Incorporation of 2-arylhexa-1,5-diene into pentasil zeolite: A distorted 1-arylcyclohexane-l,4-diyl radical cation at room temperature

Article abstract of DOI:10.1021/jp046877s

Incorporation into a redox-active pentasil zeolite [(Na,H)-ZSM-5] converted 2-arylhexa-1,5-dienes (9a-c; aryl = phenyl, tolyl, anisyl) into 1-arylcyclohexane-1,4-diyl radical cations, 10a-c^ì?+. The ESR spectra of 10a-c^ì?+ (six lines,

Full text of DOI:10.1021/jp046877s

2504
J. Phys. Chem. B 2005, 109, 2504-2511
Incorporation of 2-Arylhexa-1,5-diene into Pentasil Zeolite: A Distorted
1-Arylcyclohexane-1,4-diyl Radical Cation at Room Temperature
Hiroshi Ikeda,*,† Tomonori Minegishi,† Tsutomu Miyashi,† Prasad S. Lakkaraju,‡
Ronald R. Sauers,§ and Heinz D. Roth*,§
Department of Chemistry, Graduate School of Science, Tohoku UniVersity, Sendai 980-8578, Japan,
Department of Chemistry, Georgian Court College, Lakewood, New Jersey 08701, and
Department of Chemistry and Chemical Biology, Rutgers UniVersity, Wright-Rieman Laboratories,
New Brunswick, New Jersey 08854-8087
ReceiVed: July 14, 2004; In Final Form: October 15, 2004
Incorporation into a redox-active pentasil zeolite [(Na,H)-ZSM-5] converted 2-arylhexa-1,5-dienes (9a-c; aryl
) phenyl, tolyl, anisyl) into 1-arylcyclohexane-1,4-diyl radical cations, 10a-c•+. The ESR spectra of 10a-c•+
(six lines, g ) 2.0026; a ) 9.0 G) indicated the presence of five essentially equivalent nuclei, indicating
limited delocalization of spin and charge into the phenyl group. Sequestered in the pores of ZSM-5, the three
species 10a-c•+ are stable at room temperature, in striking contrast to the parent radical cation in cryogenic
matrices: cyclohexane-1,4-diyl radical cation is converted to cyclohexene radical cation above 90 K. The
structures of radical cation 10a•+ (X ) H) and of the unsubstituted parent were probed by density functional
theory (DFT) and ab initio calculations.
Introduction
Chemical evidence for substituted radical cations of the
cyclohexane-1,4-diyl structure type was obtained in solution
during the electron transfer initiated photoreaction of 2,5-
diarylhexa-1,5-dienes, 4, by the distribution of a deuterium label
between the terminal olefinic (C-1, C-6-) and the allylic (C-3,
C-4-) positions.3b-e Diarylcyclohexane-1,4-diyl radical cations,
5•+, were intercepted by molecular oxygen (O2); the structure
of the resulting endo-peroxides revealed the steric course of the
ring closure. For example, the isomeric 1,4-diaryl-5,6-dimethyl
derivatives, 8, obtained from 3,6-diarylocta-2,6-dienes, 6,
indicate that 6•+ generated the chair form, 7•+. Accordingly,
the ring closure of hexa-1,5-diene radical cations occurs in the
same stereospecific manner3b-e established for the thermal
rearrangement of the neutral parent.4 The 1,4-diarylcyclohexane-
1,4-diyl radical cations exist in the chair form as does the parent
radical cation, 3•+, in cryogenic matrices.3a
One-electron oxidation converts hexadiene systems to a fam-
ily of interesting radical cation intermediates, which are related
to three mechanistic extremes of the Cope rearrangement. Pos-
sible pathways include a dissociative, an associative, and a con-
certed mechanism; radical cation structures representing all three
mechanisms have been characterized. For example, one-electron
oxidation of dicyclopentadiene yields radical cation, 1•+, a spe-
cies containing two noninteracting allyl functions. In this species
the cleavage of C-3-C-4 is complete, whereas bond formation
between C-1 and C-6 has not begun.1 Similar oxidation of tri-
cyclooctadienes (2, X ) s, CdO, CH2) generates radical cat-
ions, 2•+, containing two allyl groups in close contact: addition
and cleavage have proceeded to a similar degree (in “concert”).2
Finally, radiolysis of 1,5-hexadiene in cryogenic matrices gives
rise to cyclohexane-1,4-diyl radical cation, 3•+, in which
C-1-C-6-addition is complete without significant weakening
of the C-3-C-4-bond.3 These structures illustrate remarkable
differences between the potential surfaces of radical cations and
neutral precursors: states of intermediate geometry are minima
on the radical cation potential surface but saddle points
(transition structures) on the parent potential surface. In essence,
the parent molecules undergo concerted Cope rearrangements
via a transition structure,4 whereas the cycloaddition or cleavage
of the radical cations is “arrested” at intermediate geometries.
It appeared to be of interest to generate radical cations of the
hexa-1,5-diene family in a zeolite host to study their behavior
confined within the limiting interior. Various organic radical
cations can be generated spontaneously by inclusion of their
precursors into zeolites.5,6 The rigid microporous solids stabi-
lize the otherwise highly reactive radical cations due to the
* Corresponding authors.
† Tohoku University.
‡ Georgian Court College.
§ Rutgers University.
