (80% hexane)/(20% zeolite supercage)
after approximately 30 min.9 Despite the greater in-
trazeolite affinity of the carbonyl-containing substrates,
the photooxidations were sluggish and required both
extended irradiation times and larger doping levels of the
sensitizer Methylene Blue (MB) ( S
) 0.01 molecules
MB
of MB per supercage) to achieve comparable percent
conversions. Over-photooxidation and/or peroxide decom-
position was absent in all cases. In addition, the products
were extracted from the zeolite with excellent mass
balances in excess of 94%. The products were identified
by comparison to authentic materials, and their ratios
were determined by proton NMR spectroscopy. Each
product ratio represents the average of at least two
independent determinations.
The intrazeolite product distributions for 1, 4, 5, 8, and
11 are compared to the solution results in Scheme 3. The
allylic hydroperoxide, 6, formed in the photooxidation of
5 is very susceptible to zeolite-catalyzed closure to the
R-methylene-â-peroxylactone, 14 (Scheme 4). The extent
of cyclization is a sensitive function of the residence time
in the zeolite. We were unable to completely eliminate
(10) Eliel, E. L.; Wilen, S. H. Stereochemistry of Organic Compounds;
J ohn Wiley & Sons: New York, NY, 1994.
(11) Adam, W.; Griesbeck, A. Angew. Chem., Int. Ed. Engl. 1985,
24, 1070-1071.
(12) Rao, V. J .; Perlstein, D. L.; Robbins, R. J .; Lakshminarasimhan,
P. H.; Kao, H.-M.; Grey, C. P.; Ramamurthy, V. J . Chem. Soc., Chem.
Commun. 1998, 269-270.
(9) The affinity for the interior of the zeolite appears to be in part
a function of the cation-substrate interaction. For leading references
to cation-π interactions, see: Ma, J . C.; Dougherty, D. A. Chem. Rev.
1997, 97, 1303-1324.
(13) This does not imply that complexes D or E exist at unique sites
since it is well established that the cations (Na+) accessible for
complexation in NaY occupy two distinct sites. Barthomeuf, D. Catal.
Rev. 1996, 38, 521-612.

Notes
J . Org. Chem., Vol. 67, No. 11, 2002 3977
Sch em e 5. In tr a zeolite Mu lticom p lexa tion Mod el
for th e P h otooxid a tion s of Alk en es
Sch em e 6. P er ep oxid es fr om In tr a zeolite
P h otooxid a tion of 5
expected to decrease the nucleophilicity of the alkene. The
rate constants k1 and k3 for substrates 8 and 11, which
have methylene groups as insulators between the car-
bonyl and olefinic linkage, are expected to be comparable
and both larger than k2.
The absence of any regiochemical influence of the
zeolite during photooxidations of the very electron-poor
R,â-unsaturated carbonyl compounds 1 and 4 can be
rationalized using two alternative explanations. (1) All
three species, C, D, and E, are populated, and k3[E] is
smaller than the dominant k1[C] process despite the
larger population of E. This analysis is based on the
suggestion that reaction of C will give “solutionlike”
regiochemistry,20 while reaction of complex D or E will
give “zeolite-unique”21 regiochemistry (vide infra). (2)19
Only complexes D and E are significantly populated with
k3[E] > k2[D] (i.e., k1[C] is too small to have any
influence); complex D gives zeolite-unique regiochemis-
try, while complex E gives solutionlike regiochemistry
since the cation is not associated with the double bond.
On the other hand, the real and measurable regio-
chemical influence of the zeolite environment on the
photooxidation of tetrasubstituted alkene 5 suggests that
k2[D] is large enough relative to k1[C] and k3[E] to affect
the product distribution. This can be attributed to an
increase in the population of complex D for the more
nucleophilic tetrasubstituted alkene 5 in comparison to
the less nucleophilic trisubstituted alkenes 1 and 4.
Addition of singlet oxygen to D during photooxidation
of 5 is expected to give both perepoxides F and G.
(Scheme 6) Both of these perepoxides are expected to
decompose by cleavage of the C1-O bond because the
C2-O bond is shorter and stronger in order to minimize
the ability of the carbonyl group to electrostatically
destabilize the incipient positive charge on the carbon
framework. This will result in enhanced hydrogen ab-
straction from the gem-dimethyl group to give the
experimentally observed results (Scheme 3).
