Divergent oxidative rearrangements in solution and in a zeolite: Distal vs proximal bond cleavage of methylenecyclopropanes

Article abstract of DOI:10.1021/ja045930n

Irradiation of 9,10-dicyanoanthracene (DCA) or p-chloranil in the presence of E-1-benzylidene-2-phenylcyclopropane (E-5) in CH₂CI₂ causes E-5 to undergo methylenecyclopropane rearrangement. An adduct, Z-7, between DCA and 5 firmly supports the involvement of a bifunctional trimethylenemethane radical cation. In contrast, incorporation of E-5 into HZSM-5 produces trans,trans-1,4-diphenyl-1,3-butadiene radical cation sequestered in the HZSM-5 interior, tt-8^·+ @ HZSM-5, identified by ESR and diffuse reflectance spectroscopy. In addition, low yields of tt-8, its cis, trans-isomer (ct-8), and 1-phenyl-1,2-dihydronaphthalene (9) were isolated from the supernatant solution. The sharp contrast between the photoinduced electrontransfer reaction with photosensitizers in solution and the spontaneous reaction with redox-active acidic zeolite offers the prospect of further zeolite-induced regiodivergent reactions in a range of additional substrates.

Full text of DOI:10.1021/ja045930n

Published on Web 09/17/2005
Divergent Oxidative Rearrangements in Solution and in a
Zeolite: Distal vs Proximal Bond Cleavage of
Methylenecyclopropanes
Hiroshi Ikeda,*,† Tsuyoshi Nomura,† Kimio Akiyama,§ Mitsuhiro Oshima,†
Heinz D. Roth,*,‡ Shozo Tero-Kubota,§ and Tsutomu Miyashi†
Contribution from the Department of Chemistry, Graduate School of Science, Tohoku UniVersity,
Sendai 980-8578, Japan, Institute of Multidisciplinary Research for AdVanced Materials, Tohoku
UniVersity, Sendai 980-8577, Japan, and Department of Chemistry and Chemical Biology,
Rutgers UniVersity, Wright-Rieman Laboratories, New Brunswick, New Jersey 08854-8087
Received July 8, 2004; E-mail: ikeda@org.chem.tohoku.ac.jp; roth@rutchem.rutgers.edu
Abstract: Irradiation of 9,10-dicyanoanthracene (DCA) or p-chloranil in the presence of E-1-benzylidene-
2-phenylcyclopropane (E-5) in CH2Cl2 causes E-5 to undergo methylenecyclopropane rearrangement. An
adduct, Z-7, between DCA and 5 firmly supports the involvement of a bifunctional trimethylenemethane
radical cation. In contrast, incorporation of E-5 into HZSM-5 produces trans,trans-1,4-diphenyl-1,3-butadiene
radical cation sequestered in the HZSM-5 interior, tt-8•+@HZSM-5, identified by ESR and diffuse reflectance
spectroscopy. In addition, low yields of tt-8, its cis,trans-isomer (ct-8), and 1-phenyl-1,2-dihydronaphthalene
(9) were isolated from the supernatant solution. The sharp contrast between the photoinduced electron-
transfer reaction with photosensitizers in solution and the spontaneous reaction with redox-active acidic
zeolite offers the prospect of further zeolite-induced regiodivergent reactions in a range of additional
substrates.
Scheme 1 a
Introduction
The thermal rearrangement of methylenecyclopropane (MCP)
systems and the underlying mechanism have been investigated
in detail,1 as has the nature of the related trimethylenemethane
(TMM), one of the most thoroughly investigated diradical
systems.2 Subsequently, the developing interest in electron
transfer and radical ion chemistry caused the structures and
reactivities of suitable MCP radical cations to be scrutinized.3
For example, electron transfer (ET) from 1,1-dianisyl-2-meth-
ylenecyclopropane (1, Scheme 1) to triplet p-chloranil (CA)
generates a bifunctional TMM radical cation, 2•+; CIDNP effects
observed during this reaction showed that the unpaired electron
spin density and the charge of 2•+ are residing in separate
a Ar; anisyl.
π-systems, the spin in the allyl group, and the charge in the
diphenylmethylene function.3b The limited delocalization of spin
and charge is due to the attachment of the diphenylmethylene
moiety in the node of the allyl function; the two π-systems need
not be orthogonal, as initially suggested.3a,b The radical cation,
2•+, forms two spiro-annelated CA-adducts3b via zwitterions,
3+- and 4+-,4 formed by coupling of singlet ion pairs, and
regenerates the starting material, 1. Recent spectroscopic studies,
using time-resolved absorption and ESR spectroscopy and
photoacoustic calorimetry following nanosecond laser flash
photolysis support the intermediacy of a triplet biradical, 2••,
formed by triplet return (R) ET.3c,d In summary, the photo-
induced (P) ET MCP rearrangement of 1 proceeds in three
stages: radical cation cleavage (RCCl), triplet RET, and
* To whom correspondence should be addressed.
