Controlling the reactive state through cation binding: Photochemistry of enones within zeolites

Article abstract of DOI:10.1016/S0040-4039(01)00100-9

The nature of the lowest triplet state of enones is altered by the cations present within Y zeolites. Alkali metal ions, such as Li⁺, are predicted to interact with the carbonyl unit of enones in a collinear fashion and significantly lower both the p-type n and π-2 orbitals. Excited state energies, estimated at the CIS(D)/6-31+G* level, show that the lowest triplet is n-π* in character for the enones, but switch to π-π* on coordination with Li⁺. Observed product distribution within zeolite is consistent with this theoretical prediction.

Full text of DOI:10.1016/S0040-4039(01)00100-9

TETRAHEDRON
LETTERS
Pergamon
Tetrahedron Letters 42 (2001) 2079–2083
Controlling the reactive state through cation binding:
photochemistry of enones within zeolites
Sundararajan Uppili,a Shinsuke Takagi,a R. B. Sunoj,b P. Lakshminarasimhan,a J. Chandrasekharb,*
and V. Ramamurthya,*
aDepartment of Chemistry, Tulane University, New Orleans, LA 70118, USA
bDepartment of Organic Chemistry, Indian Institute of Science, Bangalore 560012, India
Received 17 November 2000; accepted 4 January 2001
Abstract—The nature of the lowest triplet state of enones is altered by the cations present within Y zeolites. Alkali metal ions,
such as Li+, are predicted to interact with the carbonyl unit of enones in a collinear fashion and significantly lower both the p-type
n and p-2 orbitals. Excited state energies, estimated at the CIS(D)/6-31+G* level, show that the lowest triplet is n-p* in character
for the enones, but switch to p–p* on coordination with Li+. Observed product distribution within zeolite is consistent with this
theoretical prediction. © 2001 Elsevier Science Ltd. All rights reserved.
Photoreactions of molecules containing a carbonyl
chromophore can occur from both the n-p* and p–p*
excited triplet states if they are fairly proximal so that
an equilibrium exists between these two states at ambi-
ent temperature.1 Since the chemical reactivity of the
two triplet states can differ markedly, the nature and
yield of the products obtained depend on: (a) the
relative rates of reaction from the two excited states,
and (b) the energy gap between the two reactive states.
Solvent polarity strongly influences the relative energies
of n-p* and p–p* states of carbonyl compounds,2 and
can be employed to control the selectivity in product
distribution.3 However, this approach has limited value
because most organic compounds do not dissolve in the
most polar solvent, water. We believe zeolites offer a
convenient polarizable environment4 in which most
organic molecules, insoluble in water, can be readily
included, if the dimensions of the supercage window
zeolites can interact with the included carbonyl
compounds5,6 and potentially alter the excited state
energetics. We now demonstrate that enones, which are
known to give different products from the n-p* and
p–p* triplets in the solution phase,7 can indeed be
directed to react preferentially via the p–p* triplet by
employing a zeolite as the reaction medium. We also
offer computational evidence that Li+ complexation
leads to a reversal of triplet state ordering in enones,
accounting for the observed selectivity.
Triplet sensitization of 3-methyl-3-(1-cyclopentenyl)-
butan-2-one, 1, yields the 1,3-acyl migration product 2
from the n-p* triplet (and n-p* singlet) and the oxa-di-
p-methane product 3 from the p–p* triplet (Scheme
1).7c,d In the present investigation, accumulation of the
1,3-acyl migration product from the excited singlet state
was avoided by using the triplet sensitizer, 4-methoxy-
acetophenone. Triplet sensitization of 1 by 4-methoxy-
,
(diameter ꢀ8 A) permit. Further, cations present in
Scheme 1.
* Corresponding authors. Tel.: 504 862 8135; fax: 504 865 5596; e-mail: murthy@tulane.edu
0040-4039/01/$ - see front matter © 2001 Elsevier Science Ltd. All rights reserved.