10.1021/jp046877s CCC: $30.25 © 2005 American Chemical Society
Published on Web 01/22/2005

Incorporation of 2-Arylhexa-1,5-diene into Pentasil Zeolite
J. Phys. Chem. B, Vol. 109, No. 7, 2005 2505
SCHEME 1
SCHEME 2
Figure 1. (top) X-band EPR generated by sequestering 2-p-anisylhexa-
1,5-diene, 9c, into a redox active pentasil zeolite, (Na,H)-ZSM-5. The
spectrum is split into a sextet (relative intensities ∼1:4:6:6:4:1) due to
1
hyperfine interaction with 5 essentially equivalent H nuclei (a ) 9.0
G) and is identified as that of 1-p-anisylcyclohexane-1,4-diyl radical
cation, 10c•+. (bottom) A simulated second-order spectrum consisting
of two isotropic spectra (in the ratio of 1:1): a spectrum with g )
2.0026 and an isotropic coupling constant, a(5H) ) 9.0 G and a
Lorentzian spectrum with a line width of 32.0 G.
combined effects of (a) the intense electrostatic fields inside
zeolites, (b) topological restrictions that prevent the access of
external reagents, and (c) the limiting dimensions of the zeolite
channels, which may restrict the shape of the enclosed inter-
mediates.6,7 Typically, radical cations generated in zeolites
have extended lifetimes and can, therefore, be studied by
conventional spectroscopic techniques. On the other hand, some
radical cations sequestered in zeolites undergo rapid conver-
sions, which are without precedent in cryogenic matrices or
in solution.8 For example, p-propylanisole gave rise to p-pro-
penylanisole radical cation;8a 2-phenyl-1,3-dithiane formed
1,2-dithiolane radical cation;8b trans-1,2-diphenylcyclopro-
pane was converted to exo,exo-1,3-diphenylallyl radical;8c and
p-cyclopropylanisole gave rise to p-propenylanisole radical
cation.8d
lithium; acid workup furnished the p-anisylcyclohexenone, 21c.
Reduction with sodium borohydride furnished the alcohol, 22c;
finally, the corresponding tosylate, 23c, was converted to the
cyclohexadiene, 12c, with potassium t.-butoxide in tetrahydro-
furan (Scheme 2).
Incorporation Procedures. The neutral diamagnetic sub-
strates were incorporated/adsorbed into pentasil zeolite or
mordenite by stirring solutions of 10 mg in 15 mL anhydrous
2,2,4-trimethylpentane in the presence of 250 mg thermally (500
°C, 12 h) dehydrated H-ZSM-5 for 30 min at room temperature.
The loaded zeolite was collected by filtration and washed with
dry n-hexane; the solids so obtained were dried at reduced
pressure (e1 Torr) for 1 h and stored in closed vials.
ESR Spectra. ESR spectra of the dried zeolite samples
were recorded on an X-band spectrometer (9.3 GHz) in the CW
mode. Typically, a single scan gave a satisfactory spectrum;
the spectra reproduced in the figures are accumulations of eight
scans.
In light of these findings we expected interesting results
from the incorporation of hexa-1,5-dienes into appropriate
zeolites. We selected 2-aryl-substituted derivatives, 9a-c,
because their redox potentials fall into the range that can be
oxidized by redox-active acidic pentasil zeolite (H-ZSM-5) or
mordenite (H-mor).
Results and Discussion
Three arylhexadienes, 2-phenyl- (9a, X ) H), 2-tolyl- (9b,
X ) CH3), or 2-anisylhexa-1,5-dienes (9c, X ) OCH3), were
incorporated from 2,2,4-trimethylpentane solutions into ther-
mally dehydrated samples of sodium or hydrogen pentasil zeolite
[(Na,H)-ZSM-5] or hydrogen mordenite (H-mor). Incorporation
into H-ZSM-5 and H-mor caused the colorless suspensions
to turn light blue and the zeolite surfaces more intense blue,
whereas the Na-ZSM-5 samples showed little or no coloration.
After washing and evaporating the solvent, the loaded H-ZSM-5
and H-mor samples showed ESR spectra, assigned to radical
cations, 10a-c•+, whereas the Na-ZSM-5 samples failed to
show any ESR signals with appreciable signal-to-noise ratios
(g3). Apparently Na-ZSM-5 is not sufficiently redox-active
to oxidize the substrates 9a-c. The ESR spectra observed in
H-ZSM-5 for 10a-c•+ show differing signal intensities, the
samples prepared from 9c (e.g., Figure 1) being the strongest
and those prepared from 9a being weakest; the EPR spectra
are persistent at room temperature. The spectrum assigned to
Experimental Section
Materials. H-ZSM-5 was prepared from synthetic
Na-ZSM-5 (hydrothermal crystallization of silica and alum-
ina gels in aqueous NaOH medium)10 by Na+-to-NH4 ion
+
exchange followed by 12 h deep-bed calcination at 500 °C
under air. Similarly, H-mor was prepared from commercial
NH4+-mor by 12 h calcination at 500 °C and stored in vacuo.
Commercial samples of 1-p-anisylcyclohex-1-ene, 11a (Aldrich),
and p-methoxybiphenyl, 13 (Aldrich), were used as received.
Donor molecule 9c was synthesized from p-anisaldehyde, 17c,
by reaction with 4-butenylmagnesium bromide, oxidation of the
resulting alcohol, 18c, with MnO2, and Wittig reaction of the
resulting ketone, 19c (Scheme 1).
Donor molecule 12c was synthesized from cyclohexane-1,4-
dione mono-ethylene ketal, 20, by the reaction with p-anisyl-

2506 J. Phys. Chem. B, Vol. 109, No. 7, 2005
Ikeda et al.
oxidation in the zeolite interior.8a,d The spectra obtained from
11 and 12 were essentially identical, suggesting that 11•+ is
readily converted to 12•+ in the zeolite channels. Because the
spectrum of 12•+ closely resembled the secondary spectrum in
H-mor (Figure 2, bottom), that spectrum was assigned as due
to 12•+. Further comparison with the ESR spectrum from the
sample prepared from 13c argued against the subsequent
conversion of 12c•+ to 13c•+.
In addition to the acyclic precursors, 9a-c, we also attempted
to generate the 1-arylcyclohexane-1,4-diyl system by allowing
1-anisyl-2,3-diazabicyclo[2.2.2]oct-2-ene, 14c, to interact with
H-ZSM-5. Loss of nitrogen from the potential radical cation,
14c•+, would also give rise to 10c•+. Of course, the rigid
structure of this precursor might prevent its incorporation into
the zeolite channels. This attempt gave rise to a weak, broad
ESR spectrum with essentially no fine structure. The absence
of any characteristic 14N splitting suggested that the spectrum
might be due to a deazetized species; obviously 14c was not
converted to 10c•+ in the zeolite interior. The broad signal might
be due to a species on the external zeolite surface, generated
by oxidation or protonation on an appropriate zeolite site.