† Graduate School of Science, Tohoku University.
‡ Institute of Multidisciplinary Research for Advanced Materials, Tohoku
University.
§ Rutgers University.
(1) (a) Ullman, E. F. J. Am. Chem. Soc. 1960, 82, 505-506. (b) Chesick, J. P.
J. Am. Chem. Soc. 1963, 85, 2720-2723. (c) Doering, W. v. E.; Roth, H.
D. Tetrahedron 1970, 26, 2825-2835. (d) Gajewski, J. J. J. Am. Chem.
Soc. 1971, 93, 4450-4458.
(2) (a) Dowd, P. J. Am. Chem. Soc. 1966, 88, 2587-2589. (b) Dowd, P.; Chow,
M. J. Am. Chem. Soc. 1977, 99, 6438-6440. (c) Dowd, P. Acc. Chem.
Res. 1972, 5, 242-248. (d) Berson, J. A.; Bushby, R. J.; McBride, J. M.;
Tremelling, M. J. Am. Chem. Soc. 1971, 93, 1544-1546.
(3) (a) Takahashi, Y.; Miyashi, T.; Mukai, T. J. Am. Chem. Soc. 1983, 105,
6511-6513. (b) Miyashi, T.; Takahashi, Y.; Mukai, T.; Roth, H. D.;
Schilling, M. L. M. J. Am. Chem. Soc. 1985, 107, 1079-1080. (c) Ikeda,
H.; Nakamura, T.; Miyashi, T.; Goodman, J. L.; Akiyama, K.; Tero-Kubota,
S.; Houman, A.; Wayner, D. D. M. J. Am. Chem. Soc. 1998, 120, 5832-
5833. (d) Ikeda, H.; Akiyama, K.; Takahashi, Y.; Nakamura, T.; Ishizaki,
S.; Shiratori, Y.; Ohaku, H.; Goodman, J. L.; Houmam, A.; Wayner, D. D.
M.; Tero-Kubota, S.; Miyashi, T. J. Am. Chem. Soc. 2003, 125, 9147-
9157.
(4) (a) Roth, H. D. J. Photochem. Photobiol., C 2001, 2, 93-116. (b) Roth,
H. D. Pure Appl. Chem. 2005, 77, 1075-1085.
9
10.1021/ja045930n CCC: $30.25 © 2005 American Chemical Society
J. AM. CHEM. SOC. 2005, 127, 14497-14504
14497

A R T I C L E S
Ikeda et al.
Scheme 2
diradical cyclization (DRCy; Scheme 1); all steps have been
firmly established.
The catalytic properties of zeolites and their redox and acid-
base reactivities have been investigated extensively, and a wide
range of chemical reactions has been studied on the surface and
in the micropores of these materials.5 Interestingly, the nearly
cylindrical channels [internal diameter (ID) ∼5.5 Å] of pentasil
zeolite (HZSM-5) can generate π-type radical cations from
suitable substrates at ambient temperatures.6 The supramolecular
interaction with the rigid anisotropic microporous framework
stabilizes the otherwise highly reactive radical cations due to
the combined effects of (a) the intense electrostatic fields inside
zeolite pores, (b) topological restrictions that prevent the access
of external reagents, and (c) the limiting dimensions of the
zeolite channels, restricting the shape of the enclosed species.7-9
Because of dramatically increased lifetimes (days to years at
room temperature), the sequestered radical cations can be studied
by conventional spectroscopic techniques.6,10
Table 1. Eoxap, ∆Get of ET to 1DCA*, and kq for DCA
Fluorescence Quenching in CH2Cl2 for Selected Donors
The dual reactivity of HZSM-5, i.e., its ability to function as
an acid-base catalyst as well as a redox agent, offers the
intriguing prospect that incorporation of an MCP system may
lead to a selective bond cleavage differing from that occurring
in homogeneous, isotropic media. Such regiodivergent reactions
might be of significant value. On the other hand, the dual
reactivity of HZSM-5 complicates the potential mechanisms of
zeolite-induced reactions. Accordingly, few systems have been
found that show contrasting ET reactivity in solution vs in a
zeolite.8,11 To examine a potentially medium-dependent regio-
selective bond cleavage, we studied the reactions of E-1-
benzylidene-2-phenylcyclopropane (E-5, smallest molecular
diameter ) 5.5 Å) with ET photosensitizers, 9,10-dicyano-
anthracene (DCA) or CA, and its spontaneous reaction upon
incorporation into the redox-active acidic zeolite, HZSM-5. Our
results illustrate divergent reactions for the DCA-sensitized and
the HZSM-5-induced reactions of E-5; the homogeneous
solution was found to promote distal bond cleavage and an MCP
rearrangement, whereas the heterogeneous medium causes
proximal bond cleavage (Scheme 2).