PII: S0040-4039(01)00100-9

2080
S. Uppili et al. / Tetrahedron Letters 42 (2001) 2079–2083
acetophenone in hexane gave exclusively the product
from the n-p* triplet, 2. However, in polar solvents,
such as methanol and acetonitrile, a mixture of 2 and 3
was obtained (Scheme 1). The solvent dependence of
product distribution can be rationalized on the basis of
a two-state model.7d In hexane, the lowest triplet of
enone 1 is n-p* in character and the next p–p* triplet is
much higher in energy. The products are derived exclu-
sively from the n-p* triplet. In polar solvents, the p–p*
state is relatively stabilized enabling an equilibrium to
be established with the n-p* triplet, resulting in prod-
ucts from both states. It is of interest to explore the
effect of zeolites in controlling the relative energetics of
n-p* and p–p* states and, in turn, the product distribu-
tion.
established to arise from the n-p* triplet and products
7–9 from the p–p* triplet.9 The ratio of the two sets of
products [(7+8+9)/(5+6)] has been reported to depend
on solvent polarity (Scheme 2).9 Similar to enone 1, in
a non-polar hydrocarbon solvent, products from the
n-p* triplet alone were obtained suggesting that the
lowest triplet is n-p* in character and the second p–p*
triplet is not close enough to establish an equilibrium
and react. With increasing polarity, the two states
apparently are brought closer in energy such that prod-
ucts from both states are formed (Scheme 2). Consis-
tent with the behavior of enone 1, direct irradiation of
4 included within MY and MX zeolites (6 mg of enone
in 300 mg of zeolite, 450 W medium pressure mercury
lamp, Pyrex filter, zeolite–hexane slurry, 5 h irradiation,
15% conversion) gave higher yields of products 7–9
derived from the p–p* triplet (Scheme 2) than in non-
polar benzene (0%) or moderately polar acetonitrile
(42%).10 In LiY the combined yield of [(7+8+9)] was
>85%, even higher than in a 30% water–methanol
mixture (75%). The results obtained in Y-Sil and
MCM-41 (25% of [(7+8+9)]), zeolites with no cations,
reveal the key role of cations in enhancing the yield of
p–p* triplet products.
Enone 1 and the triplet sensitizer 4-methoxyacetophe-
none included (2 mg 1 and 5 mg of sensitizer in 300 mg
of zeolite) in various M+ exchanged X and Y zeolites
were irradiated as a hexane slurry (450 W medium
pressure mercury lamp, Pyrex filter, continuous nitro-
gen bubbling, 15 min, 15% conversion).8 The oxa-di-p-
methane product 3 was obtained in higher yield within
the zeolite than in the non-polar hexane or in other
polar solvents used in this investigation (Scheme 1).
The results obtained in zeolites are consistent with the
p–p* triplet being near the n-p* triplet state. In the
absence of rates and efficiencies of reactions from these
two states we cannot be certain about which of the two
states is lower in energy. The selectivity in favor of the
p–p* triplet product observed in zeolites is unmatched
in any organic solvent, attesting to the uniqueness of
zeolites.
The final examples involve steroidal cyclohexenones,
testosterone acetate, 10a, and cholestenone, 10b
(Scheme 3).11 These enones, upon excitation in hexane,
are unreactive. On the other hand, irradiation of 10a
and b in 2-propanol results in cis- and trans-fused
hydrogen addition products 11 and 12.12 The absence
of reduction products in hexane is consistent with the
expectation that the rate of hydrogen abstraction from
hexane by the triplet enone would be low. On the other
hand, the solvent 2-propanol is a good hydrogen donor
and reduction products as observed are anticipated.
Based on reported studies on related systems, 4a-
methyl - 4,4a,9,10 - tetrahydro - 2(3H) - phenanthrenone,
and 4a-methyl-4,4a,5,6,7,8-hexahydro-2(3H)-naphtha-
The above strategy of controlling product distributions
by inclusion in a zeolite was also attempted with 4-
methyl-4-phenyl-2-cyclohexenone 4. As shown in
Scheme 2, of the several products (5–9) that this
molecule gives upon excitation, 5 and 6 have been
Scheme 2.
Scheme 3.