Figure 2. X-band EPR spectra obtained upon sequestering 2-p-
anisylhexa-1,5-diene, 9c, into (Na,H)-mor (center), and of the same
sample after 6 h at room temperature (bottom). Comparison with the
spectrum obtained in (Na,H)-ZSM-5 (top) shows that species 10c•+ is
present but decays slowly to a secondary species, assigned to 1-p-
anisylcyclohexadiene radical cation, 12c•+ (bottom), by comparison with
a spectrum obtained by incorporation of authentic 12c.
The six-line spectra (Figure 1, top) obtained from 2-p-
anisylhexa-1,5-diene, 9c, and from 9a,b in the zeolite indicate
that five H nuclei have equivalent or very similar hyperfine
1
coupling constants (hfcc). Such spectra are formally compatible
with the chair conformer of cyclized radical cations, e.g., 1-p-
anisylcyclohexane-1,4-diyl radical cation, 10c•+, in which the
R-proton at C-4 and the two pairs of axial â-protons (H2,6ax
and H3,5ax) would interact strongly with the unpaired spin. The
generation of 10a-c•+ from 9a-c is analogous to the formation
of 3•+ from 1,5-hexadiene in cryogenic matrices,3a however,
with two significant noteworthy features: the persistence,
especially of 10c•+, and the unexpected apparent equivalence
of the hfccs for the four axial â-protons.
10c•+ obtained from 9c (g ) 2.0026, a ) 9.0 G), indicated the
presence of five essentially equivalent nuclei.
The observation that 10a-c•+ are persistent at room tem-
perature is surprising, given that 3•+ was converted to cyclo-
hexene radical cation above 90 K, by a (1,3-) shift of an axial
hydrogen or its mechanistic equivalent.3a The persistence of
10a-c•+ is all the more remarkable, as deprotonations, dehy-
drogenations, or net hydrogen migrations readily occur in zeo-
lites. For example, oximes readily form iminoxyls,9 p-propyl-
anisole gives rise to p-propenylanisole radical cation,8a and
p-anisylcyclopropane is converted to 1-p-anisylpropene radical
cation.8d In light of these results, one might have expected
radical cations 10a-c•+ to rearrange to 1-p-arylcyclohexene
radical cations, 11a-c•+, or be dehydrogenated to 1-p-arylcy-
clohexa-1,3-dienes, 12a-c•+, or p-substituted biphenyl radical
cations, 13a-c•+. The unexpected stability of 10a-c•+ in
ZSM-5 can be explained if it is generated in an alignment that
is unfavorable for deprotonation/dehydrogenation. Thus, the
stabilization would be ascribed to the limiting geometry of
ZSM-5.
The ESR spectra arising from incorporation of 9a-c into
(Na,H)-mor were similar to those obtained in H-ZSM-5;
however, they were less clearly defined and had somewhat
reduced wing signals (e.g., Figure 2, center, from 9c). The EPR
spectrum of 10c•+ observed initially in H-mor (Figure 2, center)
changed slowly over a period of several hours. The outer (wing)
signals of 10c•+ decayed further, and the overall spectrum was
replaced by a secondary spectrum, a poorly resolved multiplet
(Figure 2, bottom).
In an attempt to probe the identity of the secondary species
observed in H-mor, we incorporated three potential products,
1-p-anisylcyclohexene, 11c, 1-p-anisylcyclohexa-1,3-diene, 12c,
and p-methoxybiphenyl, 13c, into H-ZSM-5 and compared the
resulting spectra with that spectrum (Figure 2, bottom). These
substrates were chosen because results of previous studies
suggested that a radical cation of structure 10c•+ might undergo
a combination of deprotonation-oxidation and dehydrogenation-
Support for this interpretation is derived from the observation
that 10a-c•+ are indeed dehydrogenated in acidic mordenite,
which has larger pores than pentasil zeolite and, thus, allows
the sequestered entity greater mobility and flexibility. The EPR
spectrum of 10c•+ observed initially upon incorporation of 9c
into H-mor (Figure 2, center) changes slowly over a period of

Incorporation of 2-Arylhexa-1,5-diene into Pentasil Zeolite
J. Phys. Chem. B, Vol. 109, No. 7, 2005 2507
TABLE 1: C-C Bond Lengths (Å) Calculated for Cyclohexane-1,4-diyl (3•+) and 1-Phenyl-cyclohexane-1,4-diyl Radical
Cations (10c•+)
1-Phenylcyclohexane-1,4-diyl
Cyclohexane-1,4-diyl
conjugated
rotated 60˚
orthogonalb
pseudo twist boat
bond
UMP2
UB3LYP
UB3LYPc
UMP2
UB3LYP
UMP2
UB3LYP
UMP2
UB3LYP
UMP2
UB3LYP
C1-C2
C2-C3
C3-C4
C4-C5
C5-C6
C1-C6
C1-C1′
1.449
1.664
1.449
1.449
1.664
1.449
1.454
1.672
1.454
1.454
1.672
1.454
1.451
1.672
1.451
1.451
1.672
1.451
1.476
1.579
1.485
1.485
1.579
1.476
1.421
1.483
1.614
1.472
1.472
1.614
1.483
1.426
1.449
1.579
1.449
1.49
1.664
1.476
1.430
1.464
1.655
1.461
1.461
1.655
1.464
1.476
1.467
1.669
1.439
1.439
1.660
1.467
1.472
1.464
1.655
1.461
1.461
1.655
1.464
1.476
1.504
1.534
1.436
1.436
1.534
1.404
1.426
1.489
1.553
1.485
1.487
1.540
1.503
1.422
a Unless otherwise noted, all calculations were carried out with the 6-31G* basis set. b Optimized with the phenyl group held rigidly in the plane
bisecting the cyclohexanediyl ring system through C1 and C4. c Calculated with the 6-311G* basis set.
several hours. The outer (wing) signals of 10c•+ decay further,
and the overall spectrum is replaced by a secondary spectrum
(Figure 2, bottom) assigned to 1-p-anisylcyclohexa-1,3-diene
radical cation, 12c•+. This conversion requires the net loss of
H2 or of two electrons and two protons in any sequence or
combination. Mechanistic details about this conversion, e.g.,
whether 1-arylcyclohexene radical cation, 11•+, is an intermedi-
ate, could not be determined.