a
b
Eox
∆
eV
Get
kq
ap
1
substrate
V
1010 M-1 s-
E-5
Z-5
trans-6
tt-8
+1.62
+1.71
+1.51
+1.37
-0.62
-0.53
-0.73
c
1.0
0.78
0.76
c
a In CH2Cl2 containing 0.1 M n-Bu4NClO4. b ∆Get ) Eox1/2 - Ered(DCA)
- 1E*(DCA) - e2/ꢀr.13 Eox1/2 ) Eoxap - 0.03 V; Ered1/2(DCA) ) -0.89 V
vs SCE; 1E*(DCA) ) 2.87 eV in CH2Cl2. The coulomb term (e2/ꢀr) is
0.23 eV in CH2Cl2. c Not relevant to this study.
according to literature procedures.12 See the Supporting Information
for physical data of E-5, Z-5, and related compounds. The anodic
oxidation potential of E-5 (Eox ) +1.62 V vs SCE) is sufficiently
ap
low to quench excited singlet DCA at nearly the diffusion-controlled
rate in CH2Cl2 (kq ) 1.0 × 1010 M-1 s-1). Oxidation potentials (Eoxap),
free energy changes (∆Get), and kq for ET from E-5 and related donors
to singlet-excited DCA are summarized in Table 1.
General Procedure for the DCA-Sensitized Photoreactions.
Solutions containing 0.05 mmol each of a substrate and DCA in 0.5
mL of CD2Cl2 in Pyrex tubes (0.5 cm ID) were degassed by five
freeze-pump-thaw cycles (77 K; 10-2 Torr; ambient temperature) and
sealed at 10-2 Torr. The samples, containing a slight suspension of
DCA, were irradiated with the output of a 2 kW Xe lamp through a
cutoff filter (λ > 360 nm) at 20 ( 1 °C. Product yields were determined
by 1H NMR analyses. Details of product analysis and physical data of
the photoproducts are shown in the Supporting Information.
Experimental Section
Preparation of Organic Materials. Substrate E-5 was obtained in
50% yield together with its geometric isomer, Z-5 (16%), by oxidation
of dibenzylidenemethane dianion, generated from 1,1-dibenzylethylene,
Preparation of the Zeolite. Pentasil zeolites (HZSM-5; Si/Al ratio
20, 34) were obtained by calcination (500 °C, 10 h) of corresponding
samples of NH4ZSM-5 (TOSOH). The electron-accepting ability of
these HZSM-5 samples and of NaZSM-5 (TOSOH, Si/Al ) 11.8,
similarly calcined) was examined by mixing with 2,2,4-trimethylpentane
(TMP) solutions of π-type electron donors; these were inspected visually
for coloration and examined by ESR spectroscopy.6a,14 The HZSM-5
sample with Si/Al ) 20 was the most powerful oxidizer with an
estimated reduction potential (Eredap) of ∼+1.65 V vs SCE; this sample
was used to probe the reactions of E-5.
General Adsorption Procedure onto the Zeolite. The substrates
(1.05 mg in 10 mL anhydrous TMP) were adsorbed onto pentasil zeolite
by stirring with 300 mg HZSM-5 for 2 h at room temperature. The
loaded zeolite was collected by filtration, washed with dry TMP to
remove material adsorbed on the surface, and dried at reduced pressure
(5) (a) Rabo, J. A., Ed. Zeolite Chemistry and Catalysis; ACS Monograph:
1976; Vol. 171. (b) Stucky, G. D., Dwyer, F. D., Eds. Intrazeolite
Chemistry; ACS Monograph: 1983; Vol. 218. (c) Derouane, E. G., Lemos,
F., Naccache, C., Rebeiro, F. R., Eds. Zeolite Microporous Solids:
Synthesis, Structure, and ReactiVity; Kluwer Academics: Dordrecht,
1991.
(6) (a) Ramamurthy, V.; Casper, J. V.; Corbin, D. R. J. Am. Chem. Soc. 1991,
113, 594-600. (b) Casper, J. V.; Ramamurthy, V.; Corbin, D. R. J. Am.