In theory substrates in all three complexation modes
depicted in Scheme 5 can react with singlet oxygen. Since
interconversion between these complexes is expected to
be a rapid process,15 the Curtin Hammett Principle/
Winstein-Holness Equation applies.16 The observed rate
constant for reaction will be a weighted average as given
by eq 1 where ni is the mole fraction of the ith complex-
ation mode and ki is the rate constant for reaction of
singlet oxygen with that complex. Consequently, the
products derived from each complex will be a function of
the population of that complex and the magnitude of ki.
kobs
)
n k
(1)
∑
i
i
i
The population distribution is expected to be complex
E > complex D > C. This conclusion is based on limited
binding energy data that suggest that binding is thermo-
dynamically favorable and that binding to a carbonyl
group is more favorable than binding to an olefinic
linkage. For example, the interaction of Li+ with ethylene
(24.3 kcal/mol) is approximately 21 kcal/mol less favor-
able than with acetone.17,18
As pointed out earlier, the regiochemical effect of the
zeolite environment on the photooxidations of 8 and 11
is reminiscent of the behavior of electron-rich trisubsti-
tuted alkenes where complexation to the olefinic linkage
has been invoked (Scheme 3). This observation along with
the preferred population of site E (Scheme 5) suggests
(19) We thank a reviewer for his/her thoughtful comments on the
regiochemistry of the reaction.
The reaction rate constants k2 and k3 are expected to
be smaller than k1 for 1, 4, and 5 since complexation is
(20) Reactions of singlet oxygen with “C” in the photooxidations of
unfunctionalized alkyl-substituted alkenes8 might also be responsible
for the less than perfect regiochemistries observed in their reactions.
(21) We suggest that the zeolite-unique regiochemistry associated
with site E would be an increase in geminal hydrogen abstraction to
give more 2 rather than less 2 as observed. This is based upon the
suggestion of Stratakis and Orfanopoulos (Stratakis, M.; Orfanopoulos,
M. Tetrahedron 2000, 56, 1595-1615) that geminal hyrogen abstrac-
tion in R,â-unsaturated carbonyl compounds is enhanced by the
incipient conjugation in the ene product. This incipient conjugation
should have an even greater effect when sodium is complexed to the
carbonyl group.
(14) Sram, J. P. Ph.D. Dissertation, University of Wyoming, Laramie,
WY, 2001.
(15) Bo¨hlmann, W.; Michel, D.; Roland, J . Magn. Reson. Chem. 1999,
37, S126-S134.
(16) Eliel, E. L.; Wilen, S. H. Stereochemistry of Organic Compounds;
J ohn Wiley & Sons: New York, NY, 1994; p 651-654.
(17) Staley, R. H.; Beauchamp, J . L. J . Am. Chem. Soc. 1975, 97,
5920-5921.
(18) Caldwell, J . W.; Kollman, P. A. J . Am. Chem. Soc. 1995, 117,
4177-4178.

3978 J . Org. Chem., Vol. 67, No. 11, 2002
Notes
(E)-4-Hyd r op er oxy-4-m eth yl-2-p en ten yl Aceta te (9): 1H
NMR (CDCl3) δ 1.36 (s, 6H), 2.09 (s, 3H), 4.59 (d, J ) 6 Hz, 2H),
5.78 (dt, J ) 6, 16 Hz, 1H), 5.88 (d, J ) 16 Hz, 1H), 7.51(s, 1H).
Sch em e 7. Bid en ta te Com p lexa tion
3-Hyd r op er oxy-4-m eth yl-4-p en ten yl Aceta te (10): 1H
NMR (CDCl3) δ 1.76 (s, 3H), 1.84-1.95 (m, 2H), 2.06 (s, 3H),
4.11-4.18 (m, 2H), 4.44 (t, J ) 6 Hz, 1H), 5.05 (bs, 2H), 8.03
(s, 1H).
(E)-6-Hydr oper oxy-6-m eth yl-4-h epten -2-on e (12): 1H NMR
(CDCl3) δ 1.35 (s, 6H), 2.18 (s, 3H), 3.17 (d, J ) 7 Hz, 2H), 5.65
(d, J ) 16 Hz, 1H), 5.80 (dt, J ) 7, 16 Hz, 1H), 9.34 (s, 1H).