The second unexpected finding, that the two pairs of axial
â-protons have essentially identical hfccs, requires that the
unpaired spin is distributed essentially evenly between the 2p
orbitals at C-1 and C-4 and, notably, that the distribution of
spin and charge in the radical cation is not significantly affected
by the aryl group at one of the spin-bearing centers. The fact
that the coupling constant of the five 1H nuclei is only slightly
reduced (9.0 G) relative to those of 3•+ (11.9 G)3a also seems
to suggest limited delocalization of unpaired spin into the anisyl
group. Further, the hfccs for the anisyl (as well as the equatorial)
protons fall within the line width of the spectrum (∼8.0 G).
Although the lines are broad, this width is given by the sum of
the hfcs of all these protons. The apparent limited delocalization
could be due to rotation of the aryl moiety from the conjugated
orientation, 10•+con, into an orthogonal one, 10•+ort. This result,
too, could be ascribed to the limiting geometry of the zeolite.
methodology as incorporated in the NBO program;14 how-
ever, the resulting charge densities provide limited insights
(Table 4).
In our experience, Møller-Plesset perturbation theory (MP2)
1
reproduces positiVe H hyperfine coupling constants satisfac-
torily, but overestimates spin densities on carbon and negative
hfcs significantly, in some cases by factors >2.15 On the other
hand, density functional theory methods16 often give satisfactory
agreement17 with experimental results. Indeed, positive and
negative hfcs of norbornadiene,17a,b quadricyclane,17a,b and bi-
cyclobutane radical cations17b are reproduced accurately with
both (B3LYP/6-31G*//MP2/6-31G*) and (B3LYP/6-31G*//
B3LYP/6-31G*) methods.17b
Cyclohexane-1,4-diyl Radical Cation. The radical cation,
3•+, has a 2Y′ electronic state and Cs symmetry; the minimization
converges to this symmetry if no symmetry is imposed. The
carbon skeleton has four shorter [C-1-C-2, C-1-C-6,
C-3-C-4, C-4-C-5 ) 1.449 Å (UMP2); ) 1.454 Å (UB3LYP)]
and two “long” C-C bonds [(C-2-C-3, C-5-C-6 ) 1.664 Å
(UMP2); ) 1.672 Å (UB3LYP)]. These bond lengths (Table
1) reflect an intermediate in which the C-3-C-4-bond is partially
cleaved and the C-1-C-6-bond is partially formed. The presence
of equivalent spin density on C-1 and C-4 supports a delocalized
structure. The actual spin densities on these carbons depend on
the level of perturbation theory included; they are higher with
the MP2 (F1,4 ) 0.60) than with the B3LYP method (F1,4
0.56; Table 2).
)
Not surprisingly, the structure type derived by considering
the unpaired spin density distribution is also fully revealed in
the hyperfine coupling pattern. The protons at the sp2 hybridized
carbons, C-1 and C-4, show significant negative hfcs, whereas
the two pairs of axial â-protons have sizable positive hfcs; the
In an attempt to probe this aspect we carried out DFT and ab
initio calculations on the 1-phenylcyclohexane-1,4-diyl radical
cation, 10a•+, in two limiting orientations, 10a•+con and 10a•+
;
ort
equatorial â-protons have minor or negligible hfcs [|a2,3,5,6eq
|
additional conformers were also considered. For comparison and
to calibrate the calculations, the unsubstituted cyclohexane-1,4-
diyl radical cation, 3•+, was calculated.
< 1.5 G (UMP2); < 0.3 G (UB3LYP)]. The degree of
delocalization depends on the level of perturbation theory (Table
3). At the MP2 level, the negative coupling (a1,4 ) -15.2 G)
is slightly larger than the positive one (a2,3,5,6ax ) 12.2 G),
whereas at the B3LYP level the negative and the major positive
couplings are more similar (a1,4 ) -12.5 G; a2,3,5,6ax ) 13.5
G). These results are in satisfactory agreement with the
experimental values (|a| ) 12.0 G).3a A calculation using the
6-311G* basis reduced the negative hfccs but left the positive
ones essentially unchanged (a1,4 ) -11.3 G; a2,3,5,6ax ) 13.5
G), widening the mismatch between them. In summary, the
UB3LYP calculations reproduce the hyperfine coupling con-
stants and, by inference, the structure of cyclohexane-1,4-diyl
radical cation satisfactorily.
Computational Details. Ab initio and density functional
theory (DFT) calculations for the radical cations, 3•+, 10a•+
,
con
and 10a•+ort, were carried out with the Gaussian 03 suite of
electronic structure programs11 using extended basis sets, in-
cluding p-type polarization functions on carbon. The geometries
of the radical cations were optimized at the unrestricted
B3LYP12 level with the standard 6-31G* basis set. The im-
portance of higher degrees of electron correlation was in-
vestigated at the unrestricted MP213 level of theory (UMP2/
6-31G*//UMP2/6-31G*). Vibrational analyses confirmed two
DFT stationary points as energy minima (no imaginary frequen-
cies). Wave function analyses for charge and spin density
distributions used the conventional Mulliken partitioning scheme.11
In addition, atomic charges were also calculated using NPA
1-Phenylcyclohexane-1,4-diyl Radical Cation. Calculating
the phenyl-substituted radical cation, 10a•+con, proved more

2508 J. Phys. Chem. B, Vol. 109, No. 7, 2005
Ikeda et al.