Chem. Soc. 1991, 113, 600-610. (c) Liu, X.; Iu, K.-K.; Thomas, J. K.;
He, H.; Klinowski, J. J. Am. Chem. Soc. 1994, 116, 11811-11818. (d)
Garc´ıa, H.; Roth, H. D. Chem. ReV. 2002, 102, 3947-4007.
(7) (a) Lakkaraju, P. S.; Zhou, D.; Roth, H. D. J. Chem. Soc., Perkin Trans.
2 1998, 1119-1121. (b) Mart´ı, V.; Ferna´ndez, L.; Forne´s, V.; Garc´ıa, H.;
Roth H. D. J. Chem. Soc., Perkin Trans. 2 1999, 145-152.
(8) Herbertz, T.; Lakkaraju, P. S.; Blume, F.; Blume, M.; Roth, H. D. Eur. J.
Org. Chem. 2000, 467-472.
(9) (a) Hirano, T.; Li, W.; Abrams, L.; Krusic, P. J.; Ottaviani, M. F.; Turro,
N. J. J. Org. Chem. 2000, 65, 1319-1330. (b) Turro, N. J. Acc. Chem.
Res. 2000, 33, 637-646.
(10) Vaidyalingam, A. S.; Coutant, M. A.; Dutta, P. K. In Electron Transfer in
Chemistry; Balzani, V., Ed.; Wiley: Weinheim, 2001; Vol. 4, pp 412-486.
(11) For example, see: (a) Kojima, M.; Takeya, H.; Kuriyama, Y.; Oishi, S.
Chem. Lett. 1997, 997-998. (b) Roth, H. D.; Herbertz, T.; Lakkaraju, P.
S.; Sluggett, G.; Turro, N. J. J. Phys. Chem. A 1999, 103, 11350-11354.
(c) Ikeda, H.; Minegishi, T.; Miyashi, T.; Lakkaraju, P. S.; Sauers, R. R.;
Roth, H. D. J. Phys. Chem. B 2005, 109, 2504-2511.
(12) Witt, O.; Mauser, H.; Friedl, T.; Wilhelm, D.; Clark, T. J. Org. Chem.
1998, 63, 959-967.
(13) Rehm, D.; Weller, A. Isr. J. Chem. 1970, 8, 259-271.
(14) See the Supporting Information for details.
9
14498 J. AM. CHEM. SOC. VOL. 127, NO. 41, 2005

Methylenecyclopropane, Distal vs Proximal Bond Cleavage
A R T I C L E S
(e1 Torr). For product analysis, 30 mg of E-5 (0.15 mmol) and 8.0 g
of HZSM-5 were stirred at room temperature for 5 h. After filtration
on a sintered glass disk, the loaded, colored zeolite was washed with
TMP. After removing the solvent from the filtrate the products were
separated by preparative HPLC; structures and yields were deter-
1
mined by 200 MHz H NMR analysis. Physical data of the products
obtained in the zeolite-catalyzed reactions are shown in the Supporting
Information.
γ-Irradiation of Glassy Matrices. n-BuCl solutions (1 mL) of the
substrates (10 mM) in a flat quartz cell (2 × 10 × 40 mm) were
degassed by five freeze-pump-thaw cycles (77 K; 10-2 Torr; ambient
temperature) and sealed at 10-2 Torr. A glassy matrix was obtained by
steeping the cell into liquid nitrogen. This cell was irradiated with a
5.1 TBq 60Co source in liquid nitrogen at 77 K for 40 h. The absorption
spectra before and after irradiation were recorded at 77 K.
ESR Studies. 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 were accumulations of eight scans. Details of the CIDEP
experiments have been described in the Supporting Information of refs
3d and 15.
Quantum Chemical Calculations. Density functional theory (DFT)
and TD-DFT calculations for key radical cations were carried out with
the Gaussian 98 suite of electronic structure programs using extended
basis sets, including d-type polarization functions on carbon.16 The
geometries of the radical cations were optimized at the unrestricted
B3LYP17 level with the standard 6-31G** basis set.
Figure 1. Product distribution obtained upon irradiation of DCA in
degassed CH2Cl2 in the presence of E-5 (top: black circles) or Z-5 (bottom;
green filled circles) as a function of time: trans-6 (blue triangles), cis-6
(light blue triangles), and Z-7 (red filled diamonds).
Results
Photoinduced Reaction of E-5. Irradiation (2 kW Xe lamp,
λ > 360 nm) of DCA (1E* ) 2.87 eV; Ered ) -0.89 V vs
1/2
2.0058), assigned to the tetrachlorobenzosemiquinone radical
anion (CA•-)20 and a weak, complex emission signal at H )
3350-3400 G (g ∼2.002) due to (a) species arising from E-5
(Figure 2, bottom).
