Photochemical and thermal behavior of styrenes within acidic and nonacidic zeolites. Radical cation versus carbocation formation

Article abstract of DOI:10.1021/jp9708963

Laser flash photolysis of a series of substituted styrenes embedded within the cavities of the large pore zeolite NaY leads to the formation of the corresponding styrene radical cation. The reactivity and spectra of these radical cations embedded within NaY are examined and compared to the reactivity of the same radical cations in solution. It is found that for the highly reactive parent styrene radical cation the zeolite framework provides a strong stabilizing effect. For the 4-methoxy-substituted styrene radical cation the zeolite framework plays less of a role in stabilizing the radical cation as compared to the reactivity of the same radical cation in acetonitrile solution. Rigorous analysis of the thermal stability of 4-methoxystyrene, 4-methylstyrene, and anethole in the zeolite micropores was carried out using two sources of NaY zeolite (Aldrich and The PQ Corporation). It was found that the thermal stability was surprisingly dependent on the source of the NaY zeolite. 4-Methoxystyrene, 4-methylstyrene, and anethole were thermally stable in NaY (Aldrich) but rapidly dimerized in NaY (PQ) upon incorporation with dichloromethane. We observed the formation of the same type of dimers not only for 4-methoxystyrene but also for 4-methylstyrene and anethole. In addition, 4-methoxystyrene was incorporated into a series of different acid zeolites (HZSM-5, HMordenite, HBeta, and HY) varying in the shape and size of their micropores where rapid thermal protonation occurs. Dimerization of the thermally formed 4-methoxyphenethyl cation with a neutral molecule of 4-methoxystyrene took place within all the acid zeolites examined. The generation of this secondary 1,3-bis(4-methoxyphenyl)-1-butylium ion was clearly observed in the medium pore ZSM-5. This carbocation was found to be thermally unstable in the acidic environment provided by the four acidic zeolites and underwent a proton and hydride transfer to form the more stable allylic 1,3-bis(4-methoxyphenyl)buten-1-ylium cation. In the large round cavities of HY a competing cyclization reaction took place which led to the formation of the 3-methyl-5-methoxy-1,4-methoxyphenylindanyl cation.

Full text of DOI:10.1021/jp9708963

J. Phys. Chem. B 1997, 101, 6921-6928
6921
Photochemical and Thermal Behavior of Styrenes within Acidic and Nonacidic Zeolites.
Radical Cation Wersus Carbocation Formation
Frances L. Cozens,*,† Roumiana Bogdanova,† Miche`le Re´gimbald,‡ Hermenegildo Garc´ıa,§
Vinente Mart´ı,§ and J. C. Scaiano‡
Departments of Chemistry, Dalhousie UniVersity, Halifax, NoVa Scotia, Canada, B3H 4J3, UniVersity of
Ottawa, Ottawa, Ontario, Canada, K1N 6N5, and Instituto de Tecnolog´ıa Qu´ımica UPV-CSIC, UniVersidad
Polite´cnica de Valencia, Apartado de Correos 22012, Camino de Vera. s/n, 46071 Valencia, Spain
ReceiVed: March 11, 1997; In Final Form: June 3, 1997X
Laser flash photolysis of a series of substituted styrenes embedded within the cavities of the large pore zeolite
NaY leads to the formation of the corresponding styrene radical cation. The reactivity and spectra of these
radical cations embedded within NaY are examined and compared to the reactivity of the same radical cations
in solution. It is found that for the highly reactive parent styrene radical cation the zeolite framework provides
a strong stabilizing effect. For the 4-methoxy-substituted styrene radical cation the zeolite framework plays
less of a role in stabilizing the radical cation as compared to the reactivity of the same radical cation in
acetonitrile solution. Rigorous analysis of the thermal stability of 4-methoxystyrene, 4-methylstyrene, and
anethole in the zeolite micropores was carried out using two sources of NaY zeolite (Aldrich and The PQ
Corporation). It was found that the thermal stability was surprisingly dependent on the source of the NaY
zeolite. 4-Methoxystyrene, 4-methylstyrene, and anethole were thermally stable in NaY (Aldrich) but rapidly
dimerized in NaY (PQ) upon incorporation with dichloromethane. We observed the formation of the same
type of dimers not only for 4-methoxystyrene but also for 4-methylstyrene and anethole. In addition,
4-methoxystyrene was incorporated into a series of different acid zeolites (HZSM-5, HMordenite, HBeta,
and HY) varying in the shape and size of their micropores where rapid thermal protonation occurs. Dimerization
of the thermally formed 4-methoxyphenethyl cation with a neutral molecule of 4-methoxystyrene took place
within all the acid zeolites examined. The generation of this secondary 1,3-bis(4-methoxyphenyl)-1-butylium
ion was clearly observed in the medium pore ZSM-5. This carbocation was found to be thermally unstable
in the acidic environment provided by the four acidic zeolites and underwent a proton and hydride transfer
to form the more stable allylic 1,3-bis(4-methoxyphenyl)buten-1-ylium cation. In the large round cavities of
HY a competing cyclization reaction took place which led to the formation of the 3-methyl-5-methoxy-1,4-
methoxyphenylindanyl cation.
