Supramolecular effects on the dynamics of radicals in MFI zeolites: A direct EPR investigation

Article abstract of DOI:10.1021/jo020141e

Photolysis of the supramolecular complexes (dibenzyl ketones@ZSM-5) produced supramolecular complexes of benzyl radicals@ZSM-5, which were directly detected by CW-EPR spectroscopy, and provided information on the dynamics of the radicals. The lifetimes of

Full text of DOI:10.1021/jo020141e

Su p r a m olecu la r Effects on th e Dyn a m ics of Ra d ica ls in MF I
Zeolites: A Dir ect EP R In vestiga tion †
Nicholas J . Turro,* Steffen J ockusch, and Xue-Gong Lei
Chemistry Department, Columbia University, 3000 Broadway, New York, New York 10027
turro@chem.columbia.edu
Received February 28, 2002
Photolysis of the supramolecular complexes (dibenzyl ketones@ZSM-5) produced supramolecular
complexes of benzyl radicals@ZSM-5, which were directly detected by CW-EPR spectroscopy, and
provided information on the dynamics of the radicals. The lifetimes of the radicals increased as
the group X attached to the carbon atom at the radical center increases from X ) H (t1/2 ca. 2 min)
to X ) (CH2)4CH3 (t1/2 > 200 min). In addition, line broadening of the EPR signal was observed as
the group X increases. Experiments involving cation-exchanged zeolites (MZSM-5; M ) Li, Na, K,
Rb, Cs) showed a strong dependence of the radical lifetime on the size of the cation (t1/2 ca. 10 min
for Li and t1/2 > 200 min for Cs). The results are discussed in terms of supramolecular steric effects
on the radical-radical reactions in the zeolite supercages.
SCHEME 1
In tr od u ction
In inert solvents and in the absence of scavengers the
lifetimes of reactive radical intermediates are limited by
self-reaction, i.e., radical-radical coupling or radical-
radical disproportionation.1 Classically, radical-radical
reactions may be minimized or even eliminated by
severely reducing diffusional motion or by imposing
significant steric effects on the radical center.2 Carbon-
centered radicals produced by the photolysis of ketones
adsorbed on MFI zeolites are sufficiently persistent at
room temperature that they may be readily characterized
by steady-state CW EPR.3-5 Although a number of
studies of the EPR of carbon-centered radicals have been
reported, there have been no reports on the relationship
between the dynamics of radical decay and the supramo-
lecular structure of the radicals. We report here a
systematic investigation of the molecular structure of
guest radicals and the exchangeable cations of zeolite
host structures on the steady-state CW EPR spectra and
the decay of carbon-centered radicals adsorbed on MFI
zeolites. A series of symmetrically substituted dibenzyl
ketones, 1-4 (Scheme 1), were selected as radical precur-
sors. After photoexcitation, the ketones undergo R-cleav-
age from the triplet state to produce an acyl-benzyl
radical pair in a primary process, followed by rapid
decarbonylation of the acyl radical to generate a second-
ary radical pair consisting of benzyl radicals 1R-4R
(Scheme 1).6-8
Resu lts
The ketones 1-4 (Scheme 1) were adsorbed on
NaZSM-5 (Si/Al ) 20).9,10 In addition, ketone 2 was
adsorbed on cation-exchanged zeolites MZSM-5 (M ) Li,
Na, K, Rb, and Cs). The complexes formed by the
adsorbed ketones are termed 1-4@MZSM-5. Photolysis
of 1-4@MZSM-5 produces radicals 1R-4R (Scheme 1),5
and the complexes of the radicals with zeolites are termed
1R-4R@MZSM-5.
Figure 1 displays the decay (left) and CW-EPR spectra
(right) of radicals 1R-4R@NaZSM-5. The most salient
† Dedicated to Waldemar Adam on the occasion of his 65th birthday.
(1) Schuh, H.-H.; Fischer, H. Helv. Chim. Acta 1978, 61, 2130-2164.
(2) Griller, D.; Ingold, K. U. Acc. Chem. Res. 1976, 9, 13-19.
(3) Turro, N. J . Acc. Chem. Res. 2000, 33, 637-646.
(7) Robbins, W. K.; Eastman, R. H. J . Am. Chem. Soc. 1970, 92,
6076-6077.
(4) Hirano, T.; Li, W.; Abrams, L.; Krusic, P. L.; Ottaviani, M. F.;
Turro, N. J . J . Am. Chem. Soc. 1999, 121, 7170-7171.
(8) Robbins, W. K.; Eastman, R. H. J . Am. Chem. Soc. 1970, 92,
6077-6079.
(5) Turro, N. J .; Lei, X.-G.; J ockusch, S.; Li, W.; Liu, Z.; Abrams, L.;
Ottaviani, M. F. J . Org. Chem. 2002, 67, 2606-2618.
(9) Meier, W. M.; Olson, D. H. Atlas of Zeolite Structure Types;
Butterworth-Heinemann: London, 1992.
(6) Engel, P. S. J . Am. Chem. Soc. 1970, 92, 6074-6076.
(10) Davis, M. E.; Lobo, R. F. Chem. Mater. 1992, 4, 756-768.
10.1021/jo020141e CCC: $22.00 © 2002 American Chemical Society
Published on Web 07/04/2002
J . Org. Chem. 2002, 67, 5779-5782
5779