Zeolite-Induced Reactions. The interaction of E-5 with
HZSM-5 produced significantly different results. Upon stirring
a suspension of HZSM-5 (Si/Al ) 20, 300 mg) in TMP
containing E-5 (5 µmol) for 8 h, three products were isolated
from the supernatant liquid: trans,trans- and cis,trans-1,4-
diphenyl-1,3-butadiene (tt-8 and ct-8) and 1-phenyl-1,2-dihy-
dronaphthalene (9, Scheme 2) in 10, 3, and 10% yield,
respectively. In addition, 65% of E-5 was incorporated into the
zeolite, giving it a pink hue.
An ESR spectrum (Figure 3, top) of the dried pink zeolite
was persistent at room temperature; this spectrum was identified
as that of tt-8•+@HZSM-5, because it is essentially identical
with one obtained from authentic tt-8 upon incorporation into
HZSM-5.6a,14 Furthermore, the diffuse reflectance (DR) spec-
trum of the pink zeolite sample obtained upon incorporation of
E-5 was identical with the DR spectrum observed upon
incorporation of tt-8 (Figure 4, top), further supporting the
identification of the sequestered ion and confirming the HZSM-
5-catalyzed isomerization of E-5 to tt-8•+@HZSM-5.
SCE in CH2Cl2) in degassed CH2Cl2 in the presence of E-5
(Eox ) +1.62 V vs SCE) for 15 min led to the formation of
ap
the geometric isomer Z-5, as well as two structural isomers,
cis-6 and trans-6, in 9, 13, and 33% yield, respectively, after 1
h. In addition, a DCA-adduct, Z-7, was formed (Scheme 2;
Figure 1, top). The figure clearly shows three phases in the
photoreaction. In the earliest phase, E-5 is rapidly depleted and
trans-6 is built up, along with cis-6 and Z-5; trans-6 reaches a
maximum at ∼15 min. In the second phase, trans-6 is depleted
in favor of cis-6 and Z-5; these products show near-plateaus
between 30 and 90 min. Finally, continued irradiation causes
the three isomers, Z-5, cis-6, and trans-6 to be depleted in favor
of the adduct Z-7, which became the predominant product after
4 h irradiation. Similar results were observed during the reaction
of DCA with E-5 in DMSO-d6.
The photoinduced reaction of the geometric isomer, Z-5, with
DCA under otherwise identical conditions also gave rise to cis-
and trans-6 as well as the adduct Z-7. The significant difference
to the reaction of E-5 lies in the observation that the rise of
adduct Z-7 does not show an “induction period” but occurs
without delay (Figure 1, bottom).
To probe the nature of the radical cation intermediate, CA
(3E* ) 2.7 eV; Ered1/2 ) +0.02 V vs SCE) was irradiated (355
nm) in DMSO in the presence of E-5 at ambient temperature
in an ESR cavity. This reaction gave rise to a CIDEP spectrum18
featuring an intense emission signal at H ) 3365 G (g )
(18) For recent CIDEP studies on PET reactions of organic substrates, see ref
3d and 19.
(19) (a) Ishiguro, K.; Khudyakov, I. V.; McGarry, P. F.; Turro, N. J.; Roth, H.
D. J. Am. Chem. Soc. 1994, 116, 6933-6934. (b) Kobori, Y.; Yago, T.;
Akiyama, K.; Tero-Kubota, S. J. Am. Chem. Soc. 2001, 123, 9722-9723.
(c) Akiyama, K.; Tero-Kubota, S. J. Phys. Chem. B 2002, 106, 2398-
2403. (d) Yago, T.; Kobori, Y.; Akiyama, K.; Tero-Kubota, S. J. Phys.
Chem. B 2003, 107, 13255-13257. (e) Kobori, Y.; Yago, T.; Akiyama,
K.; Tero-Kubota, S.; Sato, H.; Hirata, F.; Norris, J. R., Jr. J. Phys. Chem.
B 2004, 108, 10226-10240.
(15) (a) Akiyama, K.; Tero-Kubota, S.; Ikegami, Y.; Ikenoue, T. J. Am. Chem.
Soc. 1984, 106, 8322-8323. (b) Akiyama, K.; Tero-Kubota, S.; Ikoma,
T.; Ikegami, Y. J. Am. Chem. Soc. 1994, 116, 5324-5327.
(16) Frisch, M. J. et al. Gaussian 98, revision A.11.4; Gaussian Inc.; Pittsburgh,
PA, 1998.
(17) (a) Becke, A. D. J. Chem. Phys. 1993, 98, 5648-5652. (b) Lee, C.; Yang,
(20) Segal, B. G.; Kaplan, M.; Fraenkel, G. K. J. Chem. Phys. 1965, 43, 4191-
4200.