Introduction
studying these species in strong acidic media such as concen-
trated H2SO4 and magic acid solutions. Under these harsh
Our current knowledge of the chemistry of organic com-
pounds is largely based on understanding the reactive intermedi-
ates which are involved in the transformation of starting
materials to products. The reactive intermediate generated from
a given compound may vary with the operating mechanism and
the conditions under which the reaction is carried out. Recently,
a large amount of work has focused on the photochemistry of
styrene derivatives and it has been found that irradiation of
substituted styrenes leads to the formation of two different types
of cationic reactive intermediates depending on the conditions
employed. Using laser flash photolysis, researchers have been
able to study the reactivity of phenethyl carbocations generated
by photoprotonation of the corresponding styrene derivative
when the photolysis was carried out in nonnucleophilic fluori-
nated alcohols such as 2,2,2-trifluoroethanol and 1,1,1,3,3,3-
hexafluoro-2-propanol.1 On the other hand, the correspond-
ing styrene radical cations have been generated by direct
photoionization of the styrene derivative when the laser pho-
tolysis was carried out in water or acetonitrile.2 These studies
have led to a deeper insight on the reactivity of both carbocations
and radical cations under solvolytic conditions. A large amount
of information on carbocations has also been obtained by
conditions many carbocations are sufficiently stable for their
spectral properties to be obtained.3
Zeolites are microporous crystalline aluminosilicates made
up of [SiO4]4- and [AlO4]5- tetrahedra.4-6 The tetrahedra are
arranged in a three-dimensional array, which gives the zeolite
its open framework structure. The framework geometry of
zeolites consists of pores and channels whose size and shape
vary widely among different types of zeolites. Zeolites are often
classified by their pore size.7 For example, ZSM-5 has a pore
containing 10 oxygen atoms and is classified as a medium pore
zeolite whereas Beta has a pore containing 12 oxygen atoms
and is known as a large pore zeolite.8 For every aluminum
present in the framework there is a net negative charge which
must be compensated by a charge balancing cation. These
charge-balancing cations are often Na+ or other alkali earth
metals. Bro¨nsted acid sites can be easily introduced into the
framework by exchanging the Na+ cations with protons. These
acidic solids have been the subject of extensive research due to
their catalytic ability.8
Zeolites have proven to be good hosts for the study of cationic
intermediates presumably due to the stabilization provided by
the internal electrostatic fields within the voids of the mi-
croporous solids9 and to the confined environment, which
provides protection of the cationic center from nucleophilic
species.10 Owing to the need to generate thermally the
† Dalhousie University.
‡ University of Ottawa.
§ Universidad Polite´ncia de Valencia.
X Abstract published in AdVance ACS Abstracts, July 15, 1997.
S1089-5647(97)00896-1 CCC: $14.00 © 1997 American Chemical Society

6922 J. Phys. Chem. B, Vol. 101, No. 35, 1997
Cozens et al.
SCHEME 1
carbocation upon adsorption of a neutral precursor, studies
involving the stability of carbocations within zeolites have
focused on acidic zeolites.11 In some cases, carbocations have
been found to be sufficiently stable in acid zeolites to obtain
their spectral properties by conventional techniques.12-16 It has
recently been suggested that the acidity of zeolites compares
better to 70% aqueous solutions of H2SO4 than to that of
superacid media as some researchers have postulated.11 There-
fore, it is not only the extreme acidity of zeolites but other
special properties of the internal voids, in terms of polarity,
mobility restriction, presence of acid sites, that makes these
solids very appropriate to stabilize positively charged reaction
intermediates.17,18 When a positively charged species is incor-
porated inside the voids of these solids, the whole lattice of the
zeolite acts as a counteranion, insulating and protecting the
carbocation from attack by external nucleophiles.19 Stabilization
has been found to be dramatic in many cases in which the tight
fit between the organic guest and the zeolite occurs.20 These
appropriate characteristics combined with the wide range of
crystalline structure and the possibility to introduce active sites
have made zeolites the host of choice for many studies of
organic cations.
Organic radical cations formed within the zeolite framework
have been the subject of much interest.21-28 Electron acceptor
sites may be present within the zeolite framework, and therefore,
zeolites have been established as convenient host materials to
promote the formation of radical cations.18 Spontaneous radical
cation formation and/or precursor decomposition has been
reported upon the incorporation of several organic guests into
Na+-exchanged zeolites.29-32 However, it is now accepted that
acid sites in these solids are related to their electron acceptor
ability. Recent studies have supported that Bro¨nsted acid sites
of zeolites can also be responsible for the thermal radical cation
formation, thus implying that the H+-form of zeolites could
exhibit a dual behavior as acid and oxidizing solids.8,10,33,34 On
the other hand, a few examples of organic radical cations within
the zeolite framework generated by photolysis or pulse radiolysis
have also been reported; however, radical cations of styrenes
in zeolites have never been studied.29,33-39
Figure 1. Transient diffuse reflectance spectrum after 266 nm
excitation of an O2 purged (A) and vacuum-sealed (B) sample of
anethole incorporated with NaY (Aldrich). Spectra were recorded (b)
1.6 µs, (9) 4.8 µs, ([) 24 µs, and (2) 75 µs after the laser pulse. The
samples were prepared using CH2Cl2 as the carrier solvent.
anethole into the framework of the Y faujasite (average loading
level: 1 anethole/10 supercages). Once the incorporation
procedure was complete, the sample was thoroughly washed
with CH2Cl2 and dried under vacuum. Throughout the incor-
poration procedure the bulk sample remained white, showing
no obvious signs of styrene decomposition (vide infra differ-
ences in the NaY samples). The laser photolysis experiments
were performed on the same day as sample preparation. Laser
flash photolysis of the anethole-NaY complex under an
atmosphere of oxygen led to the transient reflectance spectrum
shown in Figure 1A. The spectrum shows two strong absorption
bands, one centered at 620 nm and the other at 400 nm, and is
almost identical to that observed for the anethole radical cation
in solution.2,42 On the basis of this similarity, the intermediate
generated upon 266 nm excitation of anethole encapsulated
within NaY can be confidently identified as the anethole radical
cation, R1 ) OCH3 and R2 ) CH3, formed by photoionization
of anethole, eq 1. The assignment of the two reflectance bands
It is, therefore, of interest to determine which is the prevalent
pathway when both processes, either protonation or single
electron abstraction, Scheme 1, can occur concurrently.40,41 As
previously mentioned, CdC double bonds, and in particular
simple styrenes, are suitable probe molecules to establish the
nature of the actual intermediate (e.g., a radical cation or a
carbocation) generated within zeolite media. In this study two
complementary strategies have been devised: either time-
resolved diffuse-reflectance techniques for the Na+-form of the
large pore Y zeolite where radical cation intermediates appear
as short-lived transients, and conventional diffuse reflectance
for acid zeolites, where thermally generated carbocations appear
as persistent, long-lived species.
to one species is in agreement with the observation that both
the 400 nm and the 620 nm bands decay with identical kinetics.
The decay was fit to a double-exponential expression where
the fast component was analyzed for the decay of the radical
cation embedded within the framework of NaY and was on the
order of 4 × 104 s-1
.