S. Uppili et al. / Tetrahedron Letters 42 (2001) 2079–2083
2081
Figure 1. From right to left: schematic representations of n, p2 and p* orbitals of cyclopentenone (top) and cyclopentenone–Li+
complex (bottom).
similar to the polarity of bulk water.4 Hence, the
included substrate is exposed to a highly polar or
polarizable environment. By analogy with the known
solvent dependency of photoreactions of enones, the
p–p* triplet state is expected to be preferentially stabi-
lized. In addition, we suggest that the cations present in
zeolites play a direct role. The binding energies for Li+
to formaldehyde and acetone have been experimentally
measured to be ꢀ36 and 45 kcal mol−1, respectively.5
Although the corresponding data for enones are not
available, the values are likely to be in a similar range.
At the MP2/6-31G* level,14 we compute the binding
energies of Li+ to cyclopentenone and cyclohexenone to
be 54 and 54.5 kcal mol−1, respectively (for reference,
the corresponding value for acetone is computed to be
48 kcal mol−1). Although the strength of interaction is
likely to be reduced due to the presence of oxyanionic
counterions, enones adsorbed within a zeolite are
expected to be bound to M+ ions. We therefore probed
the effect of metal ion complexation on the orbital and
excitation energies of the model systems, cyclopen-
tenone and cyclohexenone, along with acetone for
comparison.
lenone,13 we assume that products 11 and 12 arise from
the p–p* and n-p* triplet states, respectively. In 2-
propanol the major products from 10a and b were 12a
and b, respectively, derived from the n-p* triplet. For-
mation of 11 along with 12 suggests that in 2-propanol
the p–p* and n-p* triplets are near one another and
hydrogen abstraction by both states occur. Thus, these
molecules are, once again, ideal examples for exploring
the use of zeolites to alter the energetic ordering of n-p*
and p–p* triplet states. As shown in Scheme 3, when
the irradiation was conducted in NaY (5 mg of enone
in 300 mg of zeolite, 450 W medium pressure mercury
lamp, Pyrex filter, zeolite–hexane slurry, continuous
nitrogen bubbling, 5 h irradiation, 15% conversion),10
consistent with the behavior of enones 1 and 4, the
major product switched from 12 to 11 (from n-p* to the
p–p* triplet product). Interestingly, the steroidal enones
10a and b abstract hydrogen from the solvent hexane
only when they are present within the NaY zeolite. The
absence of photoreduction of 10 in Y-Sil and MCM-41
zeolites suggests that cations are required for hydrogen
abstraction from the co-guest hexane.
The observed changes in product distributions, shown
in Schemes 1–3, reflect the remarkable influence of the
zeolite medium on the location of the n-p* and p–p*
triplets of enones. A simple interpretation involves the
local polarity within zeolites. Based on measurement
with at least five independent probes, the micropolarity
of M+Y and M+X zeolites has been determined to be
As in earlier studies on simple carbonyl compounds,15
the Li+ ion is computed to be aligned nearly collinear
with the CꢀO bond, suggesting a primarily ion–dipolar
electrostatic interaction between the metal ion and the
enone (Fig. 1). While the nature and relative coeffi-
cients of the MOs are not altered in any significant

2082
S. Uppili et al. / Tetrahedron Letters 42 (2001) 2079–2083
Table 1. Ground state orbital energies (HF/6-31G*) and energies of triplet states relative to the ground state (CIS(D)/6-31+
G*) for carbonyl compounds and their Li+ complexes
Molecule/ion
Orbital energy (eV)
n
Triplet energy (eV)a
p
p*
4.28
−2.08
2.93
−1.99
2.80
−2.02
n-p*
p–p*
Acetone
−13.03
−18.45
−10.21
−14.31
−10.08
−14.03
−11.19
−16.67
−10.86
−15.99
−10.90
−15.74
4.04 (4.41)
4.51 (5.16)
3.69 (4.33)
4.35 (5.29)
3.50 (4.13)
4.15 (5.11)
6.28 (5.19)
Acetone+Li+
6.63 (6.38)
4.23 (3.36)
4.10 (3.43)
4.05 (3.17)
3.84 (3.19)
Cyclopentenone
Cyclopentenone+Li+
Cyclohexenone
Cyclohexenone+Li+
a Results obtained at CIS/6-31+G* level (without doubles corrections) are given in parentheses.
Acknowledgements
manner, all the MOs are shifted to lower energies
through coordination. The key MOs of importance in
the present context are the p-type n orbital on the
The authors at Tulane University thank the Division of
Chemical Sciences, Office of Basic Energy Sciences, US
Department of Energy for support of this program.