Turro et al.
F IGURE 1. CW-EPR spectra (right) of radicals 1R-4R@NaZSM-5 produced by photolysis of 1-4@NaZSM-5 (λirr > 280 nm) and
corresponding decay kinetics of the EPR signal (left) after short irradiation (10 s, λirr > 280 nm).
F IGURE 2. CW-EPR spectra (right) of the radical 2R on cation-exchanged zeolites (2R@MZSM-5) produced by photolysis of
2@MZSM-5 (λirr > 280 nm) and corresponding decay kinetics of the EPR signal (left) after short irradiation (10 s, λirr > 280 nm).
structural feature associated with the EPR decay curves
is the clear correlation of the rate of the decay with the
size of the radical: the larger the radical, the slower the
decay. For example, the smallest radical 1R shows the
fastest decay (t1/2 ca. 2 min) and the largest radical 4R
shows the slowest decay (t1/2 > 200 min). The most salient
structural feature associated with the EPR spectra is the
clear correlation of the line width of the spectrum with
the size of the radical: the larger the radical, the broader
the signals. It is also noted that the signal/noise is poorest
for the benzyl radical spectrum, which decays the fastest.
Figure 2 displays the decay (left) and steady-state CW-
EPR spectra (right) of radical 2R@MZSM-5. The most
salient structural feature associated with the EPR decay
curves is the clear correlation of the rate of decay with
the size of the cation: the larger the cation, the slower
the decay. For M ) Li the radical shows the fastest decay
(t1/2 ca. 10 min), and for M ) Cs the radical shows the
slowest decay (t1/2 > 200 min). The line widths of the EPR
spectra are similar for M ) Li, Na, and K. However, a
definite increase in line width occurs with M ) Rb and a
significant increase in line width occurs with M ) Cs.
Discu ssion
The external surface of ZSM-5 crystals consists of pores
(circular diameter of pores ca. 5.5 Å) and a solid frame-
5780 J . Org. Chem., Vol. 67, No. 16, 2002