Results and Discussion
Photoionization of anethole must be accompanied by the
ejection of an electron into the zeolite framework. Electrons
within the pores of faujasites typically complex with Na+
Radical Cation Generation. The free diffusion of 20 mg
of anethole in dichloromethane, CH2Cl2, with 600 mg of
dehydrated NaY (Aldrich), led to the incorporation of 5 mg of
3+
counterions to generate Na4 ions that have absorption in the

Intrazeolite Photochemistry
J. Phys. Chem. B, Vol. 101, No. 35, 1997 6923
Figure 2. Transient diffuse reflectance spectrum after 266 nm
excitation of an O2 purged sample of 4-methoxystyrene incorporated
with NaY (Aldrich). Spectra were recorded (b) 0.4 µs, (9) 1.3 µs, ([)
4.8 µs, and (2) 14 µs after the laser pulse. The sample was prepared
using CH2Cl2 as the carrier solvent.
500 nm region, eq 2.43,44 We investigated the possibility that
3+
e- + 4Na+ f Na4
(2)
we could observe such a sodium complex by carrying out the
photolysis of the anethole-NaY complex in the absence of
oxygen. Figure 1B shows the transient diffuse reflectance
spectrum generated upon 266 nm excitation of anethole in NaY
under reduced pressure (0.005 Torr). Additional absorption in
the 500 nm region is clearly seen in the spectrum and has a
maximum at ∼520 nm. The decay of this transient was found
to be 1.5 × 106 s-1, which is much faster than the decay of the
anethole radical cation. The transient reacted very rapidly with
oxygen as seen by its complete quenching upon the addition of
oxygen to the sample. We therefore assign this transient species
Figure 3. Transient diffuse reflectance spectrum after 266 nm
excitation of 4-methylstyrene (A) and styrene (B) in O2 purged NaY
(Aldrich). Spectra were recorded (b) 0.28 µs, (9) 0.72 µs, ([) 1.4 µs,
and (2) 6.0 µs (4-methylstyrene) and (b) 0.28 µs, (9) 0.56 µs, ([)
0.80 µs, and (2) 6.0 µs (styrene) after the laser pulse. The samples
were prepared using CH2Cl2 as the carrier solvent.
bands followed the same kinetics and was fit to a double
exponential expression where the fast component was analyzed
for the decay of the radical cation within NaY and was on the
order of 1.6 × 106 s-1. Figure 3 shows the transient diffuse
reflectance spectra of the radical cations generated upon
photoionization of 4-methylstyrene, R1 ) CH3 and R2 ) H,
and parent styrene, R1 ) H and R2 ) H, eq 1. In both cases,
the λmax of the transients at ca. 350-360 nm and ca. 600 nm
are in accordance with the spectra of the appropriate radical
cation,2 indicating that the NaY framework clearly acts as a
highly polar, solid matrix for the generation and detection of
these reactive cationic species. Additional absorption near 500
nm is also observed in these spectra. Absorption near this region
has been observed upon the generation of styrene radical cations
in solution, and attributed to the presence of “dimer” radical
cations produced by addition of the radical cation to the neutral
styrene.2,46 Such a process may take place rapidly in zeolites,
especially under conditions where a radical cation is generated
in a doubly occupied cavity.
The traces for the decay at 600 nm of the photogenerated
4-methoxy, 4-methyl, and unsubstituted styrene radical cations
are shown in Figure 4. These traces do not follow simple first-
order behavior, but each can be adequately fit to an expression
for two consecutive first-order decays. While the long-lived
component for the 4-methoxystyrene radical cation is more
pronounced than that for the 4-methyl and the unsubstituted
derivatives, the rate constants calculated from the decays of each
radical cation, summarized in Table 1, are remarkably simi-
lar.19,47 In particular, a factor of about 4 separates the rate
constant for the highly unstable parent radical cation from that
of the stabilized 4-methoxystyrene radical cation when the
radical cations are generated in the confined environment of
the zeolite. This contrasts sharply with results in homogeneous
solutions, such as AcN, where the difference in reactivity is
greater than 106.2 It is also interesting that the observed decrease
in the range of reactivity in the zeolite can be mostly attributed
3+
to the Na4 complex generated when the electron is ejected
from anethole into the zeolite framework. The fact that we
simultaneously observe both the radical cation of anethole and
the trapped electron provides compelling evidence that photo-
ionization is occurring in the large pore Na+-form Y zeolite
upon 266 nm laser excitation.
A sharper absorption band in the 380-400 nm region is
observed upon going from a vacuum to an O2 environment.
Such narrowing has been observed before for diffuse reflectance
studies in zeolites.13 In the present work, the small change in
the maximum may be due to the presence of the radical anion
of the styrene produced by trapping of the photoejected electron
by neutral 4-methoxystyrene. A similar process has been
observed in the formation of the pyrene radical anion upon
photoionization of pyrene in zeolites.45
These results established that styrene type radical cations can
be readily generated and observed in the zeolite framework using
time-resolved diffuse reflectance and encouraged us to attempt
to generate radical cations of other styrene derivatives in NaY.
In particular, this would allow us to address questions concern-
ing how the zeolite framework influences the lifetime of the
photogenerated radical cations with decreasing thermodynamic
stability. Thus we examined the photolysis of 4-methoxysty-
rene, 4-methylstyrene, and styrene incorporated in NaY. In all
cases, incorporation procedures involved using CH2Cl2 as the
carrier solvent, and the samples remained colorless throughout
sample preparation. The photolysis experiments were carried
out within 24 h of sample preparation.
Laser photolysis of 4-methoxystyrene incorporated within
NaY under an atmosphere of oxygen led to the transient diffuse
reflectance spectrum shown in Figure 2. The spectrum shows
strong bands centered at 600 and 360 nm that are consistent
with assignment of the transient to the 4-methoxystyrene radical
cation,2 R1 ) OCH3 and R2 ) H, eq 1. The decay of the two

6924 J. Phys. Chem. B, Vol. 101, No. 35, 1997
Cozens et al.
and once prepared, the sample was left under vacuum for a
period of 1 day. After this time the 4-methoxystyrene was
reextracted from the zeolite using CH2Cl2 as the solvent.
According to GC analysis the only compound retrieved from
the zeolite was 4-methoxystyrene in greater than 95% purity.
