Absolute reactivity of the 4-methoxycumyl cation in non-acid zeolites

Article abstract of DOI:10.1021/ja993619d

The reactivity of the 4-methoxycumyl cation in a series of alkali metal cation-exchanged zeolites (LiY, NaY, KY, RbY CsY, NaX, NaMor, and Naβ) in the absence and presence of coadsorbed alcohols and water is examined using nanosecond laser flash photolysis. In dry zeolites, the absolute reactivity of the carbocation is found to be strongly dependent on the nature of the alkali counterion, the Si/Al ratio, and the framework morphology, with the lifetime of the carbocation in Naβ being almost 10000-fold longer than in CsY. The results suggest a mechanism for carbocation decay involving direct participation of the zeolite framework as a nucleophile, leading to the generation of a framework-bound alkoxy species. Intrazeolite addition reactions of alcohols and water to the 4-methoxycumyl cation can be described in terms of both dynamic and static quenching involving molecular diffusion through the heterogeneous topology and rapid coupling between the alcohol and the carbocation encapsulated within the same cavity. The dynamics of the quenching reactions are different from similar reactions in homogeneous solution due to both the passive and active influences of the zeolite environment. In a passive sense, the zeolite decreases the reactivity of the nucleophilic quencher by hindering molecular diffusion. However, the zeolite actively promotes the efficiency of intracavity coupling by enhancing the deprotonation of the oxonium ion intermediate, allowing the reaction to go to completion.

Full text of DOI:10.1021/ja993619d

J. Am. Chem. Soc. 2000, 122, 6017-6027
6017
Absolute Reactivity of the 4-Methoxycumyl Cation in Non-Acid
Zeolites
Melanie A. O’Neill, Frances L. Cozens,* and Norman P. Schepp
Contribution from the Department of Chemistry, Dalhousie UniVersity, Halifax,
NoVa Scotia, Canada B3H 4J3
ReceiVed October 8, 1999
Abstract: The reactivity of the 4-methoxycumyl cation in a series of alkali metal cation-exchanged zeolites
(LiY, NaY, KY, RbY CsY, NaX, NaMor, and Naâ) in the absence and presence of coadsorbed alcohols and
water is examined using nanosecond laser flash photolysis. In dry zeolites, the absolute reactivity of the
carbocation is found to be strongly dependent on the nature of the alkali counterion, the Si/Al ratio, and the
framework morphology, with the lifetime of the carbocation in Naâ being almost 10000-fold longer than in
CsY. The results suggest a mechanism for carbocation decay involving direct participation of the zeolite
framework as a nucleophile, leading to the generation of a framework-bound alkoxy species. Intrazeolite addition
reactions of alcohols and water to the 4-methoxycumyl cation can be described in terms of both dynamic and
static quenching involving molecular diffusion through the heterogeneous topology and rapid coupling between
the alcohol and the carbocation encapsulated within the same cavity. The dynamics of the quenching reactions
are different from similar reactions in homogeneous solution due to both the passive and active influences of
the zeolite environment. In a passive sense, the zeolite decreases the reactivity of the nucleophilic quencher by
hindering molecular diffusion. However, the zeolite actively promotes the efficiency of intracavity coupling
by enhancing the deprotonation of the oxonium ion intermediate, allowing the reaction to go to completion.
Introduction
associated counter-balancing cations generates a polar matrix
with localized electrostatic fields and active acidic and basic
sites. As a result, there has been considerable interest in the
ability of these solid materials to modify and/or promote a
Studies of intrazeolite reaction dynamics continue to improve
our understanding of host-guest interactions and the role that
zeolites play in the mechanisms and kinetics of photochemical
16-28
variety of fundamental reactions in organic chemistry.
and thermal transformations.1
-11
Zeolites are microporous
In particular, much effort has been directed toward under-
standing the chemistry of carbocations generated within the
channels and cavities of zeolites. These studies have generally
4-
crystalline aluminosilicate materials constructed of [SiO4] and
5
-
[AlO4]
tetrahedra where every aluminum present in the
framework results in a net negative charge that must be balanced
with an exchangeable cation.12-15 The geometry of the open
framework structure leads to an interesting architecture with
an array of molecular-sized pores, channels, and cavities of
various shapes and sizes depending on the morphology of the
zeolite. This anionic network of well-defined void spaces and
been restricted to carbocations produced in proton-exchanged
7,11,29-45
Brønsted zeolites,
or zeolites with divalent metal ions
(16) Alvaro, M.; Garcia, H.; Garcia, S.; Marquez, F.; Scaiano, J. C. J.
Phys. Chem. B 1997, 101, 3043-3051.
(17) Corma, A.; Fornes, V.; Garcia, H.; Marti, V.; Miranda, M. A. Chem.
Mater. 1995, 7, 2136-2143.
(18) Baldovi, M. V.; Cozens, F. L.; Fornes, V.; Garcia, H.; Scaiano, J.
(
1) Turro, N. J. Pure Appl. Chem. 1986, 58, 1219-1228.
C. Chem. Mater. 1996, 8, 152-160.
(2) Ramamurthy, V.; Turro, N. J. J. Inclusion Phenom. Mol. Recognit.
(19) Cozens, F. L.; Regimbald, M.; Garica, H.; Scaiano, J. C. J. Phys.
Chem. 1996, 100, 18165-18172.
Chem. 1995, 21, 239-282.
3) Ramamurthy, V. In Photochemistry in Organized and Constrained
Media; Ramamurthy, V., Ed.; VCH: New York, 1991; pp 429-493.
4) Weiss, R. G.; Ramamurthy, V.; Hammond, G. S. Acc. Chem. Res.
993, 26, 530-536.
5) Ramamurthy, V.; Eaton, D. F.; Caspar, J. V. Acc. Chem. Res. 1992,
5, 299-307.
6) Dutta, P. K. J. Inclusion Phenom. Mol. Recognit. Chem. 1995, 21,
15-237.
(
(20) Cozens, F. L.; Cano, M. L.; Garcia, H.; Schepp, N. P. J. Am. Chem.
Soc. 1998, 120, 5667-5673.
(21) Ramamurthy, V.; Lakshminarasimhan, P.; Grey, C. P.; Johnston,
L. J. Chem. Commun. 1998, 2411-2425.
(
1
2
2
(
(22) Brancaleon, L.; Brousmiche, D.; Rao, V. J.; Johnston, L. J.;
Ramamurthy, V. J. Am. Chem. Soc. 1998, 120, 4926-4933.
(23) Turro, N. J.; Lei, X.; Li, W.; McDermott, A.; Abrams, L.; Ottavianai,
M. F.; Beard, H. S. Chem. Commun. 1998, 695-696.
(24) Turro, N. J.; McDermott, A.; Lei, X.; Li, W.; Abrams, L.; Ottavianai,
M. F.; Beard, H. S.; Houk, K. N.; Beno, B. R.; Lee, P. S. Chem. Commun.
1998, 697-698.
(
(
(
(
(
(
7) Corma, A. Chem. ReV. 1995, 95, 559-614.
8) Thomas, J. K. Chem. ReV. 1993, 93, 301-320.
9) Yoon, K. B. Chem. ReV. 1993, 93, 321-339.
10) Hattori, H. Chem. ReV. 1995, 95, 537-558.
11) Kiricsi, I.; Horst, F.; Tasi, G.; Nagy, J. B. Chem. ReV. 1999, 99,
(25) Yoon, K. B.; Park, Y. S.; Kochi, J. K. J. Am. Chem. Soc. 1996,
118, 12710-12718.
2
085-2114.
(26) Yoon, K. B.; Hubig, S. M.; Kochi, J. K. J. Phys. Chem. 1994, 98,
3865-3871.
(
12) Breck, D. W. J. Chem. Educ. 1964, 41, 678-689.
(13) Breck, D. W. Zeolite Molecular SieVes: Structure, Chemistry and
(27) Liu, X.; Iu, K.; Thomas, J. K. J. Phys. Chem. 1994, 98, 7877-
Use; John Wiley and Sons: New York, 1974.
7884.
(
14) Dyer, A. An Introduction to Zeolite Molecular SieVes; John Wiley
and Sons: Bath, 1988.
15) van Bekkum, H.; Flanigen, E. M.; Jansen, J. C. Introduction to
Zeolite Science and Practice: Elsevier: Amsterdam and New York, 1991.
(28) Hashimoto, S. J. Chem. Soc., Faraday Trans. 1997, 93, 4401-4408.
(29) Corma, A.; Garcia, H. Top. Catal. 1998, 127-140.
(30) Cano, M. L.; Corma, A.; Fornes, V.; Garcia, H.; Miranda, M. A.;
Baerlocher, C.; Lengauer, C. J. Am. Chem. Soc. 1996, 118, 11006-11013.
(
1
0.1021/ja993619d CCC: $19.00 © 2000 American Chemical Society
Published on Web 06/08/2000

6018 J. Am. Chem. Soc., Vol. 122, No. 25, 2000
O’Neill et al.
2
+
such as Ca that also contain Brønsted acid sites upon thermal
reduced by protonation. In addition, the stability of some
nucleophiles like alkenes will be low within the strongly acidic
environment of the proton-exchanged zeolites.