Dynamics of Radicals in MFI Zeolites
work between the pores. The internal void space consists
of channels (ca. 5.5 Å in diameter) that intersect to
produce roughly spherical supercages of ca. 9 Å in
diameter.9,10 The molecular cross section of the guest
ketone 1 is small enough to enter the pores on the
external surface of ZSM-5. Under low-loading conditions
1 is exclusively adsorbed on the internal surface. Con-
versely, the guest ketones 2-4 are adsorbed on the
external surface, because the molecular cross sections of
the ketones are too large to enter the pores of ZSM-5.
Evidence has been produced that the ketones 2-4
predominantly adsorb at the pore openings of the exter-
nal surface.5
the radicals.13 A decrease in rotational motion is consis-
tent with a concomitant decrease in diffusional motion.
The data in Figure 1 are consistent with a correlation
of the variation of the molecular structure of the guest
on the observed decay of the guest@zeolite system.
However, since we are dealing with a supramolecular
system, the decay is expected to depend on the variation
of the structure of the host zeolite also. The most subtle
variation of the host structure is the variation of the
exchangeable cations associated with the negatively
charged Al atoms of the zeolite framework. The cations
can be expected to interact with the radicals, both
sterically and electrostatically. Previous studies have
indicated that electrostatic interactions of the cations
with hydrocarbon radicals are not as important as steric
interactions.14 The data in Figure 2 speak for a dominat-
ing steric effect on the decay of the radicals through
radical-radical reactions in the zeolite supercages. The
larger the cation, the slower the decay. However, for
cations Li, Na, and K, the line widths of the EPR spectra
are similar and relatively sharp, indicating a similar and
relatively high degree of rotational motion (and, by
inference, diffusional motion) for these radicals. Only in
the case of M ) Rb and Cs is line broadening significant.
These results are consistent with the main effect of the
cation as being due to steric inhibition of radical-radical
reaction in the zeolite supercages. This conclusion in turn
suggests that the cations are associated with the super-
cages so that they are effective at the time of radical-
radical reaction. It is also possible that Rb and Cs cause
an increase in line broadening due to an increase in
radical relaxation resulting from increased spin-orbit
coupling.15
Photolysis of the complexes 1-4@ZSM-5 generated the
benzyl radical pairs 1R-4R@ZSM-5. The decay curves
shown in Figures 1 and 2 represent an example of “flash
photolysis” with a time resolution in the orders of
seconds. In this experimental configuration only radicals
with a lifetime longer than ca. 1 s can be observed. The
highly mobile benzyl radicals 2R-4R on the external
surface are expected to recombine rapidly and are,
therefore, “invisible” in our EPR measurements. It is
expected that only the radicals, which can escape into
the pores on the internal surface, possess a lifetime long
enough to be detected by EPR in our experimental
configuration.
The resulting lifetimes followed by EPR of radicals
1R-4R increase monotonically in the order 1R < 2R <
3R < 4R; i.e., as the group X attached to the radical
center increases from X ) H to X ) (CH2)4CH3, the rate
of decay of the radicals decreases. We attribute this
correlation to a supramolecular steric effect on radical-
radical reactions which determine the lifetime of the
radicals under observation. From computer-simulated
molecular models, only the radical-radical combination
of benzyl radicals, 1R, to form 1,2-diphenylethane is
sterically plausible in the supercages of the internal
surface of the MFI zeolite. Product studies showed that
the radicals 2R-4R predominantly undergo radical-
radical disproportionation on ZSM-5 zeolites.5 The con-
cept of a steric effect on the radical-radical reactions in
the internal supercages explains the correlation between
radical size and radical decay. This is a dramatic su-
pramolecular steric effect on the radical-radical reaction,
since in ordinary solvents all of the radicals investigated
undergo radical-radical combination at close to the
diffusion-controlled rate.11,12
Con clu sion s
CW-EPR analysis of supramolecular complexes 1R-
4R@MZSM-5, generated by photolysis of 1-4@MZSM-
5, provided information on the supramolecular dynamics
of the radicals. The lifetimes of the radicals increased as
the group X attached to the radical center increases from
X ) H to X ) (CH2)4CH3. This correlation was attributed
to a supramolecular steric effect on the radical-radical
reactions, which determine the radical lifetimes. In
addition, line broadening of the EPR signal (associated
with a decrease in rotational motion) is consistent with
the notion of decreased mobility as the group X attached
to the radical center increases. EPR experiments involv-
ing cation-exchanged zeolites (MZSM-5; M ) Li, Na, K,
Rb, Cs) showed a strong dependence of the radical
lifetime on the size of the cation; i.e., the larger the cation,
the slower the decay. This is consistent with steric
inhibition of radical-radical reaction in the supercages.
In addition to a steric effect on the reaction step of two
radicals, there may also be a steric effect on the diffusion
of the radicals into a supercage. In other words, in the
zeolite internal surface, the diffusion of the smallest
radical 1R may be much faster than the diffusion of the
largest radical 4R. A dramatic dependence of the diffu-
sion rate on subtle changes in size is well established for
molecular diffusion in zeolites. This possibility is consis-
tent with the observed increase in line broadening in the
EPR spectra as the size of the radical increases. The
increased line broadening can be associated with a
decrease in the correlation time for rotational motion of
Exp er im en ta l Section
1,3-Diphenyl-2-propanone (1) was obtained from Aldrich and
purified by recrystallization from 5% (v/v) ether in hexane. 2,4-
Diphenylpentan-3-one (2), 3,5-diphenylheptan-4-one (3), and
(13) Berliner, L. J ., Ed. Spin Labeling. Theory and Applications;
Academic Press: New York, 1976.
(14) Turro, N. J .; Zhang, Z. Tetrahedron Lett. 1987, 28, 5637-5640.
(15) Warrier, M.; Turro, N. J .; Ramamurthy, V. Tetrahedron Lett.
2000, 41, 7163-7167.
(11) Baretz, B. H.; Turro, N. J . J . Am. Chem. Soc. 1983, 105, 1309-
1316.
(12) Ghatlia, N. D.; Turro, N. J . J . Photochem. Photobiol., A: Chem.
1991, 57, 7-19.
J . Org. Chem, Vol. 67, No. 16, 2002 5781

