Remarkable product selectivity during photo-Fries and photo-Claisen rearrangements within zeolites

Full text of DOI:10.1021/ja961620w

9428
J. Am. Chem. Soc. 1996, 118, 9428-9429
Scheme 1
Remarkable Product Selectivity During Photo-Fries
and Photo-Claisen Rearrangements within Zeolites
K. Pitchumani, M. Warrier, and V. Ramamurthy*
Department of Chemistry, Tulane UniVersity
New Orleans, Louisiana 70118
ReceiVed May 14, 1996
Shape selectivity consists in often subtle matching of size
and shape of reactants, transition states, and/or products with
the size and shape of pores, cages, and pore volumes of the
intracrystalline zeolite phase.1 We provide below examples of
how this feature can be a useful tool in controlling product
distribution in a photochemical reaction.2 The reactions we have
chosen to investigate are photo-Fries rearrangement of phenyl
acetate and phenyl benzoate and photo-Claisen rearrangement
of allyl phenyl ether (Scheme 1). We have chosen these
reactions for the following reasons: (a) These reactions have
been extensively investigated in isotropic solution media.3-5 (b)
Attempts have been made earlier to control product distribution
within various ordered media with varied success.6 (c) The
products of the Fries reaction, hydroxyacetophenones and
hydroxybenzophenones, are used in the manufacture of phar-
maceuticals. For example, the para selective rearrangement of
phenyl acetate is a key step in the Hoechst-Celanese process
for the manufacture of 4-acetaminophenol.7 (d) Fries rear-
rangements of aryl esters are often carried out on an industrial
scale with conventional catalysts such as AlCl3 and mineral acids
like HF and H2SO4. In such reactions, selectivity is poor and
the catalyst used generates a large amount of toxic waste
products. A combination of zeolite as a medium and light as
the reagent offer a new approach towards “clean chemistry”.
(e) The goal of achieving selectively a single isomer within
zeolites by thermal Fries reaction has not been realized.8
The internal structure of the two types of zeolites (faujasites
X and Y and pentasils ZSM-5 and ZSM-11) that we have
utilized as media in this study vary in size and shape.
Differences in product selectivity obtained in this study between
faujasites and pentasils as the media reveal a powerful message.
In order to obtain a high product selectivity, one should utilize
a reaction cavity that is large enough to respond to the shape
changes that occur along the reaction coordinate but at the same
time hard and small enough to provide relatively different
extents of restriction on various reaction pathways available to
the reactive intermediates.9
The experimental procedure consisted of stirring known
amounts of 1 or 2 and activated zeolite (X or Y) in hexane,
followed by filtering, washing with excess hexane, and drying
under reduced pressure (10-4 Torr). These dried samples were
irradiated (∼2 h) in hexane as a slurry. For inclusion within
ZSM-5 and ZSM-11, 2,2,4-trimethylpentane was used as the
solvent and the irradiations were conducted as a slurry in either
hexane, 2,2,4-trimethylpentane, or water. Following irradiation,
the products were extracted with ether and analyzed by GC.
Mass balances were excellent (>90%). Spectral data for the
products matched well with the reported data,10 and results are
presented in Table 1.11
One of the most remarkable observations is that in all three
cases the product distribution is altered within zeolites from
that in isotropic solvents. Furthermore, while in solution, nearly
a 1:1 mixture of ortho and para isomers (Scheme 1) is the norm.
Within zeolites, one is able to direct the photoreaction selectively
toward either the ortho or the para products by conducting the
reaction either within faujasites or pentasils, respectively (Table
1). Both photo-Fries and photo-Claisen rearrangements proceed
via a similar mechanism (Scheme 2).4,5 Excitation to the excited
singlet state results in fragmentation of the phenyl esters and
allyl phenyl ether. Cage escape, recombination, and hydrogen
migration results in both ortho and para products. While in
solution, the ratio of the ortho and the para products is
determined by the electron densities at the ortho and para
positions in the phenoxy radical; clearly, this is not the
controlling factor within zeolites. Phenol could form either by
an in-cage or out-of-cage process. We are unable to obtain any
evidence in favor of ketene, which would be an expected product
during the incage process.