46-49
activation.
This is primarily due to the fact that in the highly
acidic environment provided by acid zeolites, carbocations are
easily generated and, more importantly, are thermodynamically
stabilized to the extent that they can be readily examined by
steady-state techniques such as solid-state NMR,
diffuse reflectance,
To address these issues, we have recently developed a method
to generate reactive carbocations within nonprotic cation-
38-41
50
UV-vis
exchanged zeolites. With this technique, we can now examine
29-37
and UV-vis absorption spectros-
the reactivity of carbocations in various zeolite environments.
As well, the dynamics and mechanisms for the addition of
nucleophiles such as alcohols can be investigated without having
to take into account problems associated with the protonation
state and stability of the nucleophile. In the present work, we
describe results concerning the effect of the alkali metal
counterion, aluminum content, and framework morphology on
the absolute reactivity of the 4-methoxycumyl cation embedded
within the cavities of cation-exchanged zeolites. Our results
demonstrate for the first time that the carbocation lifetime is
indeed strongly influenced by the nature, and in particular the
nucleophilicity, of the active site within these alkali metal cation-
exchanged zeolites. We also address questions concerning the
effect of zeolite composition on the rate constant for bimolecular
reactions of carbocations with alcohol nucleophiles within the
zeolite lattice.
4
2,43
copy.
While tremendous advances in our understanding of the effect
of zeolite structure and morphology on the chemistry of
carbocations have been made from studies in acid zeolites,
studies using nonprotic alkali metal cation-exchanged zeolites
would be extremely useful in providing further insights into
the nature of the interactions between the zeolite framework
and the embedded carbocation. For example, the effect of
properties such as electrostatic field strength and cation size on
the chemistry of carbocations located within zeolite cavities can
be readily examined using alkali metal exchanged zeolites, but
not with Brønsted-acid zeolites, simply by looking at the
behavior of carbocations as a function of different exchangeable
counterions. Furthermore, the [Si-O-Al] bridges in nonprotic
cation-exchanged zeolites have distinctly enhanced Lewis base
(i.e. nucleophilic) activity that is expected to be strongly
Results
dependent on the nature of the counterion. Thus, the possibility
that reactions of unstable carbocations within zeolites involve
a direct interaction between the anionic sites on the framework
and the carbocation to generate framework bound alkoxy species
can be explored with alkali metal cation-exchanged zeolites by
looking at the absolute reactivity of the carbocation as a function
of the nature of the charge balancing cation. In addition,
nonprotic cation-exchanged zeolites are suitable media for
determining the dynamics of second-order reactions of co-
incorporated nucleophiles with carbocations. These reactions
are difficult to study in Brønsted zeolites due to the reversible
addition of nucleophiles such as water and alcohols in the highly
acidic environment, as well as the probability that the nucleo-
philicity of many common nucleophiles will be considerably
Formation of the 4-Methoxycumyl Cation in Dry Alkali
Metal Cation-Exchanged Zeolites. Laser photolysis of 4,4′-
dimethoxybicumene incorporated within the cavities of NaY
-
3
under vacuum (10 Torr) conditions leads to the formation of
5
0,51
the 4-methoxycumyl cation, λmax ) 360 nm.
The 4-meth-
oxycumyl cation is formed via rapid fragmentation of the 4,4′-
5
2-54
dimethoxybicumene radical cation
that is generated upon
laser-induced photoionization of zeolite-encapsulated 4,4′-
5
0
dimethoxybicumene, eq 1. The same laser-initiated photore-
(
31) Cano, M. L.; Cozens, F. L.; Fornes, V.; Garcia, H.; Scaiano, J. C.
J. Phys. Chem. 1996, 100, 18145-18151.
32) Cano, M. L.; Cozens, F. L.; Garcia, H.; Vicente, M.; Scaiano, J. C.
J. Phys. Chem. 1996, 100, 18152-18157.
33) Garcia, H.; Garcia, S.; Perez-Prieto, J.; Scaiano, J. C. J. Phys. Chem.
996, 100, 18158-18164.
34) Cozens, F. L.; Garcia, H.; Scaiano, J. C. Langmuir 1994, 10, 2246-
249.
(
(
1
2
1
(
action readily generates the 4-methoxycumyl cation within other
alkali metal cation-exchanged Y zeolites (LiY, KY, RbY, and
CsY), and within zeolites that differ in their Si/Al ratio and
framework morphology (NaX, Si/Al ) 1.2; NaMordenite, Si/
Al ) 6.5; and Naâ, Si/Al ) 18), Table 1. In each case, the
transient diffuse reflectance spectrum is dominated by the
presence of a strong absorption band centered at 360 nm, Figure
(35) Cozens, F. L.; Garcia, H.; Scaiano, J. C. J. Am. Chem. Soc. 1993,
15, 11134-11140.
(36) Cozens, F. L.; Bogdanov, R.; Regimbald, M.; Garcia, H.; Marti,
V.; Scaiano, J. C. J. Phys. Chem. B 1997, 101, 6821-6829.
37) Liu, X. L.; Iu, K.; Thomas, J. K.; He, H.; Klinowski, J. J. Am. Chem.
Soc. 1994, 116, 118111-11811.
38) Nicholas, J. B.; Haw, J. F. J. Am. Chem. Soc. 1998, 120, 11804-
1805.
39) Xu, T.; Barich, D. H.; Goguen, P. W.; Song, W.; Wang, Z.; Nicholas,
J. B.; Haw, J. F. J. Am. Chem. Soc. 1998, 120, 4025-4026.
40) Haw, J. F.; Nicholas, J. B.; Xu, T.; Beck, L. W.; Ferguson, D. B.
(
(
1
(
1
, which is the known absorption maximum for the 4-meth-
oxycumyl cation, that behaves in a manner consistent with the
formation of the carbocation. In particular, the transient species
is not affected by the addition of molecular oxygen, but decays
more rapidly in the presence of nucleophiles such as water and
alcohols (vide infra). The possibility that the transient is the
4,4′-dimethoxybicumene radical cation can be ruled out as
(
Acc. Chem. Res. 1996, 29, 259-267.
(
(
41) Tao, T.; Maciel, G. E. J. Am. Chem. Soc. 1995, 117, 12889-12890.
42) Kiricsi, I.; Tasi, G.; Fejes, P.; Forster, H. J. Chem. Soc., Faraday
Trans. 1993, 89, 4221-4224.
(
7.
43) Forster, H.; Kiricsi, I.; Hannus, I. J. Mol. Struct. 1993, 296, 61-
6
(
(
(
44) van Santen, R. A.; Kramer, G. J. Chem. ReV. 1995, 95, 637-660.
45) Kazansky, V. B. Acc. Chem. Res. 1991, 24, 379-383.
(50) Cozens, F. L.; O’Neill, M.; Schepp, N. J. Am. Chem. Soc. 1997,
119, 7583-7584.
46) Kao, H.; Grey, C. P.; Pitchumani, K.; Lakshminarasimhan, P. H.;
Ramamurthy, V. J. Phys. Chem. A 1998, 102, 5627-5639.
47) Pitchumani, K.; Joy, A.; Prevost, N.; Ramamurthy, V. Chem.
Commun. 1997, 127-128.
48) Pitchumani, K.; Lakshminarasimhan, P. H.; Turner, G.; Bakker, M.
G.; Ramamurthy, V. Tetrahedron Lett. 1997, 38, 371-374.
49) Pitchumani, K.; Ramamurthy, V. Chem. Commun. 1996, 2763-
764.
(51) McClelland, R. A.; Chan, R. A.; Cozens, F. L.; Modro, A.; Steenken,
S. Angew. Chem., Int. Ed. Engl. 1991, 30, 1337-1339.
(52) Maslak, P.; Chapman, W. H., Jr. J. Org. Chem. 1996, 61, 2647-
2656.
(53) Maslak, P. Top. Curr. Chem. 1993, 168, 1-65.
(54) Maslak, P.; Chapman, W. H., Jr. Tetrahedron 1990, 46, 2715-
2724.
(
(
(
2

Absolute ReactiVity of the 4-Methoxycumyl Cation
Table 1. Description of Zeolites Used in the Present Study
J. Am. Chem. Soc., Vol. 122, No. 25, 2000 6019
zeolite
Si/Al
counterion
pore diameter/Å
channel system/cage diameter
+
a
Faujasite Y
Faujasite X
â
2.4
1.2
18
alkali M
7.4
7.4
3-D channel with cage/13 Å
3-D channel with cage/13 Å
tridirectional channel system
unidirectional channel system
+
Na
Na
+
7.6 × 6.4; 5.5 × 5.5
+
Mordenite
6.5
Na
6.5 × 7.0; 2.6 × 5.7
a
M+ ) Li , Na , K , Rb , Cs .
+
+
+
+
+
Further evidence for the mechanism of carbocation formation
comes from experiments carried out with a co-incorporated
photoinduced electron-transfer sensitizer. Figure 1 (insert) shows
the transient diffuse reflectance spectrum obtained upon 308-
nm laser photolysis of chloranil in the presence of 4,4′-
dimethoxybicumene in NaY under vacuum conditions, eq 2.