To test further for the stability of 4-methoxystyrene within the
Na+-form of Y zeolite, the sample was allowed to remain under
vacuum for a period of 1 week and then was reextracted from
the zeolite. GC analysis showed that 4-methoxystyrene was
still the major compound (85%); however, a new peak which
had a longer retention time (12.5 min) than that of 4-methoxy-
styrene (4.3 min) and accounted for 10% of the organic matter
was clearly observed.
Figure 4. Transient decay kinetics generated upon 266 nm excitation
for 4-methoxystyrene (9), 4-methylstyrene (O), and styrene (b) within
NaY (Aldrich). The normalized traces were monitored for the 600 nm
band under an atmosphere of oxygen.
The same sample preparation procedure with 4-methoxy-
styrene was followed using NaY Si:Al ) 2.6 purchased from
the PQ corporation. In this case, very different results were
obtained. When 4-methoxystyrene dissolved in either CH2Cl2
or hexane was mixed with the activated NaY (PQ) zeolite, the
solid turned light purple immediately as the organic mixture
came into contact with the zeolite. Diffuse reflectance spec-
troscopy showed the presence of a weak band at 585 nm
accompanied with an intense absorption at 280 nm owing to
unaltered starting 4-methoxystyrene. The characteristic 585 nm
absorption is consistent with the results attained using acid
zeolites (vide infra). The purple color was persistent and
remained during the sample preparation procedure. Once the
incorporation procedure was finished, no 4-methoxystyrene
could be recovered from the zeolite upon extracting out the
organic material using CH2Cl2 as the solvent. Two products
(ratio 10:1) were extracted according to GC analysis, and the
retention time of the major peak (12.5 min) was identical to
that of the minor product extracted from the Aldrich sample
after 1 week. The new compounds formed upon incorporation
of 4-methoxystyrene into NaY (PQ) were examined and
TABLE 1: First-Order Rate Constants for the Decay at 600
nm of Substituted Styrene Radical Cations in NaY and in
Acetonitrile (AcN)
NaY (kobs/s-1
)
substituted styrene
fast
slowa
AcNb (kobs/s-1
)
styrene
4.5 × 106
2.6 × 106
1.4 × 106
0.5 × 106
0.4 × 106
0.1 × 106
c,d
4-methylstyrene
4-methoxystyrene
anethole
3 × 108
1.3 × 105
4 × 104
4.5 × 104
a Fast and slow refer to the fast component and the slow component,
respectively, obtained by fitting the decay traces to a double first-order
exponential expression. b Taken from reference 2. c Too fast to be
directly measured. d The estimated value by extrapolation of the 4-Me
and 4-MeO data is 5 × 1010 s-1
.
to the effect of the zeolite environment on the reactivity of the
less stable radical cation.19 Thus, while the decay of anethole
and 4-methoxystyrene radical cations in NaY is similar to their
decay in AcN, the more unstable parent styrene radical cation
is about 4 orders of magnitude less reactive in the zeolite
compared to AcN. These results suggest that the kinetic
stabilizing effect of the zeolite framework is far more effective
for the more unstable cationic species and that for slightly
stabilized radical cations the zeolite framework is no more
effective than solvents such as acetonitrile. The stabilizing effect
of the zeolitic media on short-lived photogenerated radical ion
pairs has been already noted,48 but the modulation in this effect
depending on reactivity of the species is remarkable. In
addition, since the stabilized 4-methoxystyrene radical cation
and the parent styrene radical cation have very similar absolute
reactivities in the zeolite, the thermodynamic stability of the
radical cation does not seem to be a major factor in determining
the relative reactivity of the radical cations. The large kinetic
stabilizing effect of the zeolite framework to the less stable
radical cations may be due to the limiting amount of reactive
species found within the cavity. Thus the rate determining step
may be due to the availability of a nucleophile (or some other
quenching agent) and not due to the inherent stability of the
radical cation.
Influence of the NaY Source. Early in our study we found
that sample preparation was critical and that we obtained
different results depending on the source of NaY. The zeolite-
styrene samples prepared using NaY Si:Al ) 2.4 (Aldrich)
remained white throughout the sample preparation procedure.
To test that the styrenes were not reacting upon incorporation
into the zeolite micropores, we carried out rigorous product
analysis using 4-methoxystyrene. The complete incorporation
of 40 mg of 4-methoxystyrene into the NaY faujasite (800 mg)
was accomplished using hexane as the carrier solvent. The
sample was thoroughly dried under vacuum in the usual way,
1
identified using GC, GC/MS, and H NMR as trans and cis
isomers of 1,3-dianisyl-1-butene formed by the acid-catalyzed
dimerization of 4-methoxystyrene within the zeolite voids, eq
3.
These results therefore indicate that dimerization of 4-meth-
oxystyrene does occur on both Na+-form of the Y faujasites
but with enormously different rates. We also confirmed by GC/
1
MS and H NMR that 4-methylstyrene and anethole undergo
rapid dimerization upon incorporation in the NaY sample from
PQ.
The observed dimers upon incorporation of 4-methoxystyrene
into the Y faujasite are consistent with a mechanism involving
carbocation intermediates49,50 since the intermediacy of radical
cations would lead to a different dimer distribution.30 In fact,
the dimerization reaction of the 4-methoxystyrene radical cation
with its neutral precursor has been shown to yield high
concentrations of the dihydronaphthalene, 1,2-dihydro-7-meth-
oxy-1-(4′-methoxyphenyl)naphthalene and the cyclobutane,
trans-1,2-bis(4-methoxyphenyl)cyclobutane.51 These products
are clearly not observed in the product mixture extracted from

Intrazeolite Photochemistry
J. Phys. Chem. B, Vol. 101, No. 35, 1997 6925
the zeolites, which indicates that the major path of decomposi-
tion does not involve radical cation intermediates.