Photo-Fries reaction of phenyl acetate and phenyl benzoate
is highly selective within X and Y zeolites. While in Li- and
NaX and -Y zeolites a small percentage (<10%) of the para
isomer is formed, within KX and KY zeolites the ortho isomer
is the exclusive product. Phenol yield reported in the table is
a sum of thermal (dark) and photochemical reactions. Only
for phenyl acetate, the thermal reaction contributed signifi-
cantly.12 The ortho selectivity is truly remarkable. The above
selectivity is not the result of shape exclusion since both the
(1) (a) Weisz, P. B. Pure Appl. Chem. 1980, 52, 2091. (b) Csicsery, S.
M. in Zeolite Chemistry and Catalysis; Rabo, J. A., Ed.; American Chemical
Society: Washington, DC, 1974; p 680. (c) Dwyer, J. Chem. Ind., 1984, 7,
229.
(2) As far as we know, there is only one example of shape selectivity in
photochemical reactions known in the literature and this is provided by
Turro and co-workers: (a) Turro, N. J.; Lei, X.; Cheng, C. C.; Abrams, L.;
Corbin, D. R. J. Am. Chem. Soc. 1985, 107, 5824. (b) Turro, N. J. Pure
Appl. Chem. 1986, 58, 1219. (c) Turro, N. J.; Cheng, C. C.; Abrams, L.;
Corbin, D. R. J. Am. Chem. Soc. 1987, 109, 2449.
(3) (a) Bellus, D. AdV. Photochem. 1971, 8, 109. (b) Bellus, D.; Hrdlovic,
P. Chem. ReV. 1967, 67, 599. (c) Stenburg, V. I. Org. Photochem. 1967, 1,
127.
(4) (a) Anderson, J. C.; Reese, C. B. Proc. Chem. Soc. 1960, 217. (b)
Kobsa, H. J. Org. Chem. 1962, 27, 2293. (c) Arai, T.; Tobita, S.; Shizuka,
H. Chem. Phys. Lett. 1994, 223, 521. (d) Suau, R.; Torres, G.; Valpuesta,
M. Tetrahedron Lett. 1995, 36, 1311.
(5) (a) Kharasch, M. S.; Stampa, G.; Nudenberg, W. Science 1952, 116,
309. (b) Schmid, K.; Schmid, H. HelV. Chim. Acta 1953, 36, 687. (c) Carrol,
F. A.; Hammond, G. S. J. Am. Chem. Soc. 1972, 94, 7151. (d) Adam, W.;
Fischer, H.; Hansen, H. J.; Heimgartner, H.; Schmid, H.; Waespe, H. R.
Angew. Chem., Int. Ed. Engl. 1973, 12, 663.
(6) (a) Ohara, M.; Watanabe, K. Angew. Chem., Int. Ed. Engl. 1975, 14,
820. (b)Avnir, D.; de Mayo, P.; Ono, I. J. Chem. Soc., Chem. Commun.
1978, 1109. (c) Chenevert, R.; Plante, R. Can. J. Chem. 1983, 61, 1092.
(d) Abdel-Malik, M. M.; de Mayo, P. Can. J. Chem. 1984, 62, 1275. (e)
Chenevert, R.; Voyer, N. Tetrahedron Lett. 1984, 25, 500. (f) Syamala, M.
S.; Ramamurthy, V. Tetrahedron 1988, 44, 7223. (g) Syamala, M. S.;
Nageswer Rao, B.; Ramamurthy, V. Tetrahedron, 1988, 44, 7242. (h)
Pitchumani, K.; Devanathan, S.; Ramamurthy, V. J. Photochem. Photobiol.,
A 1992, 69, 201.
(9) Ramamurthy, V.; Weiss, R. G.; Hammond, G. S. AdV. Photochem.
1993, 18, 67.
(10) Products of photo-Fries reaction are all commercially available.
Spectral data of photoproducts were compared with the authentic commercial
samples. Products 6 and 7 are not commercially available but are known
in the literature. Spectral data were compared with literature reports.5
(11) In addition to hexane slurry, solid zeolites were also photolyzed.
Conversion obtained by the latter technique was significantly low. Reflection
as well as nonuniformity of absorption by solid samples contributed to the
low conversion.
(7) Szmant, H. Organic Building Blocks of the Chemical Industry, John
Wiley: New York, 1989; p 504.