The spectrum clearly shows the presence of the 4-methoxycumyl
cation located at 360 nm, as well as bands at 420 and 450 nm
Figure 1. Transient diffuse reflectance spectra generated 320 ns after
66-nm laser irradiation of 4,4′-dimethoxybicumene in LiY and RbY
5
9,60
2
due to the chloranil radical anion.
at 500 nm due to the chloranil triplet
A short-lived weak band
is also observed under
-
3
60,61
under vacuum (10 Torr) (closed circles) and oxygen (open circles)
conditions. The insert shows the transient diffuse reflectance spectrum
generated upon 308-nm laser irradiation of chloranil in NaY with 4,4′-
vacuum conditions. Observation of the carbocation under these
photoinduced electron transfer conditions directly supports the
conclusion that the carbocation is produced by rapid fragmenta-
tion of the radical cation formed by photoionization of the
precursor in the absence of the sensitizer, eq 1.
-3
dimethoxybicumene under vacuum (10 Torr) conditions. The spectra
were taken 400 ns (closed circles), 3.32 µs (open circles), and 13.6 µs
(closed squares) after the laser pulse.
substituted anisole radical cations such as that of 4-methoxy-
Absolute Reactivity of the 4-Methoxycumyl Cation in Y
Zeolites. The absolute reactivity of the 4-methoxycumyl cation
in the various dry cation-exchanged zeolites was measured by
monitoring the intensity of the reflectance at 360 nm as a
function of time using nanosecond time-resolved diffuse re-
flectance. To get an accurate picture of the decay profile of the
carbocation within the heterogeneous environment of each
zeolite, kinetic decay traces obtained over several different time
scales were combined to generate stretched exponential traces
that represent the disappearance of the carbocation over several
55
toluene have absorption maxima near 420 nm in both solution
and zeolites,56 and these radical cations are not quenched by
nucleophiles such as alcohols (vide infra). In fact, our inability
to observe the radical cation on the nanosecond time scale was
expected, since previous studies have shown that the 4,4′-
dimethoxybicumene radical cation fragments in solution to
give the carbocation and the radical with a rate constant of ca.
8
-1 52,53
1
0 s .
A band at 290 nm is also observed in the time-resolved diffuse
reflectance spectra (Figure 1). The decay of this band is much
slower than the decay of the carbocation at 360 nm, indicating
the presence of a second transient species upon laser irradiation
of 4,4′-dimethoxybicumene. Unlike the carbocation band at 360
nm, the band at 290 nm is efficiently quenched by the presence
of molecular oxygen within the zeolite framework. We assign
this band to the 4-methoxycumyl radical produced as the second
fragmentation product upon cleavage of the 4,4′-dimethoxybi-
cumene radical cation, eq 1. Quenching of the transient by
oxygen and the location of the absorption maximum at 290 nm
are in agreement with this assignment.57 In addition, the
relatively long lifetime of this transient species is consistent
with that demonstrated for similar zeolite-encapsulated benzylic
radicals.58
62
decades. Stretched decay traces for the 4-methoxycumyl cation
within the zeolites with varying charge balancing cations, Si/
Al ratios, and framework topologies are shown in Figure 2. The
stretched decay profiles were analyzed using the exponential
63
series method to account for the possibility that the carbocation
might decay with a broad distribution of lifetimes. However,
these calculations gave a surprisingly narrow range of lifetimes
evenly distributed about one or two central values, Figure 2
inset, indicating that the decay of the 4-methoxycumyl cation
within these zeolites is made up of only one or two first-order
63
processes. Thus, the decay traces were fit to either a single or
double first-order equation as required. In LiY, the decay trace
fit well to a consecutive first-order exponential equation to give
(
59) Andre, J. J.; Weill, G. Mol. Phys. 1968, 15, 97-99.
(
55) Baciocchi, E.; Del Giacco, T.; Elisei, F. J. Am. Chem. Soc. 1993,
15, 12290-12295.
56) Supporting Information: see paragraph at the end of this paper
regarding availability.
57) Tokumura, K.; Ozaki, T.; Nosaka, H.; Saigusa, Y.; Itoh, M. J. Am.
Chem. Soc. 1991, 113, 4974-4980.
58) Cozens, F. L.; Ortiz, W.; Schepp, N. P. J. Am. Chem. Soc. 1998,
20, 13543-13544.
(60) Kawai, K.; Shirota, Y.; Tsubomura, H.; Mikawa, H. Bull. Chem.
Soc. Jpn. 1972, 45, 77-81.
1
(
(61) Kemp, D. R.; Porter, G. Chem. Commun. 1969, 1029-1030.
(62) The initial reflectance change did not remain completely constant
at the different time scales. No corrections were made to account for these
small differences, and this is reflected in the noise observed in the traces in
Figure 2.
(
(
1
(63) Wagner, B. D.; Ware, W. R. J. Phys. Chem. 1990, 94, 3489-3494.

6020 J. Am. Chem. Soc., Vol. 122, No. 25, 2000
O’Neill et al.
Table 3. Kinetic Isotope Effects for the Decay of the
4
-Methoxycumyl Cation in Alkali Cation Zeolites
zeolite
H D
k /k
LiY
NaY
KY
RbY
CsY
1.46 ( 0.20
1.30 ( 0.18
1.40 ( 0.20
1.43 ( 0.20
1.39 ( 0.20
Figure 2. Top: Decay at 360 nm for the 4-methoxycumyl cation
generated upon 266-nm excitation of 4,4′-dimethoxybicumene in the
alkali metal cation-exchanged Y zeolites. The insert shows the lifetime
distribution from the decay of the 4-methoxycumyl cation in RbY.
Bottom: Decay at 360 nm for the 4-methoxycumyl cation generated
upon 266-nm excitation of 4,4′-dimethoxybicumene in NaX, NaMor,
and Naâ.
Table 2. Lifetime(s) of the 4-Methoxycumyl Cation in Various
Alkali Cation Zeolites
Figure 3. Decay at 360 nm for the 4-methoxycumyl cation generated
upon 266-nm excitation of 4,4′-dimethoxybicumene in dry (closed
circles) and hydrated (open circles) (a) LiY, (b) RbY, (c) NaX, and (d)
Naâ.
a
charge on oxygenb
zeolite
lifetimes /µs
Si/Al
LiY
NaY
KY
RbY
CsY
NaX
NaMor
Naâ
26 (81%), 5.0 (19%)
4.6 (100%)
1.0 (74%), 4.5 (26%)
0.9 (58%), 0.3 (42%)
0.4 (100%)
2.4
2.4
2.4
2.4
2.4
1.2
6.5
18
-0.350
-0.358
-0.375
-0.378
-0.391
-0.413
-0.278
-0.240
lifetimes of ≈1000 µs in NaMor and ≈3000 µs in Naâ were
obtained in this way.
Kinetic Isotope Effects on the Absolute Reactivity of the
0.5 (100%)
4
4
-Methoxycumyl Cation in Y Zeolites. Laser irradiation of
,4′-dimethoxybicumene-d12 in each of the five alkali metal-
≈1000
≈3000
exchanged Y-zeolites yields transient diffuse reflectance spectra
showing the presence of two transients with maxima at 290 and
a
Lifetimes were calculated using the exponential series method or
by fitting the data to single or double first-order rate equations. The
values in the brackets refer to the fraction of the transient that decays
3
60 nm that are identical with those observed for the nondeu-
b
with the corresponding lifetime. Calculated using Sanderson elec-
terated analogue in each alkali metal Y zeolite under both
vacuum and oxygen conditions. The two transients can therefore
be readily identified as the deuterated 4-methoxycumyl cation
and 4-methoxycumyl radical.
tronegativity and the Sanderson equalization principle.
lifetimes of 26 and 5 µs. The major decay component making
up 81% of the overall decay had the longer 26 µs lifetime, while
the minor component at 19% had a 5 µs lifetime. The
carbocation in NaY was significantly shorter-lived and decayed
in a single-exponential fashion with a lifetime of 4.6 µs. In KY
and RbY, the carbocation decays fit best to double first-order
exponential equations. The shorter lifetime of 1 µs made up
the majority, 74%, of the decay in KY, while in RbY a shorter
lifetime of 0.3 µs contributed about 58% of the overall decay
and the slightly longer lifetime component of 0.9 µs made up
The decay traces for the deuterated 4-methoxycumyl cation
in the five alkali metal-exchanged zeolites are closely similar
to the decay traces for the nondeuterated analogue in the same
56
zeolites. Kinetics for the d -carbocation were calculated in the
6
same way as described previously for the h -carbocation, and
6
the isotope effects shown in Table 3 were calculated from the
ratio of the two rate constants, k /k . In each case, the observed
H
D
rate constants for the decay of the 4-methoxycumyl-d cation
6
are only slightly slower than the corresponding nondeuterated
4
2%. The weak decay in CsY fit to a single exponential with a
carbocation with kinetic isotope effects, k /k ) 1.4 ( 0.2.
H
D
lifetime of just 0.4 µs. These results are summarized in Table
Effect of Water on Reactivity of the 4-Methoxycumyl
Cation. As indicated by the weak absorption at 360 nm, Figure
3a,b, little or no 4-methoxycumyl cation is detectable after laser
irradiation of 4,4′-dimethoxybicumene when water is co-
incorporated into the cation-exchanged Y zeolites. This is
consistent with the lifetime of the 4-methoxycumyl cation being
reduced in the presence of water to the point where the
carbocation can no longer be resolved with our laser system.