R.B.S. thanks the CSIR (New Delhi) for a research
fellowship. V.R. and P.H. thank Jing Kong of Q-Chem
and Rajeev Bendale of Hyperchem for their help and
tips during the installation of the programs.
carbonyl oxygen, the filled p (higher lying p2 for the
enones) and the vacant p* orbitals (shown in Fig. 1 for
cyclohexenone). The n orbital is stabilized by Li+ com-
plexation to a greater extent than the p MO in the
model enones (Table 1), suggesting that the n-p* triplet
will be relatively shifted to higher energy due to cation
binding. However, orbital energies are not always a
reliable guide for deriving excited state energetics. In
order to obtain more accurate estimates, configuration
interaction calculations were carried out using the
MP2/6-31G* optimized geometries. After including all
single excitations (CIS), the effect of double excitations
was taken into account in a size-consistent second-
order perturbative treatment.16 These CIS(D) calcula-
tions were carried out using the larger 6-31+G* basis
set. Since the two triplet states of interest and the
ground state singlet correspond to the lowest roots of
their spin and spatial symmetry types, the relative ener-
gies are likely to be fairly reliable.
References
1. Wagner, P. J.; Park, B.-S. Org. Photochem. 1991, 10, 227.
2. Jaffe, H. H.; Orchin, M. Theory and Applications of
Ultraviolet Spectroscopy; John Wiley: New York, 1962;
Chapter 9.
3. (a) Lamola, A. A. J. Chem. Phys. 1967, 47, 4810; (b)
Rauh, R. D.; Leermakers, P. A. J. Am. Chem. Soc. 1968,
90, 2246; (c) Yang, N. C.; McClure, D. S.; Murov, S. L.;
Houser, J. J.; Dusenbury, R. J. Am. Chem. Soc. 1967, 89,
5466; (d) Wagner, P. J.; May, M. J.; Haug, A.; Graber,
D. R. J. Am. Chem. Soc. 1970, 92, 5269.
The CIS(D)/6-31+G* calculations confirm that the n-p*
triplet is the lowest energy triplet in the three model
systems (Table 1). While this is expected for acetone on
the basis of orbital energies, the trend prevails in the
enones in spite of the fact that the n orbital is below the
p2 HOMO. Interestingly, the greater stability of the
n-p* triplet, which is consistent with experimental
trends noted above, is obtained only after the inclusion
of doubles corrections. More significant in the present
context is the effect of Li+ coordination on the energies
of the triplet states. While the n-p* triplet is clearly
shifted to higher energy, the p–p* triplet is marginally
stabilized in the enones. The lower energy triplet is now
calculated to be the p–p* state. The switch in the
ordering of the triplet states and their relative energies
are both qualitatively consistent with the observed
product selectivities in photoreactions of enones in
zeolites.
4. (a) Uppili, S.; Thomas, K. J.; Crompton, E. M.; Rama-
murthy, V. Langmuir 2000, 16, 265; (b) Ramamurthy, V.;
Sanderson, D. R.; Eaton, D. F. Photochem. Photobiol.
1992, 56, 297.
5. (a) Woodin, R. L.; Beauchamp, J. L. Chem. Phys. 1979,
41, 1; (b) Staley, R. H.; Beauchamp, J. L. J. Am. Chem.
Soc. 1975, 97, 5920.
6. For a summary of cation effects on carbonyl photochem-
istry, see: Fukuzumi, S. Ad. Photochem. 1999, 25, 154–
172.
7. (a) Schuster, D. I. Rearrangements in Ground and Excited
States; Academic Press: New York, 1980; Vol. 3, pp.
167–279; (b) Schuster, D. I. In The Chemistry of Enones;
Patai, S.; Rappoport, Z., Eds.; John Wiley & Sons: New
York, 1989; pp. 623–756; (c) Engel, P. S.; Schexnayder,
M. A. J. Am. Chem. Soc. 1975, 97, 145; (d) Schuster, D.
I.; Calcaterra, L. J. Am. Chem. Soc. 1982, 104, 6397.