(8) (a) van Bekkum, H.; Hoefnagel, A. J.; van Koten, M. A.; Gunnewegh,
E. A.; Vogt, A.; Kouwenhoven, H. W. Stud. Surf. Sci. Catal. 1994, 83,
379. (b) Vogt, A.; Kouwenhoven, H. W.; Prins, R. Appl. Catal., A 1995,
123, 37.
(12) Contribution due to thermal reaction was monitored by analyzing
an identical sample kept in the dark for the same duration as the photolyzed
sample. The dark reaction was significant only for phenyl acetate and the
percentage contribution increased with the cation size (Cs+ > Rb+ > K+).
S0002-7863(96)01620-4 CCC: $12.00 © 1996 American Chemical Society

Communications to the Editor
J. Am. Chem. Soc., Vol. 118, No. 39, 1996 9429
to the selectivity.16 This becomes clear when one compares
the results of phenyl acetate and allyl phenyl ether. In the latter
compound, a small percentage of the para isomer (∼15%)
always accompanied the major ortho isomer. While the size
and shape of acyl and allyl radicals are expected to be similar,
the strength of interaction between the cations and these two
fragments is expected to differ; the former is expected to bind
more strongly than the latter.17 Such a difference, we believe,
is translated into an increased yield of the para isomer for phenyl
allyl ether.
Table 1. Product Distribution upon Photolysis of Phenyl Esters
and Allyl Phenyl Ether
phenyl
acetate
(1a)a,b
allyl
phenyl
benzoate
(1b)a-c
phenyl
ether (2)a,b
3
4a 5a
3
6
7
3
4b 5b
hexane
methanol
benzene
20 53 27 48 21 31 12 49 37
17 44 39 16 42 42
15 54 31
6
6
1
1
2
1
5
7
49 43
53 38
LiY-hexaned
NaY-hexane
KY-hexane
LiX-hexaned
NaX-hexane
KX-hexane
Na ZSM-5-hexanef
Na ZSM-5-TMPg,h
Na ZSM-5-water
7e 87
91
6
4
6
3
4
4
6
3
76 18
79 18
78 18
81 15
78 14
83 14
92
93
98
96
93
93
7
6
5
The shape and size of pentasils are such that only the para
isomers (5 and 7, Scheme 1) fit within the channels of these
zeolites. Indeed, under conditions in which the para isomers 5
and 7 were readily incorporated within ZSM-5 and ZSM-11,
the ortho isomers 4 and 6 remained in the solvent (2,2,4-
trimethylpentane) portion. On the basis of the above shape
selectivity, one would have predicted selective formation of the
para isomers within ZSM zeolites. However, no selectivity was
achieved when the photolysis of 1a and 2 included within
ZSM-5 and ZSM-11 was conducted as a hexane slurry. What
was surprising was the fact that the unexpected ortho isomer
was obtained in significant yield. An insight into the lack of
selectivity was gained when the GC analysis of the hexane layer
prior to the photolysis was performed. This analysis revealed
that hexane displaced the reactants 1a and 2, from the interior
of ZSM. Under these conditions, most of the reactants were
present in the hexane solvent and not inside the zeolite. Absence
of selectivity is the result of reaction occurring in the isotropic
hexane medium. We reasoned that selectivity should be possible
if we can keep the reactants within the channels of ZSM zeolites.
This was achieved by using solvents which will not displace
the included 1a and 2 from the interior of ZSM. Since the
internal structure of ZSM is highly hydrophobic, water is not
expected to fill the channels of ZSM. Similarly, 2,2,4-
trimethylpentane solvent molecules, being too large, are not
expected to enter the channels of ZSM. We visualized that these
solvent molecules would serve as lids on the channels and keep
both the reactants and the reactive fragments within the channels
of ZSM. Thus, the fragments would be forced to recombine in
such a way that the product that fits within the narrow channels
of ZSM would predominate. Indeed, when photolysis was
conducted in these solvents a clear preference for the para isomer
resulted (Table 1). A small amount of the ortho isomer that is
formed, we believe, comes from reactions that occur in the
solvent portion. This is supported by the results obtained with
1b. This molecule, being too large to enter into the internal
structure of ZSM, is expected to remain on the external surface.
Under such conditions, no selectivity is expected and indeed
we did not obtain any.