5-Hyd r op er oxy-6-m eth yl-6-h ep ten -2-on e (13): 1H NMR
(CDCl3) δ 1.66-1.77 (m, 4H), 1.79 (s, 3H), 2.13 (s, 3H), 4.46-
4.44 (m, 1H), 4.97 (s, 1H), 5.00 (s, 1H), 8.84 (s, 1H).
that 8 and 11 have enough flexibility to allow complex-
ation of the cation to the carbonyl and simultaneous
stabilization of the incipient perepoxide (Scheme 7).
Examination of molecular models verifies that the length
of the tether is sufficient for this bidentate complexation
of sodium. It is not clear, however, if substrates 8 and
11 serve as bidentate ligands for sodium or if, as the
perepoxide begins to develop, the sodium migrates to the
position that ensures the most thermodynamically favor-
able perepoxide cation complex. Stratakis and Froudakis7
have suggested that prior association of sodium to the
alkene is necessary to observe unique regiochemistry of
intrazeolite photooxidations. Unfortunately, no experi-
mental data allowing a choice between these alternatives
are available.
These results indicate that the role of the interstitial
cation of preorganizing the alkene and the relative rates
of reaction at the various intrazeolite complexes are both
important in determining the regiochemical outcome of
these reactions.22 These results provide important infor-
mation for the design of zeolitic environments to control
the regio- and stereochemistry of organic reactions.
Dop in g of Na Y w ith Meth ylen e Blu e. NaY was added to
distilled water containing Methylene Blue. The occupancy ( S
) 0.01; number of methylene blue molecules per supercage) was
calculated assuming a composition of the unit cell of Na56
-
(AlO2)56(SiO2)136‚253H2O with eight supercages per unit cell. The
mixture was stirred overnight in the dark. The water was
decanted, and the colored zeolite was filtered and washed with
water and then air-dried. The NaMBY was dried at 100-120
°C for 8-10 h at 10-4 Torr on a vacuum line and then stored in
a desiccator prior to use.
Solu tion P h otooxid a tion s. CD3CN or (CD3)2CO solutions
0.05 M in alkenes 1, 4, 5, 8, or 11 and 2 × 10-4 M in Methylene
Blue were presaturated with oxygen for 15 min and then
irradiated for 30 min with a 600 W tungsten-halogen lamp
through 1 cm of a 12 M NaNO2 filter solution (cutoff 400 nm)
under continuous oxygen flow. The reactions were monitored by
proton NMR, and the product ratios were determined by
integration of appropriate peaks. The product ratios are repro-
ducible within (5%.
In tr a zeolite P h otooxid a tion s. A 300 mg sample of NaMBY
was added to 5 mL of hexane followed by the addition of
sufficient alkene to bring its concentration to between 0.012 and
0.018 M. This mixture was then stirred under a constant stream
of oxygen for 15 min followed by irradiation for 60 min through
a 400 nm solution cutoff filter. The reaction mixtures were then
centrifuged to separate the zeolite. The hexane wet zeolite
powder was then placed in a Soxhlett cup and extracted for a
minimum of 2 h with diethyl ether. The product ratios were
determined by proton NMR after careful removal of the diethyl
ether immediately following the Soxhlett extraction.
Exp er im en ta l Section
Alkenes 1,23 5,23 and 824 were prepared and purified as
reported in the literature. The allylic hydroperoxides, 2, 3, 6,
and 7, were too sensitive for isolation; however, their 1H NMR
spectra are consistent with their structures and with the data
reported in the literature.23 Compounds 4 and 11 were purchased
commercially and used without further purification.
(22) The change of the cation has a dramatic effect on the regio-
chemistry of the ene reactions of simple alkenes consistent with its
role in this reaction. Shailaja, J .; Sivaguru, J .; Robbins, R. J .;
Ramamurthy, V.; Sunoj, R. B.; Chandrasekhar, J . Tetrahedron 2000,
56, 6927-6943.
Ack n ow led gm en t. We thank the National Science
Foundation and the donors of the Petroleum Research
Fund, administered by the American Chemical Society,
for their generous support of this research.
(23) Adam, W.; Richter, M. J . Tetrahedron Lett. 1993, 34, 8423-
8426.
(24) Badet, B.; J ulia, M.; Mallet, J . M.; Schmitz, C. Tetrahedron
1988, 44, 2913-2924.
J O025657C