Turro et al.
6,8-diphenyltridecan-7-one (4) were synthesized and charac-
terized previously.5
solvent by evaporation under a stream of argon, the ketone-
loaded zeolites were further degassed under vacuum (2 × 10-5
Torr) to remove traces of water, solvents, and oxygen. The
photolysis was performed inside the cavity of a CW-EPR
spectrometer (EMX, Bruker) employing a 1000 W Xe lamp
(LX1000UV; ILC Technology, Inc.) in conjunction with a long
wave pass filter (cutoff at 280 nm). For kinetic measurements
the zeolite sample was briefly irradiated (10 s) in the cavity
of the EPR spectrometer, and the decay of the EPR signal at
a fixed magnetic field was monitored over time.
NaZSM-5 (Si/Al ) 20), a Chemie Uetikon product, was
supplied as a generous gift by Dr. V. Ramamurthy, Depart-
ment of Chemistry, Tulane University, New Orleans, LA. The
mean crystal lateral dimension was determined by scanning
electron microscopy as ca. 0.3 µm. The external surface area
was determined by mercury porosimetry16 to be 16 m2/g. This
zeolite (5 g) was then ion-exchanged two times at 70 °C for 24
h with 50 mL of 0.5 M solutions of LiCl, KCl, RbCl, and CsCl,
washed thoroughly with deionized water, and dried.17,18 The
zeolites were activated in a furnace at 500 °C for at least 8 h
and cooled to room temperature in a desiccator before use.
The ketones 1-4 were adsorbed on ZSM-5 from an isooctane
solution at ca. 1% (w/w) loading. After removal of the isooctane
Ack n ow led gm en t. We thank the National Science
Foundation for its generous support of this research
(Grant CHE 01-10655). The research was also supported
in part by the Department of Energy and the National
Science Foundation by Grant CHE 98-10367 to the
Environmental Molecular Sciences Institute (EMSI) at
Columbia University.
(16) Abrams, L.; Keane, M.; Sonnichsen, G. C. J . Catal. 1989, 1989,
410-419.
(17) Lei, G.-D.; Kevan, L. J . Phys. Chem. 1990, 94, 6384-6390.
(18) Ramamurthy, V.; Corbin, D. R.; J ohnston, L. J . J . Am. Chem.
Soc. 1992, 114, 3870-3882.
J O020141E
5782 J . Org. Chem., Vol. 67, No. 16, 2002