The short lifetime can readily be explained by a quenching
process that involves nucleophilic addition of a water molecule
already present within the cavity in which the carbocation is
born. Such an intracavity reaction, typically called a static
2
4
, and overall show a clear trend in which the lifetime of the
-methoxycumyl cation decreases substantially upon going from
LiY to CsY.
In NaX, only one component with a lifetime of 0.5 µs was
observed for the decay of the 4-methoxycumyl cation. In both
NaMor and Naâ, little decay was observed, Figure 2, over 1
ms, the longest time scale of the laser system, indicating that
the carbocation is very long-lived in these zeolites. To obtain a
rough estimate of the lifetime within these two zeolites, the small
decay was fit to a first-order expression with the assumption
that the carbocation decays completely to the baseline. Long

Absolute ReactiVity of the 4-Methoxycumyl Cation
J. Am. Chem. Soc., Vol. 122, No. 25, 2000 6021
quenching process, is expected to be rapid since no diffusion
of the two reacting species is required.
We also considered the possibilities that water inhibits the
formation of the carbocation by rapidly quenching the 4,4′-
dimethoxybicumene radical cation prior to C-C bond cleavage,
or by interfering with the photoionization step. These possibili-
ties were explored by examining the behavior of the 4-meth-
oxytoluene radical cation generated by laser-induced photoion-
ization of 4-methoxytoluene incorporated within the zeolite
cavities. Our results show that the yield of the 4-methoxytoluene
radical cation is unaffected by the presence of water or methanol
within the zeolite cavities.56 In addition, the 4-methoxytoluene
radical cation is not quenched at all in the presence of
co-incorporated nucleophiles and continues to have a very long
lifetime of ∼10 µs within the zeolite cavities even when the
water or methanol content is high with greater than 2 molecules
of methanol per supercage. These results clearly suggest that
the low yield of carbocation derived from 4,4′-dimethoxybi-
cumene in the presence of water cannot be due to reduced
efficiency of photoionization, or to rapid quenching of the 4,4′-
dimethoxybicumene radical cation prior to C-C bond cleavage.
Static quenching of the 4-methoxycumyl cation by hydration
also occurred in NaMor, but quite different results were observed
in Naâ and NaX (Figure 3c,d). In Naâ, the addition of water
had no effect on the initial signal of the 4-methoxycumyl cation
indicating that no static quenching is taking place within this
zeolite. However, the lifetime of the carbocation was signifi-
cantly reduced by more than 2 orders of magnitude from ≈3000
µs to 25 µs. The effect of water in Naâ therefore exemplifies a
dynamic quenching process that requires diffusion together of
water and carbocation for the nucleophilic addition reaction to
occur. In NaX, hydration had little effect on the initial absorption
after the laser pulse and also had little effect on the lifetime of
the carbocation which decreased slightly from 0.5 µs under dry
conditions to 0.4 µs under hydrated conditions.
Figure 4. Relationship between (a) the observed rate constant for decay
and (b) the initial absorption at 360 nm for the 4-methoxycumyl cation
and the concentration of methanol in LiY (closed circles), NaY (open
circles), KY (closed squares), and RbY (open squares).
Table 4. Quenching Rate Constants for the Reaction of the
4-Methoxycumyl Cation with Methanol in Alkali Metal
Cation-Exchanged Zeolites
1
zeolite
rate constant/M- s-1
rel quenching constant
5
5
5
5
LiY
NaY
KY
RbY
CsY
(2.0 ( 0.2) × 10
(2.7 ( 0.3) × 10
(3.8 ( 0.4) × 10
(5.9 ( 0.6) × 10
no quenching
0.7
1
1.4
2.2
no quenching
Table 5. Relative Quenching Rate Constants for the Reaction of
the 4-Methoxycumyl Cation with Simple Alcohols in NaY
alcohol
methanol
ethanol
rel quenching constant
Intrazeolite Addition of Methanol to the 4-Methoxycumyl
Cation. The addition of methanol to LiY, NaY, KY, RbY, and
CsY had distinct effects on the behavior of the 4-methoxycumyl
cation. First, in each of the zeolites except CsY, the rate constant
for the decay of the carbocation increased with increasing
amounts of methanol. This behavior indicates the presence of
a dynamic quenching process whereby a methanol molecule in
a remote cavity first diffuses over a time scale of several
microseconds into the cavity containing the carbocation. Once
the two reacting species are confined within the same cavity,
an encounter complex is rapidly formed and then immediately
collapses to quench the carbocation. Except in CsY where no
dynamic quenching was observed, plots of the rate constant as
a function of methanol content were all linear (Figure 4a). Linear
least-squares analysis gave the second-order rate constants
summarized in Table 4. The rate constants are all very similar,
1
0.51
0.31
0.25
2
1
-propanol
-butanol
2-methyl-2-propanol
tert-butyl alcohol
0.19
no quenching
function of methanol content. A linear relationship is observed,
with the slopes changing slightly upon going from LiY to CsY.
In Naâ, the rate constant for the decay of the carbocation
increased dramatically upon the addition of small amounts of
methanol indicating the presence of dynamic quenching, while
no static quenching was observed. The opposite result was
obtained for NaMor, where static quenching by methanol was
seen, but dynamic quenching was not. In NaX no quenching of
the carbocation by methanol, either dynamic or static, was
observed.
Addition of Other Alcohols to the 4-Methoxycumyl Cation
in NaY. Similar results to those described above for the addition
of methanol to the 4-methoxycumyl cation in NaY were
observed upon quenching of the carbocation with ethanol and
2-propanol. However, in both cases, dynamic quenching is
considerably reduced relative to methanol. As shown in Table
5, quenching rate constants for ethanol and 2-propanol are about
2-fold and 3-fold, respectively, slower than that for methanol.
Dynamic quenching of the 4-methoxycumyl cation by either
1-butanol or 2-butanol was even less effective and reacted with
rate constants about 4-fold and 5-fold, respectively, slower than
5
-1 -1
5
-1 -1
ranging from 2.0 × 10 M
s
in LiY to 5.9 × 10 M
s
in RbY.
In addition to the dynamic quenching, the initial intensity of
the reflectance change immediately after the laser pulse at 360
nm due to absorption by the carbocation decreases significantly
in the presence of methanol. This was observed in all of the
cation-exchanged zeolites, including CsY. Since, as described
earlier, the yield and lifetime of the radical cation are not
influenced by methanol, the decrease represents a static quench-
ing process whereby some of the carbocations are rapidly
quenched by methanol in an intracavity reaction. As shown in
Figure 4b, the magnitude of the static quenching increases as a

6022 J. Am. Chem. Soc., Vol. 122, No. 25, 2000
O’Neill et al.
methanol, indicating that the efficiency of dynamic quenching
decreases with increasing chain branching. In the presence of
tert-butyl alcohol, no dynamic quenching was observed as the
carbocation lifetime was found to be independent of alcohol
concentration.
While the size of the alcohol seems to have a significant effect
on the dynamic quenching rate constants, the magnitude of the
static quenching remains relatively unaffected. Even though
there is some scatter in the static quenching data56 it is evident
that the decrease in the intensity of the carbocation signal at
radical by selective irradiation of di-tert-butyl peroxide within
NaY in the presence of cumene under a continuous flow of
nitrogen. Observation of significant ketone formation under these
conditions indicates that the concentration of oxygen within the
zeolite framework is sufficient to trap this radical even under a
continuous flow of nitrogen. In addition, to rule out the
possibility that the ketone was produced by secondary photolysis
of R-methyl-4-methoxystyrene, this olefin was irradiated in NaY
under a continuous flow of dry nitrogen. No detectable amounts
of 4-methoxyacetophenone were observed for conversions as
high as 35%, suggesting that the concentration of oxygen within
the zeolite framework in nitrogen-saturated NaY is insufficient
to lead to competitive olefin oxidation.
3
60 nm in the presence of the larger alcohols such as tert-butyl
alcohol is very similar to that observed for the smaller alcohols
such as methanol, Figure 4b, and ethanol in NaY.
Product Studies. Steady-state irradiation of 4,4′-dimethoxy-
bicumene under a continuous flow of dry nitrogen or argon in
NaY for ∼80 h followed by addition of methanol to the zeolite
composite and extraction with dichloromethane gave the five
compounds shown in eq 3. The products, which were identified
Discussion
Mechanism for Intrazeolite Decay of the 4-Methoxycumyl
Cation. Our results demonstrate unambiguously that the 4-meth-
oxycumyl cation has a lifetime in the microsecond to millisecond
range in dry alkali metal cation-exchanged zeolites, even under
conditions where coadsorbed nucleophiles are excluded. These
relatively short lifetimes can therefore be attributed to a direct
interaction between the carbocation and the zeolite host in which
it is encapsulated. In a manner analogous to the reactions of
carbocations in solution, the mechanism for this intrazeolite
decay can be envisioned as taking place via nucleophilic addition
or deprotonation pathways. For instance, the zeolite framework
may act as a nucleophile adding to the carbocation to generate
a zeolite-carbocation adduct. Alternatively, the zeolite may act
as a base, deprotonating the carbocation to give the correspond-
ing styrene.