8. The general procedure of inclusion of enones within
cation-exchanged X and Y zeolites consisted of stirring
known amounts of enone (2 mg) and 4-methoxy-
aceotophenone (5 mg) with activated (500°C) zeolite (300
mg) in hexane. The slurry was filtered, washed with
hexane to remove the ketone adsorbed on the surface of
the zeolite and dried on a vacuum line (10−3 torr) at room
In summary, the high polarity and cation–carbonyl
interactions enable the zeolite medium to control the
product distribution from photoreactions of enones by
influencing the energetic ordering of the reactive excited
states.

S. Uppili et al. / Tetrahedron Letters 42 (2001) 2079–2083
2083
temperature. The zeolite was transferred to dry hexane
inside a dry box, sealed with rubber septa, bubbled with
dry nitrogen for 30 min (outside the dry box) and irradi-
ated (>300 nm) for 15 min. The products were extracted
by stirring the zeolite with dichloromethane and analyzed
by GC. The mass balance, as estimated by GC using a
calibration compound, was ꢀ85%. X and Y zeolites used
in this study were obtained from the PQ Corporation,
Valley Forge, PA. The source of the zeolites was Zeolyst
International, The Netherlands.
S.; Lustig, S. R.; Ramamurthy, V. J. Am. Chem. Soc.
1998, 120, 2480.
13. (a) Chan, A. C.; Schuster, D. I. J. Am. Chem. Soc. 1986,
108, 4561; (b) Bellus, D.; Kearns, D. R.; Schaffner, K.
Helv. Chim. Acta 1969, 52, 971.
14. Calculations were carried out using the Gaussian and
Hyperchem-Q-Chem series of programs: (a) Frisch, M. J.
et al. Gaussian 94, Revision C.2; Gaussian, Inc.: Pitts-
burgh, PA, 1995; (b) Head-Gordon, M.; Pople, J. A.;
Frisch, M. J. Chem. Phys. Lett. 1988, 153, 503; (c) White,
C. A. et al. Q-Chem, Version 1.2; Q-Chem, Inc.: Pitts-
burgh, PA, 1998.
9. Dauben, W. G.; Spitzer, W. A.; Kellogg, S. M. J. Am.
Chem. Soc. 1971, 93, 3674.
10. The experimental procedure was identical to the one
described in Ref. 8 except that no sensitizer was used.
11. (a) Chapman, O. L.; Rettig, T. A.; Griswold, A. A.;
Dutton, A. I.; Fitton, P. Tetrahedron Lett. 1963, 2049; (b)
Nann, B.; Gravel, D.; Schorta, R.; Wehrli, H.; Schaffner,
K.; Jeger, O. Helv. Chim. Acta 1963, 46, 2473; (c) Cor-
nell, D. G.; Avram, E.; Filipescu, N. Steroids 1979, 33,
485; (d) Shoulders, B. A.; Kwie, W. W.; Klyne, W.;
Gardner, P. D. Tetrahedron 1965, 21, 2973; (e) Bellus, D.;
Kearns, D. R.; Schaffner, K. Helv. Chim. Acta 1969, 52,
971.
15. (a) Smith, S. F.; Chandrasekhar, J.; Jorgensen, W. L. J.
Phys. Chem. 1982, 86, 3308; (b) Raber, D. J.; Raber, N.
K.; Chandrasekhar, J.; Schleyer, P.; v., R. Inorg. Chem.
1984, 23, 4076; (c) Ha, T.; Wild, U. P.; Ku¨hne, R. O.;
Loesch, C.; Schaffhauser, T.; Stachel, J.; Wokaun, A.
Inorg. Chem. 1978, 61, 1193; (d) Del Bene, J. E. Chem.
Phys. 1979, 40, 329; (e) Del Bene, J. E.; Frisch, M. J.;
Raghavachari, K.; Pople, J. A.; Schleyer, P.; v., R. J.
Phys. Chem. 1983, 87, 73.
16. (a) Head-Gordon, M.; Rico, R. J.; Oumi, M.; Lee, T. J.
Chem. Phys. Lett. 1994, 219, 21; (b) Head-Gordon, M.;
Maurice, D.; Oumi, M. Chem. Phys. Lett. 1995, 246, 114.
12. Jayathirtha Rao, V.; Uppili, S.; Corbin, D. R.; Schwarz,
.