W.; Parr, R. G. Phys. ReV. B 1988, 37, 785-789.
9
J. AM. CHEM. SOC. VOL. 127, NO. 41, 2005 14499

A R T I C L E S
Ikeda et al.
Figure 2. CIDEP spectrum observed at a delay time of 500 ns after the
laser excitation of CA in the presence of E-5 in DMSO (bottom); the asterisk
(*) represents emission due to CA•-. A simulated spectrum with hyperfine
coupling constants (aH) of bifunctional radical cation, E-10•+ (see text), is
shown above.
Figure 4. (Top) DR spectra, assigned to tt-8•+, obtained by inclusion of
E-5 (blue) and tt-8 (red) in HZSM-5 at room temperature. (Bottom)
Absorption spectra observed upon γ-irradiation of an n-BuCl glassy matrix
of E-5 (blue) and tt-8 (red) at 77 K; γ-irradiation of E-5 does not generate
tt-8•+
.
Scheme 3 a
Figure 3. (Top) Steady-state CW-ESR spectra of tt-8•+@HZSM-5 (g )
2.0023, line width ) 0.8 G) obtained upon inclusion of E-5 in HZSM-5 at
room temperature and (bottom) of tt-8•+ (g ) 2.0023, line width ) 2.3 G)
obtained upon γ-irradiation in an n-BuCl matrix at 77 K.
a Ar; 3,5-dimethylphenyl.
on HZSM-5, we conducted quantitative adsorption experiments.
Based on the quantities of holes of straight channels (Nh) and
active sites (Ns) of HZSM-5 and the molecular volume of
tt-8•+@HZSM-5 we estimated the occupation of holes and active
sites. For 100 mg of HZSM-5, Nh and Ns were estimated to be
0.082 and 33 µmol, respectively (Table 2). The value of Nh
was determined by an N2 gas absorption experiment; a calcula-
tion based on an SEM image of HZSM-5 gave a similar value
(0.11 µmol for 100 mg of HZSM-5). The Ns values are based
on the experimentally determined micropore volume of the
straight channels (9.2 × 10-8 m3/unit cell), the molecular
volume of one tt-8•+ molecule (4.9 × 10-28 m3), and the
assumption that (a) the rigid tt-8•+ molecules can be adsorbed
in the straight, but not the zigzag channels of HZSM-5, and (b)
one unit cell of HZSM-5 (19.9 Å length along a zigzag channel)
can adsorb two tt-8•+ molecules (14 Å length) because one unit
cell of the HZSM-5 has two straight channels. A similar Ns
value (37 µmol per 100 mg HZSM-5) was calculated from the
absorption volume per gram. Accordingly, each channel of
HZSM-5 has ∼400 sites (∼33 µmol/0.083 µmol) for tt-8•+.
The degree of conversion of E-5 depended strongly on the
“loading” ratio of substrate to zeolite. For example, stirring E-5
(1.03 mg, 5.0 µmol) with HZSM-5 (100 mg) produced a mixture
of E-5, tt-8, ct-8, and 9 in 53, 7, 2, and 5% yield, respectively
ESR and absorption spectra of tt-8•+ were also obtained upon
γ-ray irradiation of tt-8 in a glassy matrix (n-butyl chloride,
n-BuCl) at 77 K. In the n-BuCl matrix the absorption maximum
(548 and 822 nm, Figure 4, bottom) is slightly red-shifted
relative to the DR spectrum in the zeolite (540 and 817 nm).
Similar γ-irradiation of E-5 gave a weak 446-nm band, but not
the 548- and 822-nm bands. The ESR signals of tt-8•+@HZSM-5
were noticeably sharper than those of tt-8•+ in the frozen n-BuCl
matrix (Figure 3, bottom).
The geometric isomer, Z-5, was also allowed to interact with
HZSM-5 under the same conditions as the E-isomer; this
reaction gave rise to tt- and ct-8, 9, in the supernatant liquid
and also resulted in the pink tt-8•+@HZSM-5.
In contrast, a bulkier analogue of E-5, E-1-[(3,5-dimethyl-
phenyl)methylene]-2-(3,5-dimethylphenyl)cyclopropane, E-11,
whose smallest molecular diameter (∼7 Å) is significantly larger
than the pore size of HZSM-5, gave rise to the corresponding
1,3-butadiene (tt- and ct-12) and 1,2-dihydronaphthalene (13)
derivatives (Scheme 3) but failed to produce any perceptible
color on the surface (or inside) of the zeolite sample.