Steady-state irradiation of 4,4′-dimethoxybicumene in NaY
clearly demonstrates that the major carbocation-derived products
obtained can be accounted for by nucleophilic addition (90%),
rather than deprotonation (10%). The role of the zeolite as a
base toward carbocation deprotonation is therefore a minor
pathway. The kinetic isotope effects are also consistent with
deprotonation playing only a minor role in the decay pathway
of the carbocation within zeolites. Deprotonation of cumyl
cations in solution using various bases typically shows a kinetic
and quantified by a combination of GC, GC/MS, and HPLC,
are derived from the 4-methoxycumyl radical and the 4-meth-
oxycumyl cation in a 1:1 ratio and are consistent with the rapid
fragmentation of the photogenerated 4,4′-dimethoxybicumene
radical cation as observed by laser studies within zeolites and
5
2,53
as previously measured for the radical cation in solution.
4
-Methoxycumyl alcohol and 4-methoxycumyl methyl ether
64
together account for 90% of the carbocation-derived products.
As methanol was only added to the zeolite sample after the
photolysis was complete, formation of the ether cannot be
attributed to direct attack of methanol on the free carbocation.
It must instead be produced via reaction of an initial carboca-
tion-nucleophile adduct, such as a long-lived framework-bound
alkoxy species (vide infra). A similar mechanism also can
account for the formation of the alcohol since water is adsorbed
by the zeolite composite following the photolysis and during
the extraction. However, although great care was taken to
conduct the photolysis under anhydrous conditions, the pos-
sibility that a small amount of water adsorbed during the lengthy
photolysis period trapped some of the free carbocation cannot
be ruled out. Interestingly, deprotonation of the 4-methoxycumyl
cation to give R-methyl-4-methoxystyrene is a minor pathway,
accounting for only 10% of the carbocation-derived products.
The most abundant radical photoproduct was 4-methoxyac-
etophenone (46%). Additional experiments were conducted to
establish that this product originates from oxygen-trapping of
the 4-methoxycumyl radical. First, acetophenone was produced
as a major photoproduct (ca. 50%) upon formation of the cumyl
6
5,66
isotope effect in the range of kH/kD ≈ 4-5.
The measured
kinetic isotope effects of kH/kD ) 1.4 ( 0.2 in all of the cation-
exchanged zeolites are close to unity. These values are consider-
ably less than that expected for a reaction in which the zeolite
acts as a base to remove a proton from the carbocation, and are
more consistent with values expected for a mechanism that is
6
5-67
dominated by nucleophilic addition.
Furthermore, the
similarity of the isotope effects in each of the zeolites examined
indicates that the mechanism is consistently nucleophilic addi-
tion and that the decreased lifetime as the zeolite counterion
size is increased cannot be due to enhanced rate constants for
deprotonation.
The 4-methoxycumyl alcohol and 4-methoxycumyl methyl
ether produced under steady-state photolysis conditions arise
from the addition of water and methanol, respectively. Since
the majority of the 4-methoxycumyl cations live only about ∼5
µs in NaY and methanol was added to the photolysis mixture
after irradiation, the generation of 4-methoxycumyl methyl ether
(64) As suggested by a referee, we considered the possibility the alcohol
and ether were derived from the R-methyl-4-methoxystyrene during the
workup, especially since a small amount of protons would be generated
upon formation of the styrene. To test for this possibility, the styrene was
incorporated in NaY containing added protons and methanol, and then
extracted in the same manner as in the product studies. No methyl ether
was formed, and only a small amount (<0.01%) of alcohol was produced.
These results indicate that the styrene is not converted to a significant extent
to the carbocation derived products under our reaction conditions.
(65) Thibblin, A. J. Phys. Org. Chem. 1989, 2, 15-25.
(66) Thibblin, A. Chem. Soc. ReV. 1993, 22, 427-433.
(67) The isotope effect for nucleophilic substitution should in principle
be closer to unity than that measured in the present work, and may even be
slightly less than one. Our measured value of kH/kD ) 1.4 ( 0.2 may
therefore include a component for deprotonation which, based on product
studies, accounts for about 10% of the decay of the carbocation and should
show a substantial normal isotope effect.

Absolute ReactiVity of the 4-Methoxycumyl Cation
J. Am. Chem. Soc., Vol. 122, No. 25, 2000 6023
cannot be attributed to nucleophilic trapping of the initially
formed free carbocation. Instead, the carbocation must decay
to some metastable product which subsequently reacts with
methanol to give the observed product. A reasonable mechanism
for carbocation decay, then, involves nucleophilic attack of the
zeolite framework on the free carbocation leading to the
formation of a framework-bound alkoxy species, eq 4. The
methyl ether is then formed by a solvolysis type reaction of the
alkoxy species upon the introduction of methanol.68-71
carbocations stabilized by intermolecular interactions and the
electrostatic fields within the zeolite, to covalently bound
intermediates deriving stabilization from the attachment to
electron rich framework sites. Ultimately the transition from
one extreme to the other will depend on the inherent stability
of the carbocation within the zeolite microenvironment, and on
the nucleophilicity exerted by the zeolite framework.
Reactivity of the 4-Methoxycumyl Cation in Dry Zeo-
lites: Zeolite Nucleophilicity. The lifetimes of the 4-meth-
oxycumyl cation in the various zeolites examined in the present
work are summarized in Table 2. The lifetimes span a
remarkably wide range, with the carbocation having a lifetime
that is almost 10000-fold longer in Naâ than in CsY! These
results therefore provide dramatic evidence that carbocation
lifetime within nonacid zeolites can be effectively controlled
by changing the counterion, Si/Al ratio, and morphology of the
zeolite.
The alcohol can also be produced through the same reaction
pathway, with water being adsorbed by the zeolite after
photolysis is complete. We cannot rule out the possibility that
a sufficient amount of water was adsorbed by the zeolite during
photolysis to trap the free carbocation directly and account for
all alcohol formation. However, our precautions in keeping the
zeolite dry prior to and during photolysis make this possibility
unlikely.
One trend that is evident from the data in Table 2 is that the
lifetime decreases in a systematic manner as the cation is varied
+
+
from Li to Cs . This trend clearly illustrates that the charge-
balancing cation must play an important role in regulating the
kinetic behavior of carbocations embedded within Y zeolites.
The mechanism by which the cation exerts its influence on the
carbocation reaction can be related to the effect of cation-
exchange in Y zeolites on the Lewis basicity of the [Si-O-
Al] active sites within the framework, an effect that has recently
There is considerable evidence from infrared72,73 and solid-
state NMR studies of carbocations in Brønsted acid zeo-
4
0,68-71,74-76
lites
and a variety of uni- and divalent metal
1
6,27,82,83
been emphasized in several studies.
In particular, the
7
7-80
exchanged zeolites
for the existence of framework-bound
+
Cs -exchanged Y zeolite is a considerably stronger Lewis base
than the Li -exchanged Y zeolite by virtue of the weak
interaction between the large diffuse Cs cation and the negative
alkoxy species. Within protic zeolites, such framework-bound
alkoxy species have been proposed as mechanistic alternatives
to long-lived free carbocations which achieve considerable
stabilization through covalent attachment to nucleophilic frame-
work sites.4 Thus with the exception of a few highly stabilized
carbocations that exist as persistent free species within Brønsted
acid zeolites, carbocations are trapped by the nucleophilic lattice
oxygens, either concurrent with carbocation formation (i.e. in
an SN2-like process) or after some finite lifetime as a transient
intermediate (i.e. an SN1-like process). Similar types of reaction
dynamics have been proposed to occur within nonprotic zeolites.
For instance, the conversion of methyl iodide to ethylene within
metal ion zeolites has been described as “acid-base bifunctional
+
+
charge at the [Si-O-Al] active site. This increased Lewis
+
+
basicity as the counterion is changed from Li to Cs is
manifested as a significant increase in nucleophilicity of the
framework oxygens and gives rise to the large 65-fold decrease
in lifetime of the carbocation.
0,81
In addition to being affected by the counterion, the kinetic
behavior of the 4-methoxycumyl cation is strongly influenced
by the Si/Al ratio. For example, the lifetime decreases consider-
ably by a factor of 9 upon going from NaY, which has a Si/Al
)
2.4, to NaX, with a Si/Al ) 1.2. This is consistent with the
7
7,78
notion that the framework participates directly in the quenching
of the carbocation by nucleophilic addition. In fact, numerous
studies have shown that the basicity in zeolites does increase
catalysis”
that can be formally envisioned as an SN2-like
process whereby the zeolite framework acts as a nucleophile
while the metal cation promotes loss of the leaving group. The
1
0,16,27,82-88
emerging picture is a continuum of intermediates,40 from free
as the Si/Al ratio decreases.