17 83
15e 77
21 79
27 73
8
3
2
27 35 38 40 31 29 10 34 27
32
32
5
6
8
63
62
74
5
2
7
5
7
Na ZSM-11-TMPg,h 18
Na ZSM-11-water
20 73
27 73
18 12 7j
a Irradiations were conducted to about 30% conversion. This amount
of conversion was reached in 1 h for phenyl benzoate, 2 h for phenyl
acetate, and 4 h for allyl phenyl ether. Ratios of products were
independent of the conversion in the range of 15 to 90%. All yields
presented in the table are an average of at least five independent runs
with an error limit on yields of (2%. b Occupancy level was kept at
∼1 molecule per supercage for X and Y zeolites. Loading level for
ZSM-5 was 5 mg of substrate to 300 mg of ZSM. c Difference between
the reported yield and the total theoretical yield (100%) is due to the
formation of benzoic acid and benzaldehyde. d Ortho selectivity was
also observed when Mg-, Ca-, Sr-, and BaX or -Y zeolites were used.
Results were similar to the ones presented for alkaline metal cation
exchanged X and Y zeolites. e Phenol yield presented here is a sum of
blank thermal reaction and photochemical reaction. Control experi-
ments showed that blank reaction was higher in 1a than in 1b. See ref
f
13 for details. In hexane, the molecules remained either on the external
surface or in the solution. Product distribution obtained is close to
that in hexane in the absence of zeolite. Under such conditions,
contribution due to thermal reaction is very small (<5%). When the
reaction is conducted with TMP as the solvent, phenyl acetate remained
within the zeolite. Under such conditions thermal reaction was
significant (∼26%). g Selectivity did not depend on the Si/Al ratio (24:
400) of ZSM-5 and ZSM-11 used. h TMP refers to 2,2,4-trimethylpen-
tane.
Scheme 2
ortho and the para isomers (4 and 5) fit well within the
supercage. This is indicated by the fact that when these isomers
were included independently, they were readily incorporated
within X and Y zeolites. Also, they can be easily extracted
from the zeolite. Selectivity, we believe, results from the
restriction imposed on the mobility of the phenoxy and the acyl
(benzoyl) fragments by the supercage framework and cations.
Several features contribute to the restriction. The cage free
volume plays an important role, as evident from the increased
selectivity with the increase in size of the cation.13 It is
important to note that when cations larger than K+ were used,
the reaction was extremely slow. This complication resulted
from the heavy cation induced intersystem crossing from the
reactive excited singlet to the unreactive triplet.14,15
We have shown that with the proper choice of medium one
can achieve very high selectivity on photochemical reactions.
Detailed mechanistic studies of the above rearrangements within
zeolites are underway.
Acknowledgment. Authors thank PRF-ACS and the NSF-EPS-
COR-LBR Center for Photoinduced Processes, Tulane University for
financial support. Discussions with R. G. Weiss and A. Weedon have
been very valuable.
JA961620W
(15) Emission spectra of phenyl benzoate changed from predominantly
fluorescence to predominantly phosphorescence when the cation was
changed from Na+ to Cs+. Also, excited singlet lifetime decreased with
the change in cation from Na+ to Cs+.
(16) (a) Fitch, A. N.; Jobic, H.; Renouprez, A. J. Phys. Chem. 1986, 90,
1311. (c) Czjek, M.; Vogt, T.; Fuess, H. Angew. Chem., Int. Ed. Engl. 1989,
28, 770. (c) Hepp, M. A.; Ramamurthy, V.; Corbin, D. R.; Dybowski, C.
J. Phys. Chem. 1992, 96, 2629.
In addition to the cage free volume, an interaction between
the cation and the two reactive fragments must be contributing
(13) (a) Turro, N. J.; Zhang, Z. Tetrahedron Lett. 1987, 28, 5637. (b)
Ramamurthy, V.; Corbin, D. R.; Eaton, D. F.; Turro, N. J. Tetrahedron
Lett. 1989, 30, 5833.
(14) Ramamurthy, V.; Caspar, J. V; Kuo, E. W.; Corbin, D. R.; Eaton,
D. F. J. Am. Chem. Soc. 1992, 114, 3882.
(17) (a) Staley, R. H.; Beauchamp, J. L. J. Am. Chem. Soc. 1975, 97,
5920. (b) Sunner, J.; Nishizawa, K.; Kebarle, P. J. Phys. Chem. 1981, 85,
1814.