TABLE 2: Mulliken Atomic Spin Densities (G) Calculated for Cyclohexane-1,4-diyl (3•+) and 1-Phenylcyclohexane-1,4-diyl
Radical Cations (10c•+)a
1-Phenylcyclohexane-1,4-diyl
Cyclohexane-1,4-diyl
conjugated
rotated 60°
orthogonalb
pseudo twist boat
bond
UMP2
UB3LYP
UB3LYPc
UMP2
UB3LYP
UMP2
UB3LYP
UMP2
UB3LYP
UMP2
UB3LYP
C1
C2
C3
C4
C5
C6
0.596
-0.040
-0.040
0.596
-0.040
-0.040
0.563
-0.031
-0.031
0.563
-0.031
-0.031
0.559
-0.030
-0.030
0.559
-0.030
-0.030
0.221
-0.007
-0.082
1.061
-0.082
-0.007
0.192
0.014
-0.061
0.784
-0.061
0.014
0.364
-0.035
-0.047
0.811
-0.074
-0.032
0.320
-0.002
-0.052
0.712
-0.052
-0.084
0.746
-0.072
-0.007
0.401
-0.007
-0.072
0.436
-0.011
-0.044
0.659
0.044
-0.011
0.789
0.071
0.003
0.010
0.003
0.071
0.108
0.003
-0.069
0.898
-0.069
0.005
a Unless otherwise noted, all calculations were carried out with the 6-31G* basis set. b Optimized with the phenyl group held rigidly in the plane
bisecting the cyclohexanediyl ring system through C1 and C4. c Calculated with the 6-311G* basis set.
TABLE 3: Calculated Hyperfine Coupling Constants (a, G) for Cyclohexane-1,4-diyl (3•+) and 1-Phenylcyclohexane-1,4-diyl
Radical Cations (10c•+)a
1-Phenylcyclohexane-1,4-diyl
Cyclohexane-1,4-diyl
conjugated
rotated 60°
orthogonalb
pseudo twist boat
proton
UMP2
UB3LYP
UB3LYPc
UMP2
UB3LYP
UMP2
UB3LYP
UMP2
UB3LYP
UMP2
UB3LYP
H1
-15.22
12.20
-1.47
12.20
-1.47
-15.22
12.20
-1.47
12.20
-12.45
13.47
-0.22
13.47
-0.22
-12.45
13.47
-0.22
13.47
-11.25
13.48
-0.27
13.48
-0.27
-11.25
13.48
-0.27
13.48
H2ax
H2eq
H3ax
H3eq
H4
H5ax
H5eq
H6ax
H6eq
2.28
0.46
37.39
7.48
-29.92
37.39
7.48
2.79
0.14
24.05
1.93
-18.35
24.05
1.93
10.73
3.10
8.00
-0.21
18.78
0.48
-16.11
18.06
0.37
17.7
0.02
5.68
-3.00
-7.63
5.68
-3.00
17.7
0.02
9.32
-0.14
17.4
-0.41
-14.25
17.4
0.41
9.32
-0.14
28.33
13.75
-1.07
1.17
-0.43
-1.06
1.19
7.61
-0.54
38.51
10.35
-21.44
34.29
30.86
-0.88
1.14
14.35
-3.07
-18.92
12.59
-4.66
16.28
5.61
2.28
0.46
2.79
0.14
16.54
0.67
28.36
16.68
-1.47
-0.22
-0.27
a Unless otherwise noted, all calculations were carried out with the 6-31G* basis set. b Optimized with the phenyl group held rigidly in the plane
bisecting the cyclohexanediyl ring system through C1 and C4. c Calculated with the 6-311G* basis set.
TABLE 4: Atomic Charges (f) Calculated for Cyclohexane-1,4-diyl (3•+) and 1-Phenylcyclohexane-1,4-diyl Radical Cations
(10c•+)a
1-Phenylcyclohexane-1,4-diyl
Cyclohexane-1,4-diyl
UB3LYP
conjugated
UB3LYP
orthogonalb
atoms
UMP2
UB3LYPc
UMP2
UMP2
UB3LYP
C1
0.004
-0.405
-0.405
0.004
0.266
0.254
0.266
0.254
0.266
0.266
0.009
-0.516
-0.516
0.009
0.263
0.288
0.304
0.288
0.304
0.263
0.017
-0.349
-0.349
0.017
0.238
0.233
0.238
0.233
0.238
0.238
0.087
-0.514
-0.514
0.087
0.261
0.290
0.300
0.290
0.300
0.261
0.154
-0.404
-0.354
-0.137
0.324
-0.517
-0.501
-0.113
C2,6
C3,5
C4
9.32
-0.14
17.4
-0.41
-14.25
H1
H2,6ax
H2,6eq
H3,5ax
H3,5eq
H4
0.238
0.224
0.205
0.233
0.231
0.288
0.271
0.258
0.285
0.246
a Unless otherwise noted, all calculations were carried out with the 6-31G* basis set. b Optimized with the phenyl group held rigidly the plane
bisecting the cyclohexandiyl ring system through C1 and C4. c Calculated with the 6-311G* basis set.
problematic than the parent, 3•+, because of the increased
number of heavy atoms and because of having to impose
constraints on the geometry of the phenyl group. The dimen-
sions of the species suggest that it should be readily accom-
modated in ZSM-5. For this consideration the longest exten-
sion, the distance l ) H4-Hp ) 8.82 Å, is less important;
however, “depth”, d ) H2eq-H6eq ) 4.33 Å, and “height”, h )
H2ax-H5ax ) 3.07 Å, are compatible with the zeolite dimensions
(Figure 3).
) 1.06 (UMP2), F4 ) 0.78 (UB3LYP)] and it is delocalized
unto C-1 only to a limited degree [F1 ) 0.22 (UMP2), F1 )
0.19 (UB3LYP), Table 2]. Given this distribution of electron
spin density, the hfcc of H-4 and those of the two pairs of axial
1H nuclei are significantly different, particularly with Møller-
Plesset perturbation theory [a4 ) -29.9 G, a2,6ax ) 2.3 G, a3,5ax
) 37.4 G (UMP2); a4 ) -18.4 G, a2,6ax ) 2.8 G, a3,5ax ) 24.1
G (UB3LYP)]. Obviously, this splitting pattern is incompatible
with the experimental results.
Given that 10a•+ can be accommodated inside pentasil
One possible explanation for the mismatch between calcula-
tion and experiment involves a structure in which the zeolite
has caused the phenyl group to twist out of conjugation. To
evaluate a structure with limited delocalization of spin and
con
zeolite, we examined its spin density distribution and hyperfine
coupling pattern to judge whether they are compatible with the
ESR results. Both the UMP2 and UB3LYP methods essentially
localized spin and charge in different sections of the species:
the positive charge is largely located on the benzyl function
whereas the unpaired electron spin resides mainly on C-4 [F4
charge into the phenyl group, we constructed structure 10a•+
ort
by rotating the phenyl group 90° to position it into the plane
bisecting the cyclohexandiyl ring system through C-1 and C-4.