An increase in Lewis
base strength of the zeolite would result in an increase in the
nucleophilicity of the active sites which correlates nicely with
the increase in the reactivity of the 4-methoxycumyl cation in
NaX as compared to NaY. It must also be recognized that the
number of Si-O-Al bridges within the supercages of NaX is
greater than that in NaY. NMR studies have shown that
framework-bound carbocations covalently attach to the oxygen
atoms of Si-O-Al bridges, and theoretical studies have shown
that the oxygen atoms of Si-O-Al bridges bind carbocations
significantly more strongly than the oxygen atoms of Si-O-
(
68) Haw, J. F.; Richardson, B. R.; Oshiro, I. S.; Lazo, N. D.; Speed, J.
A. J. Am. Chem. Soc. 1989, 111, 2052-2058.
69) Lazo, N. D.; White, J. L.; Munson, E. J.; M., L.; Haw, J. F. J. Am.
Chem. Soc. 1990, 112, 4050-4052.
70) Aronson, M. T.; Gorte, R. J.; Farneth, W. E.; White, D. J. Am.
Chem. Soc. 1989, 111, 840-846.
71) Bronnimann, C. E.; Maciel, G. E. J. Am. Chem. Soc. 1986, 108,
154-7159.
72) Forester, T. R.; Howe, R. F. J. Am. Chem. Soc. 1987, 109, 5076-
082.
(
(
(
7
5
2
8
1
1
(
(
73) Ono, Y.; Mori, T. J. Chem. Soc., Faraday Trans. 1 1981, 77, 2209-
221.
(74) Oliver, F. G.; Munson, E. J.; Haw, J. F. J. Phys. Chem. 1992, 96,
106-8111.
(82) Lavalley, J.; Lamotte, J.; Travert, A.; Czyzniewska, J.; Ziolek, M.
J. Chem. Soc., Faraday Trans. 1 1998, 94, 331-335.
(83) Choi, S. Y.; Park, Y. S.; Hong, S. B.; Yoon, K. B. J. Am. Chem.
Soc. 1996, 118, 9377-9386.
(84) Barthomeuf, D. J. Phys. Chem. 1984, 88, 42-45.
(85) Haung, M.; Adnot, A.; Kaliaguine, S. J. Catal. 1992, 137, 322-
332.
(86) Okamoto, Y.; Ogawa, M.; Maezawa, A.; Imanaka, T. J. Catal. 1988,
112, 427-436.
(87) Weirzchowski, P. T.; Zatorski, L. W. Catal. Lett. 1991, 9, 411.
(88) Corma, A.; Forn e´ s, V.; Mart ´ı n-Aranda, R. M.; Garc ´ı a, H.; Primo,
J. Appl. Catal. 1990, 59, 237-248.
(75) Stepanov, A. G.; Zamaraev, K. I.; Thomas, J. M. Catal. Lett. 1992,
3, 407-422.
76) Tsiao, C.; Corbin, D. R.; Dybowski, C. J. Am. Chem. Soc. 1990,
12, 7140-7144.
77) Murray, D. K.; Howard, T.; Goguen, P. W.; Krawietz, T. R.; Haw,
J. F. J. Am. Chem. Soc. 1994, 116, 6354-6360.
78) Murray, D. K.; Chang, J.; Haw, J. F. J. Am. Chem. Soc. 1993, 115,
(
(
(
4
732-4741.
(79) Bosacek, V. Z. Phys. Chem. 1995, 189, 241-250.
(80) Bosacek, V. J. Phys. Chem. 1993, 97, 10732-10737.
(81) Xu, T. J. Am. Chem. Soc. 1994, 116, 10188-10195.

6024 J. Am. Chem. Soc., Vol. 122, No. 25, 2000
O’Neill et al.
groups to the Sanderson electronegativity of a series of nonprotic
79,80
zeolite frameworks.
The chemical shift trend was explained
in terms of zeolite nucleophilicity and the formation of the
framework-bound methoxy groups was attributed to direct
nucleophilic attack on the electrophilic guest molecule.
The ability of the various zeolites to stabilize the incorporated
carbocation must also be considered. The electrostatic fields
within the zeolite cavities are typically assumed to play an
important role in stabilizing charged organic compounds such
as carbocations. The strength of these fields depends on the
size of the counterion and decreases as the size of the counterion
Figure 5. Relationship between lifetime of the 4-methoxycumyl cation
and charge on framework oxygen.
98
increases. Thus, the observed decrease in the lifetime upon
going from LiY to CsY may, in part, reflect a decrease in the
inherent ability of the zeolites to support the presence of the
positively charged carbocation. On the other hand, for a given
counterion, the electrostatic field, and thus the stabilizing ability
of the zeolites, decreases as the Si/Al ratio increases. Thus, based
on electric field strength alone, the lifetime of the carbocation
should decrease as the Si/Al ratio increases. This is opposite
from that observed in the present work for NaX, NaY, NaMor,
and Naâ, and supports the notion that weaker nucleophilicity
and fewer nucleophilic sites as the Si/Al ratio increases play
the dominant role in determining the lifetime of the carbocation.
Si bridges.89,90 As a result, the Si-O-Al sites are considerably
more nucleophilic, and the smaller Si/Al ratio in NaX translates
into an increase in the number of potential nucleophilic sites
which presumably contributes to the higher reactivity of the
carbocation in NaX compared to NaY.
The 4-methoxycumyl cation is quite long-lived in NaMor and
Naâ, with estimated lifetimes of ca. 1000 and 3000 µs,
respectively. These zeolites have high Si/Al ratios (NaMor, Si/
Al ) 6.5 and Naâ, Si/Al ) 18) which suggests that the lower
reactivity of the carbocation is due to a combination of reduced
nucleophilicity at each active site and a large overall reduction
in the number of nucleophilic Si-O-Al sites relative to the
nonnucleophilic Si-O-Si sites. Another important factor that
must be considered, especially with respect to NaMor, is the
rigorous confinement of the carbocation within the channel-
like morphology (Table 1). This tight molecular confinement
may make it more difficult for the carbocation to locate an active
site and to adopt an orientation that satisfies the electronic and
steric requirements necessary to form the framework-bound
alkoxy species.
Nucleophilic Addition Reactions of Alcohols to the 4-Meth-
oxycumyl Cation. In the presence of coadsorbed methanol in
NaY the 4-methoxycumyl cation is quenched by two distinct
processes. Dynamic quenching is observed where the rate
constant for the disappearance of the 4-methoxycumyl cation
increases linearly with increasing concentration of methanol.
8
-1
Rapid static quenching (>10 s ) is also observed, such that
the amount of carbocation remaining after the 10 ns laser pulse
decreases with increasing methanol content. Observing static
quenching is significant because it suggests that a rapid reaction
takes place when the alcohol and the carbocation are confined
within the same cavity. As a result, it is reasonable to conclude
that dynamic quenching represents a process in which the
alcohol diffuses slowly into the cavity containing the carbocation
Since nucleophilicity is an important parameter that deter-
mines the lifetime of the carbocations within these solid
supports, the observed lifetime trend should follow the magni-
tude of the partial negative charge associated with the oxygen
atoms in the framework. Sanderson concepts of electronegativity
and charge density at the lattice oxygen atoms have previously
(Scheme 1, step a) and then undergoes a rapid intracavity
addition reaction (Scheme 1, step b), while static quenching
involves only the rapid intracavity reaction (Scheme 1, step b).
This contrasts sharply with the behavior observed in solution.
For instance, methanol quenching of the 4-methoxycumyl cation
in 2,2,2-trifluoroethanol (TFE) is activation controlled with
8
4,91-93
been applied to zeolites,
and more recently have been
used to describe basicity in alkali metal zeolites.2
7,83
Partial
charges on oxygen can be calculated using the Sanderson
equalization principle.9
4-97
These values are given in Table 2
6
-1 -1 99
typical bimolecular rate constants of ca. 5 × 10 M s ,
which are much smaller than the diffusional limit. Thus, the
encounter complex between the alcohol and the carbocation
proceeds more readily to product within the zeolite matrix than
in solution, and the zeolite actually enhances the reactivity of
the nucleophile toward static quenching of the carbocation. The
role of the zeolite may be largely passiVe, with the enhanced
reactivity being derived from the entropic effects associated with
molecular confinement. However, it is quite conceivable that
the zeolite is an actiVe participant, whereby the electron-rich
lattice oxygens promote nucleophilic addition by enhancing the
deprotonation of the nucleophile as it attacks the carbocation,
or by assisting the deprotonation of the resultant oxonium ion
adduct.
for all of the zeolites examined in the present work. A plot of
the carbocation lifetime as a function of the charge on the
framework oxygen is shown in Figure 5. When all of the zeolites
examined in the present work are included, a clear trend is
observed in which the lifetime increases as the charge on the
oxygen decreases, which is consistent with a decrease in overall
nucleophilicity of the zeolite. Interestingly, an analogous trend
has been recently observed in a study that correlated the
1
3
observed C chemical shifts of framework-bound methoxy
(
89) Kramer, G. J.; van Santen, R. A. J. Am. Chem. Soc. 1993, 115,
887-2897.
90) Mota, C. J. A.; Esteves, P. M.; de Amorim, M. B. J. Phys. Chem.
996, 100, 12418-12423.
91) Jacobs, P. A.; Mortier, W. J.; Uytterhoeven, J. B. J. Inorg. Nucl.