Quantitative Adsorption Experiments with E-5. To gain
further insight into the formation of tt-8•+@HZSM-5 from E-5
9
14500 J. AM. CHEM. SOC. VOL. 127, NO. 41, 2005

Methylenecyclopropane, Distal vs Proximal Bond Cleavage
A R T I C L E S
Table 2. Quantitative Adsorption Experiment of E-5 (1.03 mg, 5.0 µmol) onto HZSM-5 for 8 h
yield/%
a
a
HZSM-5
mg
Nh
mol
Ns
fraction of
fraction of
µ
µ
mol
E-5
tt-8
ct-8
9
SADb
tt-8•+@HZSM-5
occupied holes
occupied sites
100
200
300
0.082
0.16
0.24
33
66
100
53
22
2
7
9
10
2
3
3
5
7
10
3
9
10
30 (1.8 µmol)
50 (2.5 µmol)
65 (3.3 µmol)
∼18
∼15
∼13
∼0.05
∼0.04
∼0.03
a Quantities of holes of straight channels (Nh) and active sites (Ns) of HZSM-5 for tt-8•+ b Estimated amounts of the surface adsorption and decomposition.
.
Figure 6. A model of the adsorption of tt-8 into HZSM-5 zeolite, giving
rise to tt-8•+@HZSM-5.
Scheme 4 a
a Plain arrows; pathways supported by the reactions of E-5 and Z-5.
Dashed arrows; additional likely pathways.
Figure 5. Time-dependent changes in the product distribution obtained
by treating E-5 (black circles) with 100 mg of HZSM-5 (top) and 300 mg
of HZSM-5 (bottom) in TMP; tt-8 (lime green filled squares), ct-8 (gray
filled squares), 9 (orange filled circles), and tt-8•+@HZSM-5 (pink filled
diamonds).
apparently promotes cleavage of the proximal cyclopropane
bond. The likely course of the divergent reactions is considered
below.
Mechanism of the Photoinduced Reaction of E-5. The
significantly exergonic free energy for ET from E-5 to 1DCA*
(∆Get ) -0.62 eV) leaves little doubt that this photoreaction
proceeds via ET. The overall mechanism must accommodate
three key findings: (a) the early buildup of trans-6 from E-5
and the failure of the initial radical cation derived from E-5 to
form a DCA adduct;21 (b) the delay of Z-7 formation from E-5
in contrast to the ready adduct formation from Z-5; and (c) the
Z-stereochemistry of the adduct (Z-7).21 The unambiguously
documented conversions are summarized in Scheme 4.
The conversions between E- and Z-5 and cis- and trans-6
are typical MCP rearrangements. The corresponding well-
documented rearrangement of the dianisyl analogue, 1,3 clearly
proceeds via a ring-opened TMM-type radical cation formed
by cleavage of the distal cyclopropane bond. In addition, 1 is
known to give rise to a ring-closed as well as a ring-opened
molecular ion in the gas phase.22 It is tempting to invoke
analogous ring-opened species for E-5 and its isomers. On the
other hand, the known persistence of ring-closed radical cations
cis- and trans-14•+ and 15•+ (Chart 1), derived from 1,2-
(Table 2 and Figure 5, top). The yield and amount (30%, 1.5
µmol) of tt-8•+@HZSM-5 was determined from the yield data
and the amount of surface adsorption and decomposition (3%),
which was estimated from a control experiment with HZSM-5
and a bulky 1,3-butadiene, tt-12. This value indicated that about
18 molecules of tt-8 (∼1.5 µmol/0.082 µmol) enter a HZSM-5
channel through the same hole but that only 4.5% of the sites
(∼1.5 µmol/33 µmol) were occupied by tt-8•+. Thus, on the
average, 18 of the 400 sites are occupied by tt-8•+ in each
channel under these conditions (Figure 6). When larger amounts
(200 or 300 mg) of HZSM-5 were used with the same quantity
of E-5, the yield of tt-8•+@HZSM-5 increased (Figure 5,
bottom) but the population of occupied holes and sites decreased
accordingly (Table 2).
Discussion
The divergent results obtained in the oxidative rearrangement
of the MCP system, E-5, under the different reaction conditions
are incompatible with the involvement of only one intermediate;
the range of products obtained can be reconciled only if at least
two different intermediates are involved. The nature of the
products obtained by ET in solution (Scheme 2) supports
cleavage of the distal cyclopropane bond, whereas HZSM-5
(21) For a possible formation of E-7, a stereoisomer of Z-7, from E-10•+, see
the Supporting Information.
(22) Pineda, C.-J.; Roth, H. D.; Mujsce, A. M.; Schilling, M. L. M.; Reents, W.
D., Jr. J. Phys. Org. Chem. 1989, 2, 117-130.