Incorporation of 2-Arylhexa-1,5-diene into Pentasil Zeolite
J. Phys. Chem. B, Vol. 109, No. 7, 2005 2509
Figure 4. (top) The dimensions of the bisected 1-phenylcyclohexane-
Figure 3. (top) The dimensions of the conjugated 1-phenylcyclohex-
ane-1,4-diyl radical cation, 10a•+con. (bottom) Docking of 10a•+con inside
pentasil zeolite viewed along the cylindrical channel axis (right) and
perpendicular to it (left).
1,4-diyl radical cation, 10a•+ort. (bottom) Docking of 10a•+ inside
ort
pentasil zeolite viewed along the cylindrical channel axis (right) and
perpendicular to it (left).
The relative enthalpy and the relative free energy of 10a•+ort at
298.14 K lie 15.4 and 17.9 kcal/mol, respectively, above those
of 10a•+con. It is hardly surprising that a vibrational analysis
(UB3LYP/6-31G*) shows one imaginary frequency for the
structure with enforced Cs geometry: 10a•+ort is not an energy
minimum, at least not in the vacuum calculations. An animation
of the imaginary frequency shows twisting of the orthogonal
phenyl and cyclohexane-1,4-diyl units in the direction toward
the conjugated structure, 10a•+con. The dimensions of 10a•+
ort
also are compatible with the dimensions of the ZSM-5 chan-
nels. Of the dimensions of 10a•+ort, l ) H4-Hp ) 8.76 Å, d )
H3eq-H5eq ) 4.30 Å, and h ) Ho-Ho′ ) Hm-Hm′ ) 4.32 Å,
only the “height” is noticeably increased without, however,
affecting its compatibility (Figure 4).
Figure 5. Dimensions of the pseudo-twist boat form of 1-phenylcy-
clohexane-1,4-diyl radical cation, 10a•+
.
ptb
local minimum; its relative enthalpy and its relative free energy
at 298.14 K lie 12.0 and 13.8 kcal/mol, respectively, above those
of 10a•+con. However, its unpaired electron spin density once
again is localized either on C-1 (and the phenyl group) with
the UMP2 method (F1 ) 0.79, F4 ) 0.01), or on C-4 with the
UB3LYP method (F1 ) 0.11, F4 ) 0.90, Table 2). The resulting
hyperfine coupling patterns (Table 3) are incompatible with the
experimental data.
The unpaired electron spin density for 10a•+ort is more evenly
delocalized than for 10a•+con; however, UMP2 and UB3LYP
methods show divergent trends. With the UMP2 method the
highest spin density is found on C-1 (F1 ) 0.75, F4 ) 0.40,
Table 2), whereas it resides on C-4 with the UB3LYP method
(F1 ) 0.44, F4 ) 0.66). The calculated hyperfine splittings
cover a smaller range than for the conjugated structure [a4 )
-7.6 G, a2,6ax ) 17.7 G, a3,5ax ) 5.7 G (UMP2); a4 ) -14.3
G, a2,6ax ) 9.3 G, a3,5ax ) 17.4 G (UB3LYP), Table 3]. Still,
large differences remain between the splittings calculated for
the three types of nuclei. Also, the largest hfccs calculated with
either UMP2 (17.7 G) or UB3LYP (17.4 G, -14.3 G) are
incompatible with the experimental spectra. Accordingly,
structure 10a•+ort is not a good representation of the species in
the zeolite.
A comparison between the hyperfine coupling patterns of
10a•+ and 10a•+con, in particular the significant changes for
ort
H-4 and H-2,6ax with the UMP2 method (a4 changes from
-29.92 to -7.63 G, a2,6ax from 2.79 to 9.32 G) suggested that
there might be a structure at an intermediate angle of twist, for
which the hfcs would fit the experimental data. Accordingly,
we examined additional species with twist angles between 0
and 90°; none of these structures are minima. The species with
a twist angle of 60°, 10a•+60°, has hfccs (a4 ) -18.92 G, UMP2;
) -16.11 G, UB3LYP; a2,6ax ) 10.73 G, UMP2; ) 8.00 G,
UB3LYP) that show the closest agreement between |a4| and
|a2,6ax| of any calculated form. Alas, even in this case the
discrepancy is significant.
The obvious discrepancy between the hyperfine coupling
patterns of 10a•+ and 10a•+con, respectively, and the experi-
ort
mental results caused us to evaluate additional conformers of
10a•+, a pseudo twist boat species, 10a•+ptb, as well as structures
in which the phenyl group is rotated different degrees from the
plane bisecting the cyclohexan-1,4-diyl ring system through C-1
and C-4. The pseudo twist boat form, 10a•+ptb, (Figure 5) is a
An alternative explanation for the ESR data involves a
distortion of the electronic levels of 10a•+ by the strong

2510 J. Phys. Chem. B, Vol. 109, No. 7, 2005
Ikeda et al.
electrostatic fields inside zeolite voids.18 These forces have been
invoked to explain the remarkable stability of otherwise elusive
positively charged intermediates in zeolites and they can affect
the ordering of close-lying electronic states. For example, the
ESR spectra of cis- and trans-decalin (bicyclo[4.4.0]decane)
radical cations, cis-, trans-15•+, support the stabilization of two
different “electronic states” (structure types), depending on the
nature of the zeolite host and the temperature.19 Spectra
IBM P-series 690. The authors are indebted to Pedro Enrique
Atienzar Corvillo for the molecular modeling of the docking
pictures.
Supporting Information Available: Synthesis of donor
molecules. This material is available free of charge via the
Internet at http://pubs.acs.org.
2
corresponding to the A1 state of cis-15•+ were obtained in
References and Notes
silicalite (a ) 49.5 G, 4 H, 45K) or ZSM-34 (a ) 50 G, 4 H),
whereas spectra supporting the 2A2 state, predicted to be higher
in energy by (vacuum) calculations, were observed in silicalite
(a ) 28.1 G, 4 H, 95 K) or offretite (a ) 30.2 G, 4 H). For
trans-15•+, structures corresponding to the (lower-energy) 2Ag
state were observed in silicalite (a4H ) 50.5 G) or ZSM 34 (a
) 51.5 G, 4 H), whereas the higher-energy 2Bg state was
supported in Na-Y (a ) 28.5 G, 4 H) or Na-W-5 (a ) 29.8
G, 4 H).