2
(
1
(
Despite the rapid intracavity reaction of the alcohol and the
carbocation, the second-order rate constants for the bimolecular
reactions between the alcohol and the carbocation are quite
small. This is certainly associated with the heterogeneous
topology of the framework, which hinders the mobility of the
Chem. 1978, 40, 1919-1923.
(
92) Mortier, W. J. Catal. 1978, 55, 138-145.
(93) Mortier, W.; Schoonheydt, R. A. Prog. Solid State Chem. 1985,
1
6, 1-125.
94) Sanderson, R. T. Chemical Bonds and Bond Energy, 2nd ed.;
Academic Press: New York, 1976.
(
(95) Sanderson, R. T. J. Am. Chem. Soc. 1983, 105, 2259-2261.
(96) Sanderson, R. T. J. Chem. Educ. 1988, 65, 112-118.
(97) Sanderson, R. T. J. Chem. Educ. 1988, 65, 227-231.
(98) Ward, J. W. J. Catal. 1968, 10, 34-46.
(99) Cozens, F. L. Ph.D. Thesis, University of Toronto, 1992.

Absolute ReactiVity of the 4-Methoxycumyl Cation
J. Am. Chem. Soc., Vol. 122, No. 25, 2000 6025
Scheme 1
quencher as it diffuses through the channels and cavities of the
zeolite. As such it exemplifies the capacity of zeolitic media to
influence bimolecular reactions of incorporated reagents in a
passive sense.
observed in NaY. In addition, the 4-methoxycumyl cation reacts
quite rapidly with the NaX framework. As a result, even if
dynamic methanol quenching in NaX was comparable to that
observed in NaY, the decrease in the lifetime of the carbocation
in NaX would be marginal and difficult to observe.
The importance of hindered diffusion is also evident when
comparing the effect of alcohol structure on the rate constants
for quenching of the 4-methoxycumyl cation in NaY. From the
data summarized in Table 5, it is evident that dynamic quenching
decreases by a factor of 5 as the size of the alcohol increases
from methanol to 2-methyl-1-propanol, and that no dynamic
quenching at all is observed with the bulky tert-butyl alcohol.
At the same time, static quenching is not significantly affected,
indicating that the actual nucleophilic addition step within the
cavity remains rapid even as the steric bulk of the alcohol is
considerably increased. Thus, the large decrease in the dynamic
quenching of the 4-methoxycumyl cation within the NaY zeolite
reflects a slower diffusion of the large alcohols through the
framework to a cavity containing the carbocation.
Unlike the situation in NaY where both dynamic and static
quenching were observed upon addition of methanol, only
dynamic quenching of the 4-methoxycumyl cation was observed
in Naâ. One possible explanation for the absence of static
quenching is that the carbocation is generated at a site within
the framework that is remote from any sites containing methanol.
While this situation might exist in Naâ due to its high Si/Al
ratio, the relatively large amount of methanol used in these
studies (up to 10 wt %) renders this explanation unlikely.
Alternatively, as suggested above, basic sites of the zeolite
framework might play a role in facilitating the nucleophilic
addition of methanol to the carbocation. Since the number of
such sites and the basicity of each site is low in Naâ, the ability
of Naâ to aid in the actual nucleophilic addition step is
considerably reduced compared to NaY, and rapid static
quenching is not observed. This is consistent with the fact that
the quenching rate constant in Naâ is 130-fold smaller than that
The observation that the rate constants for dynamic quenching
+
increase as the counterion becomes larger upon going from Li
+
to Rb suggests that the rate of diffusion of methanol increases
as the size of the counterion increases. This is not consistent
with a simple steric model for diffusion of alcohols within the
open framework of the zeolite that would result in a decrease
in the rate of diffusion as the counterion becomes larger.
However, other factors are also important in defining the
mobility of alcohols. In particular, recent studies have shown
that the electrostatic interaction between hydroxyl-containing
molecules and intrazeolite cations decreases considerably as the
size of the cations increases. For example, temperature-
1
04
in NaY. Thus, since the encounter complex in Naâ does not
rapidly collapse to quench the carbocation, dynamic quenching
is not diffusion limited, but instead becomes activation con-
trolled. In NaMor, dynamic quenching was not observed. In
this case, the mobility of the methanol is presumably inhibited
by the 1-dimensional morphology of the zeolite.
Reactivity of the 4-Methoxycumyl Cation in Zeolites:
Influence of Hydration. In the Y zeolites, the addition of water
(2-12 wt %) to the 4,4′-dimethoxybicumene/zeolite composites
prior to laser irradiation quenches almost all of the photoge-
nerated 4-methoxycumyl cations during the laser pulse, and the
lifetime of the carbocations under these conditions can therefore
be estimated to be <20 ns. Since the lifetime in the presence
of water is considerably less than that in the absence of water,
the dominant mode of decay of the carbocation upon hydration
is nucleophilic addition of water. The short lifetime presumably
reflects a significant degree of static quenching whereby the
carbocations are trapped rapidly by water present within the
same cavity in which the carbocation was initially generated.
The effects of water on the lifetime of the carbocation in NaX,
Naâ, and NaMor are essentially the same as those observed
upon the addition of methanol. The explanation for these effects
is therefore similar to that described above for methanol
quenching. Thus, in NaX, the water molecules are rendered less
nucleophilic due to the interactions with the large number of
sodium cations. In Naâ, the small number of active sites reduces
the ability of the framework to assist in the nucleophilic addition
reaction, while the 1-dimensional morphology of NaMor restricts
access of the water molecules to the reactive carbocation.
+
desorption studies have shown that water molecules in Li -
containing zeolites require up to 20 kJ/mol more energy for
desorption than water molecules from zeolites containing large
+
100,101
cations such as Rb .
As a result, inhibition of alcohol
migration due to electrostatic interactions will decrease as the
cation becomes larger, causing an increase in the diffusion of
the methanol molecules and a slight increase in the rate of the
dynamic quenching. In the case of CsY where no dynamic
quenching is observed, it is likely that unfavorable steric factors
+
dominate over the electrostatic binding, with the Cs counterion
being sufficiently large to effectively block the mobility of
+
methanol. Thus, diffusion of methanol around the large Cs
cations into cavities containing the carbocation is too slow to
compete with the very fast intracavity trapping of the carbocation
by nucleophilic addition to the CsY framework.
Similar arguments can be used to provide an explanation for
the absence of dynamic quenching in NaX. NaX contains a
significantly greater number of sodium ions than NaY. Since
methanol coordinates strongly to the metal cations, the efficiency
of migration is significantly reduced in NaX,102,103 and the
magnitude of dynamic quenching in NaX is less than that
(
100) Kirmse, A.; Karger, J.; Stallmach, F.; Hunger, B. Appl. Catal. A
999, 188, 241-246.
101) Hunger, B.; Klepel, O.; Kirchhock, C.; Heuchel, M.; Toufar, H.;
1
(
(104) The quenching rate constant in Naâ was measured to be 7.5 ×
4
-1 -1
7
-1 -1
Fuess, H. Langmuir 1999, 15, 5937-5941.
10 mg mol
s
which compares to a value of 1 × 10 mg mol
s
in
(
(
102) Ramamurthy, V. J. Am. Chem. Soc. 1994, 116, 1345-1351.
103) Karger, J.; Pfeifer, H. In NMR Techniques in Catalysis; Bell, A.
NaY. These values were determined from the relationship between observed
rate constants for the decay of the 4-methoxycumyl cation as a function of
moles of methanol per mg of Naâ or NaY.
T., Pines, A., Eds.; Marcel Dekker: New York, 1994; Chapter 2.

6026 J. Am. Chem. Soc., Vol. 122, No. 25, 2000
O’Neill et al.
Conclusions
Sample Preparation. 4,4′-Dimethoxybicumene was incorporated
into NaY using hexanes as a carrier solvent. Samples for laser
experiments were prepared such that the loading level was ap-
proximately one molecule in every 10 supercages. The typical procedure
involved introducing ca. 1 mL of a stock solution of the organic
compound to ca. 250 mg of activated zeolite in 20 mL of hexane. For
steady-state irradiations and product studies, larger zeolite samples with
slightly higher concentrations of the bicumene were used. Thus, 100-
Rapid fragmentation of the photogenerated 4,4′-dimethoxy-
bicumene radical cation is a convenient, efficient technique for
the generation and study of the reactive 4-methoxycumyl cation
in nonprotic zeolites. A likely mechanism for intrazeolite decay
of the 4-methoxycumyl cations involves direct attack of electron
rich lattice oxygens in the framework, and the carbocation
lifetime is thus strongly dependent on the zeolite nucleophilicity.
Such a mechanism is consistent with numerous studies in
Brønsted acidic zeolites, where many carbocations have been
shown to exist as framework-bound alkoxy species rather than
free ions. As a result, the absolute lifetime of the carbocation
within the zeolite matrix can be tuned by altering zeolite
characteristics such as the charge-balancing cation, Si/Al ratio,
and/or framework morphology which are directly correlated to
the nucleophilicity of the framework. Similarly, bimolecular
nucleophilic addition of alcohols and water to carbocations can
be modulated by simple variations in the zeolite composition
and morphology. This is again attributed to the fact that the
zeolite is not simply a passive medium for encapsulating
electrophilic reagents, but in fact exerts an active role in the
uni- and bimolecular chemistry of carbocation guests.