We checked the acidity of the NaY sample from PQ by IR
spectroscopy following the standard pyridine adsorption-
desorption method.52 In this technique, pyridine vapors (5 Pa)
are adsorbed onto dehydrated zeolites and the presence of
pyridinium ions (Bro¨nsted sites) or pyridine adducts (Lewis
sites) are determined by their characteristic aromatic vibration
bands at 1550 and 1450 cm-1, respectively. Using this
technique, no Bro¨nsted acid sites could be observed and only
very weak Lewis sites were titrated after evacuation of the
pyridine at 150 °C for 1 h. This corresponds to the minimum
outgassing temperature that can be used in the pyridine method
in order to avoid the interference of physisorbed pyridine. In
addition, the sensitivity of the pyridine adsorption-desorption
method is estimated to 1 mmol of pyridine, while differences
in the population of active sites can be smaller than this. The
fact that weak Lewis acid sites are present in NaY (PQ sample)
was confirmed by measuring the amount of ammonia (1.39
mmol g-1) adsorbed onto a dehydrated sample at 100 °C. In
addition, the crystallinity of the PQ sample determined by XRD
was found to be 100%. In spite of the fact that we have not
been able to measure differences in the acidity of the samples
by state-of-the-art techniques in zeolite chemistry, it is clear
that minor differences in the Si:Al ratio or in the crystallinity
lead to large differences in their catalytic behavior for the
dimerization of 4-methoxystyrene. In this regard, it has been
reported that the dimerization of the parent unsubstituted styrene
by zeolite does not require the presence of strong acid sites.50
Even the ability of neutral silanol groups to catalyze olefin poly-
merization has been recently advanced.53-55
Figure 5. Ground state diffuse reflectance spectra of 4-methoxystyrene
incorporated in HZSM-5 recorded immediately after sample preparation
(A) and after being stored 72 h at open ambient atmosphere (B).
Figure 5B was recorded. Quantitative electron pair resonance
(EPR) spectroscopy established that the amount of radical ions
at this stage was less than 1015 spin g-1, which would correspond
to less than 2 × 10-4 mg of radical cation/g of zeolite.59,60 These
quantities are too low to account for the strong absorption bands.
The peak at 360 nm is still clearly observable, but there is a
much stronger absorption centered at 585 nm. This band is
analogous to the absorption band centered at 500 nm for the
unsubstituted 1,3-diphenylpropen-1-ylium cation in HZSM-561
assuming a +10 nm shift for the methyl substituent on the allyl
bridge and +25 nm shift for each 4-methoxy group.62 Therefore,
the species responsible for the 585 nm band is thought to be
due to the analogous 1,3-bis(4-methoxyphenyl)-2-buten-1-ylium
cation (spectrum B). The first step in the generation of this
allylic cation is also the dimerization reaction of the 4-meth-
oxyphenethyl cation with 4-methoxystyrene to give the linear
cation A.63 Once this cation is formed within the zeolite it can
undergo a hydride transfer from the linear dimer, resulting in
equimolar formation of the allylic cation B and 1,3-dianisyl-
butane. The fact that both the linear dimer and its dihydro
derivative are observed as reaction products upon adsorption
of 4-methoxystyrene into acid zeolites (see Table 2) as well as
detection of cation A lends support to our proposed reaction
scheme,32 eq 4. Notably, the same absorption band, but
Carbenium Ion Generation. We had anticipated that the
incorporation of 4-methoxystyrene into the acidic environment
of H+ exchanged zeolites would lead to the rapid decomposition
of the styrene by thermal protonation56,57 of the â-carbon to
generate the 4-methoxyphenethyl carbocation. The possibility
arose that if the sample was kept free of atmospheric water we
would be able to observe the thermally stable carbocation within
the strongly acidic and polar environment of the acid zeolite.
The 4-methoxyphenethyl cation has been generated by photo-
protonation of 4-methoxystyrene under solvolytic conditions and
has an absorption maximum at 340 nm.1 This cation is known
to undergo a rapid dimerization reaction with the â-position of
the alkene bond of 4-methoxystyrene to yield the 1,3-dianisyl-
but-1-ylium cation, characterized by a new absorption band
shifted ∼20 nm to longer wavelengths compared to the
monomer 4-methoxyphenethyl cation.58
Upon the incorporation of 4-methoxystyrene in H+-exchanged
medium pore ZSM-5, the zeolite-styrene complex turned yellow
immediately following the incorporation procedure. The ground
state diffuse reflectance spectrum for the HZSM-5 4-methoxy-
styrene complex is shown in Figure 5A. This spectrum clearly
shows a strong absorption band centered at 360 nm. This band
is shifted exactly 20 nm to longer wavelengths with respect to
the 4-methoxyphenethyl cation, λmax ) 340 and corresponds
nicely to that of the 1,3-dianisylbut-1-ylium cation (spectrum
A) generated by thermal protonation of 4-methoxystyrene in
the acidic environment to yield the 4-methoxyphenethyl cation
followed by its reaction with neutral 4-methoxystyrene, eq 3.
The initially formed 4-methoxyphenethyl cation is too reactive
to be observed under the conditions that the diffuse reflectance
spectra are obtained.
After incorporation of 4-methoxystyrene into HZSM-5, the
spectrum shown in Figure 5A progressively changed and after
72 h a very different diffuse reflectance spectrum shown in

6926 J. Phys. Chem. B, Vol. 101, No. 35, 1997
Cozens et al.
TABLE 2: Products Formed upon Incorporation of 4-Methoxystyrene into HZSM-5, HMordenite, and HY Using Isooctane as
the Carrier Solvent
product distributiona
An
An
CH3O
An
An
An
zeolite
% mass balance
HZSM-5
HMor
HY
44
60
58
50
43
30
15
7
47
8
a Variable amounts of unidentified polymeric material was also extracted in the case of HMordenite and HY zeolites.
Examination of the diffuse reflectance spectrum upon incor-
poration of 4-methoxystyrene into the large cavity HY zeolite
also reveals the formation of both the 1,3-dianisylbut-1-ylium
cation A and the 1,3-bis(4-methoxyphenyl)buten-1-ylium cation
B, Figure 7A. However, we found that in this case, the diffuse
reflectance spectrum was very sensitive to the adsorption
procedure. Thus, by refluxing 4-methoxystyrene in CH2Cl2 the
spectrum shown in Figure 7B is recorded. This band is
consistent with the cyclized 3-methyl-1,4-methoxyphenylindanyl
cation C formed after the initial dimerization step.63,64 The
initially formed carbocation has sufficient room in the tridirec-
tional, large pore zeolite to undergo the cyclization reaction,
which yields an unstable cyclohexadienyl cation. The cyclo-
hexadienyl cation is not observed in the zeolite and is thought
to undergo a rapid deprotonation to yield the neutral indane
which then undergoes a hydride transfer to yield the 3-methyl-
5-methoxy-1,4-methoxyphenylindanyl cation, eq 6. This reac-
Figure 6. Ground state diffuse reflectance spectra of 4-methoxystyrene
incorporated in HMordenite (A) and HBeta (B) recorded immediately
after sample preparation.
appearing at 565 nm, is observed upon dissolving 4-methoxy-
styrene in concentrated sulfuric acid.
tion mechanism is also supported by the observation of the
indanic dimer in the reaction mixture after adsorption of
4-methoxystyrene in acid zeolites.