(1) (a) Roth, H. D.; Schilling, M. L. M. J. Am. Chem. Soc. 1985, 107,
716-718. (b) Roth, H. D.; Schilling, M. L. M.; Abelt, C. J. Tetrahedron
1986, 42, 6157-6166. (c) Roth, H. D.; Schilling, M. L. M.; Abelt, C. J. J.
Am. Chem. Soc. 1986, 108, 6098-6099. (d) Momose, T.; Shida, T.;
Kobayashi, T. Tetrahedron 1986, 42, 6337-6342.
(2) (a) Roth, H. D. Proc. IUPAC Symp. Photochem. 1984, 10, 455-
456. (b) Roth, H. D.; Abelt, C. J., J. Am. Chem. Soc. 1986, 108, 2013-
2019. (c) Dai, S.; Wang, J. T.; Williams, F. J. Am. Chem. Soc. 1990, 112,
2835-2836. (d) Dai, S.; Wang, J. T.; Williams, F. J. Am. Chem. Soc. 1990,
112, 2837-2837.
(3) (a) Guo, Q.-X.; Qin, X.-Z.; Wang, J. T.; Williams, F. J. Am. Chem.
Soc. 1988, 110, 1974-1976. (b) Miyashi, T.; Konno, A.; Takahashi, Y. J.
Am. Chem. Soc. 1988, 110, 3676-3677. (c) Ikeda, H.; Minegishi, T.; Abe,
H.; Konno, A.; Goodman, J. L.; Miyashi, T. J. Am. Chem. Soc. 1998, 120,
87-95. (d) Ikeda, H.; Takasaki, T.; Takahashi, Y.; Konno, A.; Matsumoto,
M.; Hoshi, Y.; Aoki, T.; Suzuki, T.; Goodman, J. L.; Miyashi, T. J. Org.
Chem. 1999, 64, 1640-1649. (e) Miyashi, T.; Ikeda, H.; Takahashi, Y.
Acc. Chem. Res. 1999, 32, 815-824.
(4) Doering, W. v. E.; Roth, W. R. Tetrahedron 1962, 18, 67-74.
(5) (a) Ramamurthy, V.; Casper, C. V.; Corbin, D. R. J. Am. Chem.
Soc. 1991, 113, 594-600. (b) Pollack, S. S.; Sprecher, R. F.; Frommell, E.
A. J. Mol. Catal. 1991, 66, 195-203. (c) Rhodes, C. J. J. Chem. Soc.,
Faraday Trans. 1991, 87, 3179-3184. (d) Rhodes, C. J.; Standing, M. J.
Chem. Soc., Perkin Trans. 2 1992, 1455-1460. (e) Chen, F. R.; Fripiat, J.
J. J. Phys. Chem. 1992, 96, 819-823, 1993, 97, 5796-5797. (f) Crockett,
R.; Roduner, E. J. Chem. Soc., Perkin Trans. 2 1993, 1503-1509. (g)
Roduner, E.; Crockett, R.; Wu, L. M. J. Chem. Soc., Faraday Trans. 1993,
89, 2101-2105. (h) Baciocchi, E.; Doddi, G.; Ioele, M.; Ercolani, G.
Tetrahedron 1993, 49, 3793-3800.
Similarly, the ESR spectrum of p-methylphenoxyl radical,
16•, in ZSM-5 showed a well-resolved 1:3:3:1 quartet (g )
2.0042 ( 0.0001; a ) 14.6 G, 3 H; line width ) 2.1 G),9b
analogous to the known solution spectrum of 16• (aâ ) 11.95
G, 3H; a2,6 ) 6.0 G, 2 H).20 However, the species sequestered
in the zeolite failed to reveal any evidence for coupling of the
o-protons, which is significant in solution (a2,6 ) 6.0 G, 2 H).19
In fact, the EPR spectrum of 16•-aryl-d4 showed a spectrum
(aCH3 ) 14.4 G; line width ) 1.3 G) very similar to that of
16•-aryl-h4 in the zeolite. These results were also ascribed to a
distortion of the “conventional” phenoxyl radical, 16•, by a
specific interaction with the zeolite.
In light of these ESR results in zeolites we ascribe the results
for 10•+ in ZSM-5 or H-mor, particularly the apparent limited
delocalization of spin and charge into the phenyl group, to a
distortion of the electronic structure of the sequestered guest
by the strong electrostatic fields inside zeolite voids, an
additional example of such a phenomenon. The unusual and
unexpected results obtained in this study encourage further
research into generating radical cations of unusual structure types
in zeolites.
(6) Garc´ıa, H.; Roth, H. D. Chem. ReV. 2002, 102, 3947-4008.
(7) (a) Corma, A.; Forne´s, V.; Garc´ıa, H.; Mart´ı, V.; Miranda, M. A.
Chem. Mater. 1995, 7, 2136-2143. (b) Folgado, J.-V.; Garc´ıa, H.; Mart´ı,
V.; Espla´, M. Tetrahedron 1997, 53, 4947-4956.
(8) (a) Lakkaraju, P. S.; Zhou, D.; Roth, H. D. Chem. Commun. 1996,
2605-2606. (b) Roth, H. D.; Shen, K.; Lakkaraju, P. S.; Ferna´ndez, L.
Chem. Commun. 1998, 2447-2448. (c) Herbertz, T.; Lakkaraju, P. S.; Roth,
H. D. J. Phys. Chem. A 1999, 103, 11350-11354. (d) Herbertz, T.; Roth,
H. D.; Blume, M.; Blume, F.; Lakkaraju, P. S. Eur. J. Org. Chem. 2000,
467-472.
(9) (a) Lakkaraju, P. S.; Zhang, J.; Roth, H. D. J. Phys. Chem. 1994,
98, 2722-2725. (b) Roth, H. D.; Weng, H.; Zhou, D.; Lakkaraju, P. S.