1
80 mg of compound was combined with 2-3 g of NaY resulting in
loading levels of approximately one molecule in every 2-3 supercages.
The hexane slurry was stirred for 1-3 h, after which time the suspension
was centrifuged and the isolated zeolite washed using hexane and dried
-3
under vacuum, 10 Torr. UV analysis of the combined decants
indicated that 100% of the organic precursor was incorporated within
NaY. During the incorporation of the precursor into the zeolite
framework the utmost care was taken to keep the zeolite sample dry.
After vacuum-drying for several hours, the zeolite samples were
transferred to laser cells in a nitrogen-filled glovebag and then
-3
reevacuated to 10 Torr for an additional 10-12 h before the photolysis
experiments. Samples for steady-state irradiation were placed in
photolysis cells under an atmosphere of nitrogen and purged with
nitrogen or argon for at least 30 min prior to irradiation. A continuous
flow of dry nitrogen or argon was maintained throughout the photolysis.
Detailed weight analysis studies were carried out on several NaY and
NaX samples to determine the percent weight water contained in the
zeolite after sample preparation. The results indicate that after sample
preparation as described for the diffuse reflectance laser studies, the
zeolite samples typically contain little, if any, water, certainly less than
Experimental.
Materials. 4,4′-Dimethoxybicumene was prepared by radical-
1
wt %.
initiated dimerization of 4-methoxycumene according to literature
The procedure for incorporation of 4,4′-dimethoxybicumene into the
1
05
procedures and purified by sublimation to yield a white solid (mp
other cation-exchanged zeolites was essentially identical with that for
incorporation into NaY. The initial amount of 4,4′-dimethoxybicumene
was adjusted to account for the reduced percent incorporation in the
larger alkali cation zeolites, such that the final concentration was one
molecule in every 10 supercages for all faujasites examined. The co-
incorporation of 4,4′-dimethoxybicumene and chloranil was accom-
plished by initially preparing a NaY sample containing chloranil using
dichloromethane as the carrier solvent. The loading level was one
chloranil per 10 zeolite cavities. After the sample was dried under
vacuum, 4,4′-dimethoxybicumene was incorporated into the chloranil/
zeolite composites in the same manner as described above.
Hydration Experiments. Zeolite samples were hydrated by exposing
a known quantity of the vacuum-dried 4,4′-dimethoxybicumene/zeolite
composite to the atmosphere. The sample was spread into a thin layer
on a shallow dish to allow for efficient hydration. The changes in sample
mass due to water uptake were monitored until no further change in
mass was observed. The final concentration of water in the hydrated
zeolite was dependent on the charge balancing cation, Si/Al ratio, and
framework morphology. Typical values of water uptake (expressed as
weight percent) by the 4,4′-dimethoxybicumene/zeolite composites are
ca. 12%, 10%, 6%, 3%, and 2% for the alkali metal zeolites LiY through
CsY, respectively, and ca. 5%, 4%, and 3% for NaX, NaMor, and Naâ,
respectively.
1
1
6
1
73-175 °C). H NMR (250 MHz, CDCl
3
) δ 6.85 (dd, 8H), 3.85 (s,
H), 1.3 (s, 12 H); C NMR (62.5 MHz, CDCl ) δ 25.3, 43.1, 55.2,
11.9, 129.6, 139.1, 157.3. 4,4′-Dimethoxybicumene-d12 was similarly
(1,1,1,3,3,3-hexa-
1
3
3
prepared by dimerization of 4-methoxycumene-d
6
deuterio-2-(4-methoxyphenyl)propane) and purified by column chro-
matography (1% ethyl acetate in hexanes). Recrystallization from
hexane-dichloromethane mixtures afforded a white crystalline solid
1
(
mp 173-175 °C). H NMR (250 MHz, CDCl
3
) δ 6.85 (dd, 8H), 3.85
) 23.3 (septet), 42.6, 55.2, 111.9,
29.6, 139.1, 157.3. The deuterated precursor, 4-methoxycumene-d
was prepared by Grignard addition of acetone-d to 4-bromoanisole
and subsequent reduction of the corresponding alcohol using trifluo-
1
3
(s, 6H); C NMR (62.5 MHz, CDCl
3
1
6
,
6
1
06
roacetic acid and sodium borohydride.
The alcohols, methanol,
ethanol, 2-propanol, 2-methyl-1-propanol, 1-butanol, and tert-butyl
alcohol, were of spectroscopic grade (Aldrich) and used as received.
NaY, Si/Al ) 2.4, and NaX, Si/Al ) 1.2, were obtained from
Aldrich. Naâ, Si/Al ) 18, and NaMordenite, Si/Al ) 6.5, were obtained
from the P.Q. Corporation. All zeolites were used as received after
activation at 400 °C for 16 h to remove any co-incorporated water.
The exchanged zeolites were prepared by stirring NaY with 1 M
aqueous solutions of the corresponding chlorides (LiCl, KCl, RbCl,
and CsCl) at 80 °C for 1 h. The zeolites were then washed until no
chlorides appeared in the washing water and dried under vacuum. This
procedure was repeated three times and the zeolite was calcinated
between washings. NaY has three types of exchangeable cations. The
number of cations per unit cell of zeolite Y are 16 site I cations, 32
site II cations, and 8 site III cations. It is well established that only site
II and III cations are accessible to guest molecules.2 The percent
Alcohol Experiments. Zeolite samples used in alcohol-quenching
experiments were prepared by vapor-phase adsorption of the neat
alcohol in dry zeolite/dimethoxybicumene composites. This typically
involved introducing the liquid alcohol in microliter increments with a
syringe through the top of a sealed laser cell containing a small quantity
(ca. 100-200 mg) of the composite. The laser cell was then heated
exchange is typically 47% for LiY, 97% for KY, 44% for RbY, and
slightly with a low-temperature heat gun to vaporize the alcohol and
the material was shaken to thoroughly mix the contents. The sample
was allowed to cool to room temperature before conducting the laser
experiments.
+
4
7% for CsY. It is known that for the larger cations such as Rb and
+
+
Cs only the accessible Na which occupy type II and type III sites
1
5
can be readily exchanged. Thus, the maximum accessible cation
exchange is ∼70%. Values of ∼50% exchange indicate that a small
percentage of type II cations are not completely exchanged. The low
percent exchange for the LiY sample used in this study is because
Product Studies. Steady-state photolyses were conducted by ir-
radiating 4,4′-dimethoxybicumene in NaY in a sealed quartz cell under
a continuous flow of either dry nitrogen or argon for periods of 72-
+
hydrated Li also does not readily exchange the type II cations.
9
6 h. The light source was a medium-pressure 450 W mercury lamp
passed through a quartz filter. Following the photolysis small amounts
of methanol (typically 100-200 µL) were injected into the irradiation
cell and heat vaporized. The resultant products were separated from
the zeolite by continuous extraction with dichloromethane for 24-48
(
105) Kratt, G.; Beckhaus, H. D.; Lindner, H. J.; Ruchardt, C. Chem.
Ber. 1983, 116, 3235-3263.
106) Refa Nutaitis, C. F.; Bernardo, J. E. Synth. Commun. 1990, 20,
87-493.
(
4

Absolute ReactiVity of the 4-Methoxycumyl Cation
J. Am. Chem. Soc., Vol. 122, No. 25, 2000 6027
h. The photochemical conversion was consistently about 10%, while
the mass balance varied slightly between 40 and 60%. The extracts
were analyzed by a combination of GC, GC/MS, and HPLC, and the
products were identified by comparison to known standards.
pressure (10- Torr) and sealed or purged with oxygen for a minimum
of 20 min prior to photolysis. The data analysis was based on the
fraction of reflected light absorbed by the transient (reflectance change
3
o o
∆J/J ) where J is the reflectance intensity before excitation and ∆J is
Time-Resolved Diffuse Reflectance. Time-resolved diffuse reflec-
tance experiments were carried out using a nanosecond laser system.
The excitation source was the fourth harmonic (266 nm, <8 mJ, <8
ns/pulse) from a Continuum NY61-10 YAG laser. A Lambda-Physik
excimer laser (308 nm, <100 mJ, 10 ns/pulse) was used as the excitation
source in the photosensitizer experiments. The transient signals from
the monochromator-PMT detection system were captured using a
Tektronix 620 digitizer and then transferred via a GPIB interface to a
Power Macintosh 7100 computer which controls the laser system using
a program written with LabView 3.0 software. The experimental setup
for time-resolved diffuse reflectance is similar to that described in detail
the change in the reflectance after excitation.
Acknowledgment. We gratefully acknowledge the Natural
Science and Engineering Research Council of Canada (NSERC)
and Dalhousie University for financial support of this research.
M.A.O. is the recipient of an NSERC graduate scholarship and
F.L.C. is the recipient of an NSERC-WFA.
Supporting Information Available: Figures S1-S4 of
spectral and rate data (PDF). This material is available free of
charge via the Internet at http://pubs.acs.org.
2
0
previously. The samples were contained in quartz cells constructed
2
with 3 × 7 mm tubing and were either evacuated under reduced
JA993619D