An analogous dimerization/cyclization reaction has been
previously reported in the literature upon the incorporation of
styrene into the large pore HY zeolite. The resulting 3-methyl-
1-phenylindanyl cation was found to be sufficiently stable to
be characterized by 13C NMR.63
In order to unambiguously identify the 585 nm band as the
1,3-bis(4-methoxyphenyl)buten-1-ylium cation B, an alternative
preparation of this cation was carried out in solution by
dissolving the thermal labile hydroxy compound into an acidic
TFE:AcOH solution. The addition of this allylic alcohol into
the acidic solvent led to the formation of the allylic cation by
loss of water, eq 5. The cation was sufficiently stable at room
The major products obtained upon the incorporation of
4-methoxystyrene into HZSM-5, HMordenite, and HY are given
in Table 2. For the zeolites examined, products derived from
the dimerization reaction were found to include linear and
cyclized products, together with 1,3-dianisylbutane. The major
products observed upon incorporation of 4-methoxystyrene into
the acidic zeolites agree with the carbocations observed by
ground state diffuse reflectance for the zeolite complexes. In
each case products derived from the initial attack of the
4-methoxyphenethyl cation to neutral 4-methoxystyrene are
observed.63 On the other hand the relatively poor mass balances,
although not uncommon for product studies using microporous
solids, reflects the presence on the solid of a large amount of
unextractable organic material.
temperature in the acidic environment for its absorption
spectrum to be obtained. The absorption spectrum of the
resulting cation B showed a strong band centered at 565 nm,
which is in agreement with our identification of the 585 nm
species within the zeolite micropores as the same cation.
When 4-methoxystyrene was incorporated into HMordenite
or HBeta (Figure 6) by the same procedure as described for
HZSM-5, the ground state reflectance spectra obtained are very
similar to that shown in Figure 5B. In both cases, a small band
centered at 360 nm and a much stronger band at 570 nm are
observed. Thus, in these acidic zeolites both the 1,3-dianisyl-
1-butylium A and the 1,3-bis(4-methoxyphenyl)-2-buten-1-ylium
B cations are formed within zeolite environment.
Conclusion
Radical cations of styrene derivatives are readily formed upon
photoionization within the large pore NaY faujasite (Aldrich).

Intrazeolite Photochemistry
J. Phys. Chem. B, Vol. 101, No. 35, 1997 6927
Time-resolved diffuse reflectance experiments were carried
out using a nanosecond laser system similar to those previously
described.65 The excitation source was the fourth harmonic (266
nm, < 20 mJ, < 8 ns/pulse) from a Continuum NY61-10 YAG
laser. The samples were contained in quartz cells constructed
with 3 × 7 mm2 tubing and were either evacuated under reduced
pressure and sealed under vacuum or purged with oxygen for a
minimum of 30 min prior to photolysis. The data analysis was
based on the fraction of reflected light absorbed by the transient
(reflectance change ∆J/Jo) where Jo is the reflectance intensity
before excitation and ∆J is the change in the reflectance after
excitation.
The guest-host samples were prepared with the utmost of
care. The non-acidic zeolite was activated at 475 °C for at least
10 h before use. Typically 600 mg of zeolite was activated
overnight and allowed to cool to room temperature in a dry
environment. A solution of the organic styrene (about 20 mg)
dissolved in the carrier solvent (20 mL of CH2Cl2 or hexane)
was then rapidly added to the activated zeolite. The sample
was magnetically stirred under an atmosphere of dry nitrogen
for 1 h, after which time the sample was centrifuged and the
solvent was decanted. Fresh solvent was added in order to wash
the zeolite surface. The sample was stirred for another 30 min,
centrifuged, and the solvent was decanted. The styrene-zeolite
complex was then transferred to a desiccator and dried under
vacuum of 0.005 torr for at least 5 h. Laser experiments were
conducted on fresh samples generally the day of sample
preparation.
Figure 7. Ground state diffuse reflectance spectra recorded im-
mediately after incorporation from a CH2Cl2 solution of 4-methoxy-
styrene in HY stirring at room temperature (A) or at reflux temperature
(B).
The reactivities of the 4-methoxy derivatives were found to be
similar to those for the same radical cations in acetonitrile, but
less stabilized 4-methyl and parent radical cations are remarkably
long lived in the zeolite host where the rate constants of the
4-methoxy and unsubstituted derivatives vary by a factor of only
4. In contrast to the relatively high thermal stability of styrenes
adsorbed on NaY (Aldrich), 4-methoxystyrene, 4-methylstyrene,
and anethole were found to undergo a rapid dimerization
reaction upon incorporation into NaY (PQ) zeolite.
4-Methoxystyrene was also incorporated into a series of acid
zeolites where rapid thermal protonation in the â-position
occurred leading to the formation of three different carbocations.
In the medium pore HZSM-5 the 1,3-dianisyl-1-butyl cation is
clearly observable. Over time, this cation decays with the
concomitant appearance of 1,3-bis(4-methoxyphenyl)but-1-
ylium cation (λmax ) 585 nm), which is also observed after the
adsorption of 4-methoxystyrene on HMordenite and HBeta. In
HY a third carbocationic species is observed and is thought to
be due to the 3-methyl-5-methoxy-1,4-methoxyphenylindanyl
cation formed by a cyclization reaction. These different
carbocations clearly demonstrate the ability of acid zeolites to
control the products of a reaction by their shape-selectivity
nature.