Acta Chem. Scand. 1997, 51, 626-635.
Conclusion
(10) Shiralkar, V. P.; Clearfield, A. Zeolites 1989, 9, 363-370.
(11) Frisch, M. J.; Trucks, G. W.; Schlegel, H. B.; Scuseria, G. E.; Robb,
M. A.; Cheeseman, J. R.; J. A. Montgomery, J.; Vreven, T.; Kudin, K. N.;
Burant, J. C.; Millam, J. M.; Iyengar, S. S.; Tomasi, J.; Barone, V.;
Mennucci, B.; Cossi, M.; Scalmani, G.; Rega, N.; Petersson, G. A.;
Nakatsuji, H.; Hada, M.; Ehara, M.; Toyota, K.; Fukuda, R.; Hasegawa, J.;
Ishida, M.; Nakajima, T.; Honda, Y.; Kitao, O.; Nakai, H.; Klene, M.; Li,
X.; Knox, J. E.; Hratchian, H. P.; Cross, J. B.; Adamo, C.; Jaramillo, J.;
Gomperts, R.; Stratmann, R. E.; Yazyev, O.; Austin, A. J.; Cammi, R.;
Pomelli, C.; Ochterski, J. W.; Ayala, P. Y.; Morokuma, K.; Voth, G. A.;
Salvador, P.; Dannenberg, J. J.; Zakrzewski, V. G.; Dapprich, S.; Daniels,
A. D.; Strain, M. C.; Farkas, O.; Malick, D. K.; Rabuck, A. D.;
Raghavachari, K.; Foresman, J. B.; Ortiz, J. V.; Cui, Q.; Baboul, A. G.;
Clifford, S.; Cioslowski, J.; Stefanov, B. B.; Liu, G.; Liashenko, A.; Piskorz,
P.; Komaromi, I.; Martin, R. L.; Fox, D. J.; Keith, T.; Al-Laham, M. A.;
Peng, C. Y.; Nanayakkara, A.; Challacombe, M.; Gill, P. M. W.; Johnson,
B.; Chen, W.; Wong, M. W.; Gonzalez, C.; Pople, J. A. Gaussian 03,
revision B.02; Gaussian Inc.: Pittsburgh, PA, 2003.
(12) Becke’s three-parameter exchange functional with the nonlocal
gradient correlation functional of Lee-Yang-Parr was used; cf. (a) Becke,
A. D. J. Chem. Phys. 1993, 98, 5648-5652. (b) Lee, C.; Yang, W.; Parr,
R. G. Phys. ReV. B 1988, 37, 785-789.
(13) Møller, C.; Plesset, M. S. Phys. ReV. B 1934, 46, 618-622.
(14) Glendening, E. D.; Reed, A. E.; Carpenter, J. E.; Weinhold, F. NBO,
version 3.1.
(15) (a) Raghavachari, K.; Haddon, R. C.; Roth, H. D. J. Am. Chem.
Soc. 1983, 105, 3110-3114. (b) Raghavachari, K.; Roth, H. D. J. Am.
Chem. Soc. 1989, 111, 7132-7136. (c) Roth, H. D.; Schilling, M. L.
M.; Raghavachari, K. J. Am. Chem. Soc. 1984, 106, 253-255.
Upon incorporation into a redox-active pentasil zeolite
[(Na,H)-ZSM-5], 2-arylhexa-1,5-dienes (9; aryl ) anisyl, tolyl,
phenyl) are converted into 1-arylcyclohexane-1,4-diyl radi-
cal cations, 10•+. The ESR spectrum of 10•+ (sextet, g )
2.0026; a ) 9.0 G) indicated the presence of five essentially
equivalent nuclei, indicating limited delocalization of spin and
charge into the phenyl group. The limited delocalization is
ascribed to the strong electrostatic fields inside zeolites. Species
10•+ is stable at room temperature, in striking contrast to the
parent radical cation in cryogenic matrices: cyclohexane-1,4-
diyl, 3•+ is converted to cyclohexene radical cation above 90
K. The structures of radical cation, 10a•+, and of the unsubsti-
tuted parent were probed by density functional theory calcula-
tions.
Acknowledgment. H.I. thanks the Ministry of Education,
Culture, Sports, Science, and Technology of Japan for financial
support for a Grant-in-Aid (No. 1405008) for Scientific Research
in Priority Areas (Area No. 417); H.D.R. acknowledges financial
support from the National Science Foundation (Grant NSF
CHE-9714850). R.R.S. acknowledges the National Computer
Alliance (grant CHE 030060) for support of his work on the

Incorporation of 2-Arylhexa-1,5-diene into Pentasil Zeolite
J. Phys. Chem. B, Vol. 109, No. 7, 2005 2511
(d) Krogh-Jespersen, K.; Roth, H. D. J. Am. Chem. Soc. 1992, 114, 8388-
8394.
(18) (a) Rabo, J. A.; Angell, C. L.; Kasai, P. H.; Schomaker, V. Discuss.
Faraday Soc. 1966, 41, 328-349. (b) Dempsey, E. J. Phys. Chem. 1969,
73, 3660-3668. (c) Jousse, F.; Cohen de Lara, E. J. Phys. Chem. 1996,
100, 233-237.
(16) (a) Eriksson, L. A.; Malkin, V. G.; Malkina, O. L.; Salahub, D. R.
J. Chem. Phys. 1993, 99, 9756-9763. (b) Eriksson, L. A.; Malkin, V. G.;
Malkina, O. L.; Salahub, D. R. Int. J. Quantum Chem. 1994, 52, 879-901.
(17) (a) Batra, R.; Giese, B.; Spichty, M.; Gescheidt, G.; Houk, K. N.
J. Phys. Chem. 1996, 100, 18371-18379. (b) Roth, H. D.; Herbertz, T.
1996, unpublished results.
(19) Barnabas, M. V.; Trifunac, A. D. Chem. Phys. Lett. 1991, 187,
565-570.
(20) (a) Stone, T. J.; Waters, W. A. Proc. Chem. Soc. 1962, 253-254.
(b) Stone, T. J.; Waters, W. A. J. Chem. Soc. 1964, 213-218.