The amount of organic material incorporated into the zeolite
was calculated from the difference between the initial amount
of the styrene and the amount that was recovered in the
combined solutions measured by UV absorption. When hexane
was used as the carrier solvent, it was found that >90% of the
styrene was incorporated into the zeolite. The loading level
was found to be much lower when CH2Cl2 was used as the
carrier solvent, and in this case, about 25% of the styrene was
incorporated.
Determination of the stability of the styrene upon incorpora-
tion was carried out by reextracting the incorporated material
using CH2Cl2 as the solvent. The organic components were
either extracted from the zeolite by continually washing the
sample with fresh CH2Cl2 until no more organic compound was
visible with UV absorption or the sample was extracted using
a continuous solid-liquid extracting apparatus. The organic
solution was analyzed by GC-MS (Varian Saturn 11, 35 m
capillary column cross-linked 5% phenylmethylsilicone) and,
after removal of solvent, by 1H NMR (300 MHz, Varian
Gemini).
The adsorption procedure for the acid zeolites was carried
out by mixing a solution of 4-methoxystyrene (25 mg) in
isooctane at 90 °C in the presence of the thermally dehydrated
acid zeolite (1.0 g) for 1 h. After this time the solid was filtered,
washed with fresh isooctane (10 ml), and submitted to solid
liquid extraction using CH2Cl2 as solvent. The organic solution
was analyzed by GC/MS (Varian Saturn 11, 35 m capillary
Experimental Section
The two types of nonacidic NaY faujasite were obtained from
Aldrich (LZY52 molecular sieves, Si:Al ) 2.4) and from PQ
(Si:Al ) 2.6) and were used as received. Both zeolites have
previously been well characterized. The styrene derivatives
were obtained from Aldrich and distilled prior to use. All
solvents (hexane, dichloromethane, isooctane, and acetonitrile)
were of the highest grade (OmniSolv, BDH) and used without
further purification.
Ground state diffuse reflectance measurements of the dried
samples were obtained using a Shimadzu UV-2101 PC scanning
spectrophotometer. Kubelka-Munk equations F(R) were plot-
ted as a function of wavelength. GC analysis was done using
a Perkin Elmer autosystem GC with a DB5 15 meter column.
1
column crosslinked 5% phenylmethylsilicone) and H NMR.
Acknowledgment. Financial support by the Natural Sciences
and Engineering Research Council of Canada (NSERC) through
an operating grant (F.L.C. and J.C.S.) and Spanish DGICYT
(H.G., Project PB93-0380) are gratefully acknowledged. F.L.C.
is the recipient of an NSERC WFA. J.C.S. is the recipient of a
Killam Fellowship awarded by the Canada Council.

6928 J. Phys. Chem. B, Vol. 101, No. 35, 1997
Cozens et al.
(34) Werst, D. W.; Piocos, E. A.; Tartakovsky, E. E.; Trifunac, A. D.
Chem. Phys. Lett. 1994, 229, 421-428.
Supporting Information Available: Synthetic details and
1H NMR spectra of 1,3-bis(4-methoxyphenyl)-2-buten-1-one and
1,3-dianisylbutane and 1H NMR spectra of the products isolated
from the zeolites (2 pages). Ordering information is given on
any current masthead page.
(35) Barnabas, M. V.; Werst, D. W.; Trifunac, A. D. Chem. Phys. Lett.
1993, 206, 21-24.
(36) Werst, D. W.; Tartakovsky, E. E.; Piocos, E. A.; Trifunac, A. D.
J. Phys. Chem. 1994, 98, 10249-10257.
(37) Corma, A.; Forne´s, V.; Garc´ıa, H.; Mart´ı, V.; Miranda, M. A. Chem.
Mater. 1995, 7, 2136-2143.
References and Notes
(38) Liu, X.; Iu, K.-K.; Thomas, J. K. Chem. Phys. Lett. 1993, 204,
163-167.
(1) McClelland, R. A.; Chan, C.; Cozens, F. L.; Modro, A.; Steenken,
S. Angew. Chem., Int. Ed. Engl. 1991, 30, 1337-1339.
(2) Johnston, L. J.; Schepp, N. P. J. Am. Chem. Soc. 1993, 115, 6564-
6571.
(3) Olah, G. A.; Prakash, G. K.; Sommer, J. Superacids; John Wiley
& Sons: New York, 1985.
(39) Gessner, F.; Scaiano, J. C. J. Photochem. Photobiol. A 1992, 67,
91-100.
(40) Liu, X.; Iu, K.-K.; Thomas, J. K.; He, H.; Klinowski, J. J. Am.
Chem. Soc. 1994, 116, 11811-11818.
(41) Pitchumani, K.; Lakshminarasimhan, P. H.; Turner, G.; Bakker,
M. G.; Ramamurthy, V. Tetrahedron Letters 1997, 38, 371-374.
(42) Shida, T. Electronic Absorption Spectra of Radical Ions Physical
Science Data 34; Elsevier: New York, 1988.
(4) Dyer, A. An Introduction to Zeolite Molecular SieVes; John Wiley
& Sons: Bath, U.K. 1988.
(5) van Bekkum, H.; Flanigen, E. M.; Jansen, J. C., Eds. Introduction
to Zeolite Science and Practice; Elsevier: Amsterdam, 1991.
(6) Barrer, R. M. Zeolites and Clay Minerals as Sorbents and Molecular
SieVes; Academic Press: London, 1978.
(7) Davis, M. E. Acc. Chem. Res. 1993, 26, 111-115.
(8) Corma, A. Chem. ReV. 1995, 95, 559-614.
(9) Spackman, M. A.; Weber, H. P. J. Phys. Chem. 1988, 92, 794-
796.
(43) Liu, X.; Iu, K.-K.; Thomas, J. K. J. Phys. Chem. 1994, 98, 13720-
13728.
(44) Iu, K. K.; Liu, X.; Thomas, J. K. J. Phys. Chem. 1993, 97, 8165-
8170.
(45) Hashimoto, S. Chem. Phys. Lett. 1996, 236-242.
(46) Brede, O.; David, F.; Steenken, S. J. Chem. Soc., Perkin Trans. 2
1995, 23-32.
(47) The absorption intensities of the 4-methyl and the unsubstituted
styrene radical cations at 600 nm are considerably lower than the intensity
of the 4-methoxystyrene radical cation. This may suggest that a substantial
fraction of the less stable radical cations decay over a time scale too fast to
be resolved with our nanosecond equipment. However, the weaker intensity
may also be due to less efficient photoionization of the 4-methyl and styrene
radicals, as would be expected from their higher ionization potentials.
(48) Sankararaman, S.; Yoon, K. B.; Yabe, T.; Kochi, J. K. J. Am. Chem.
Soc. 1991, 113, 1419-1421.
(10) Cano, M. L.; Corma, A.; Forne´s, V.; Garc´ıa, H. J. Phys. Chem.
1995, 99, 4241-4246.
(11) Haw, J. F.; Nicholas, J. B.; Xu, T.; Beck, L. W.; Ferguson, D. B.
Acc. Chem. Res. 1995, 29, 259-267.
(12) Cozens, F. L.; Garcia, H.; Scaiano, J. C. Langmuir 1994, 10, 2246-
2249.
(13) Cozens, F. L.; Garcia, H.; Scaiano, J. C. J. Am. Chem. Soc. 1993,
115, 11134-11140.
(14) Cano, M. L.; Cozens, F. L.; Vicente, F.; Garcia, H.; Scaiano, J. C.
J. Phys. Chem. 1996, 18145-18151.
(49) Corriu, R.; Guenzet, J.; Reye, C. C.R. Seances Acad. Sci., Ser. C
1970, 270, 566-568.
(15) Cano, M. L.; Cozens, F. L.; Garcia, H.; Vicente, M.; Scaiano, J.
C. J. Phys. Chem. 1996, 18152-18157.
(50) Benito, A.; Corma, A.; Garc´ıa, H.; Primo, J. Appl. Catal., A 1994,
116, 127-135.
(16) Alvaro, M.; Corma, A.; Garc´ıa, H.; Miranda, M. A.; Primo, J. J.
Chem. Soc., Chem. Commun. 1993, 1041-1042.
(17) White, J. C.; Hess, A. C. J. Phys. Chem. 1993, 97, 9703-9706.
(18) Yoon, K. B. Chem. ReV. 1993, 93, 321-339.
(19) Cozens, F. L.; O’Neill, M.; Schepp, N. P. J. Am. Chem. Soc. 1997,
119, In press.
(51) Schepp, N. P.; Johnston, L. J. J. Am. Chem. Soc. 1994, 116, 6895-
6903.
(52) Arribas, J.; Corma, A.; Forne´s, V. J. Catalysis 1984, 88, 374-
380.
(53) Dessau, M.; Schmitt, K. D.; Kerr, G. T.; Woolery, G. L.; Alemany,
L. B. J. Catalysis 1987, 104, 484-489.
(20) Cano, M. L.; Forne´s, V.; Garc´ıa, H.; Miranda, M. A.; Pe´rez-Prieto,
J. J. Chem. Soc., Chem. Commun. 1995, 2477-2478.
(21) Rhodes, C. J. In Annual Reports on the Progress of Physical
Chemistry; Royal Society of Chemistry: London, 1993; Vol. 90; pp 82-
83.
(22) Rhodes, C. J. Colloids Surf., A 1993, 72, 111-118.
(23) Rhodes, C. J.; Reid, I. D.; Roduner, E. J. Chem. Soc., Chem.
Commun. 1993, 512-513.
(24) Rhodes, C. J. J. Chem. Soc., Faraday Trans. 1991, 87, 3179-
3184.
(54) Matsumura, K.; Hashimoto, S.; Yoshida, S. J. Catal. 1991, 131,
226-233.
(55) Matsumura, K.; Hashimoto, S.; Yoshida, S. J. Mol. Catal. 1991,
68, 73-86.
(56) Noyce, D. S.; Schiavelli, M. D. J. Am. Chem. Soc. 1968, 90, 1020-
1023.
(57) Schubert, W. M.; Keefe, J. R. J. Am. Chem. Soc. 1972, 94, 559-
566.
(58) Cozens, F. L. Ph.D. Thesis, University of Toronto, 1992.
(59) A similar low value has been reported for the adsorption of anethole
inside Na-ZSM-5.60
(25) Rooney, J. J.; Pink, R. C. Proc. Chem. Soc., London 1960, 70-71.
(26) Chen, F. R.; Fripiat, J. J. J. Phys. Chem. 1992, 96, 819-823.
(27) Chen, F. R.; Fripiat, J. J. J. Phys. Chem. 1993, 97, 5796-5797.
(28) Roduner, E.; Wu, L. M.; Crockett, R.; Rhodes, C. J. Catal. Lett.
1992, 14, 373-379.
(60) Lakkaraju, P. S.; Zhou, D.; Roth, H. D. J. Chem. Soc., Chem.
Commun. 1996, 2605-2606.
(61) Garcia, H.; Garcia, S.; Perez-Prieto, J.; Scaiano, J. C. J. Phys. Chem.
1996, 18158-18164.
(29) Liu, X.; Thomas, J. K. Langmuir 1993, 9, 727-732.
(30) Pollack, S. S.; Sprecher, R. F.; Frommell, E. A. J. Mol. Catal. 1991,
66, 195-203.
(62) Pretsch, E.; Clerc, T.; Seibl, J.; Simon, W. Tabellen zu¨r Struk-
turaufkla¨rung Organischer Verbindungen mit Spektroskopischen Metoden;
Springer Verlag: Berlin, 1976.
(31) Ramamurthy, V.; Caspar, J. V.; Corbin, D. R. J. Am. Chem. Soc.
1991, 113, 594-600.
(63) Xu, T.; Haw, J. F. J. Am. Chem. Soc. 1994, 116, 10188-10195.
(64) Das, P. K. Chem. ReV. 1993, 93, 119-144.
(65) Wilkinson, F.; Kelly, G. In Handbook of Organic Photochemistry;
Scaiano, J. C., Ed.; CRC Press: Boca Raton, FL, 1989; Vol. I; pp 293-
314.
(32) Caspar, J. V.; Ramamurthy, V.; Corbin, D. R. J. Am. Chem. Soc.
1991, 113, 600-610.
(33) Corma, A.; Forne´s, V.; Garc´ıa, H.; Mart´ı, V.; Miranda, M. A. Chem.
Mater. 1996. Accepted for publication.