9
J. AM. CHEM. SOC. VOL. 127, NO. 41, 2005 14501

A R T I C L E S
Chart 1
Ikeda et al.
aH) of the Z-1-phenylallyl radical24 (Figure 1), suggesting the
electronic structure, E-10-B•+, as shown below. However, the
observed spectrum is weak and may be complicated by the
presence of additional radical species.
Scheme 5
In summary, the true electronic structure of the principal
radical cation derived from E-5 and the detailed mechanism of
the MCP rearrangement remain open to question. However, the
combined evidence derived from product analysis, CIDEP
spectrum, and calculations together with the precedent of the
dianisyl derivative, 1, support ring-opened radical cations (10•+),
especially Z-10•+ as an incontrovertible precursor for Z-7. Our
previous work on 13c,d suggests that the products may be formed
via diradicals E-10•• and/or Z-10••; this aspect of the mechanism
is being pursued.
diphenylcyclopropanes23 and (diphenylmethylene)cyclopropane,3b
respectively, clearly indicates that ring-closed radical cations
cannot be excluded a priori. The failure of E-5 to form a DCA-
adduct (Z-7 or an isomer) argues against a ring-opened radical
cation derived from E-5•+, although its rapid conversion to
trans-6 may argue for a ring-opened species. Of course, the
formation of the DCA-adduct Z-7 provides irrefutable evidence
for a ring-opened TMM-type radical cation formed by cleavage
of the distal cyclopropane bond. The precedent of the PET
reaction between CA and 13 suggests coupling of a radical ion
pair, Z-10•+-DCA•-. In further analogy, the addition likely
involves coupling of singlet ion pairs with formation of
zwitterion, Z-16+- (Scheme 5).3
One key to the electronic structure of the principal radical
cation derived from E-5, ring-closed (E-5•+) or of the TMM-
type (E-10•+), lies in the distribution of the unpaired electron
spin density and of the positive charge. As mentioned earlier,
1 forms the unusual intermediate, 2•+, in which spin and charge
reside in separate π-systems. For the ring-opened radical cation
derived from E-5, two limiting possibilities exist to deploy spin
and charge: either between an E,E-1,3-diphenylallyl system and
a methylene function attached at the nodal C-2 (E-10-A•+) or
between a 1-phenylallyl system and a benzyl function attached
at C-2 (E-10-B•+).
Mechanism of the Zeolite-Induced Conversion of E-5. The
alternative method applied to oxidize E-5, incorporation into
HZSM-5, generated entirely different products (Scheme 2),
requiring the involvement of a different intermediate formed
by cleavage of the proximal cyclopropane bond. Two-thirds of
the starting material is incorporated into the zeolite and observed
as tt-8•+@HZSM-5. Additional quantities of tt-8 and of its
geometric isomer, ct-8, were isolated from the supernatant
solution, further supporting proximal bond cleavage. The third
product, 9, also requires cleavage of the proximal bond and
reveals either an additional pathway for the intermediate
involved or, alternatively, an additional intermediate.
The structure of the major product, tt-8•+@HZSM-5, rests
securely on a comparison of its ESR and DR spectra with those
obtained by incorporation of authentic tt-8.6a,b Furthermore, the
ESR spectra were in good agreement with those obtained by
Ramamurthy et al. upon adsorbing tt-8 onto ZSM-5.6a In
addition, an ESR spectrum simulated with hfccs calculated for
tt-8•+ on the basis of DFT reproduced the experimental spectrum
well. The hfcc’s shown below reflect the delocalization of
unpaired spin density throughout tt-8•+.
Density functional theory (DFT) and semiempirical calcula-
tions suggested different electronic structures for E-10•+. A
UB3LYP/6-31G** calculation indicated structure E-10-A•+ with
most of the unpaired spin density (71%) localized on the
methylene (CH2) carbon and the positive charge (97%) delo-
calized in the E,E-1,3-diphenylallyl subunit. On the other hand,
the UAM1 calculation favored electronic structure E-10-B•+.
The CIDEP spectrum18 (Figure 2, bottom) provides limited
insight into the structure of E-10•+. The strong emission signal
(g ) 2.0058) clearly is due to CA•-.20 The balance of the
spectrum (g ∼2.002) is ascribed to (a) radical cation(s) derived
from E-5•+. This spectrum was partially reproduced by a
spectrum simulated with the hyperfine coupling constants (hfcc,
(23) Roth, H. D.; Schilling, M. L. M. J. Am. Chem. Soc. 1980, 102, 7956-
7958.
(24) (a) Davies, A. G.; Sutcliffe, R. J. Chem. Soc., Perkin Trans. 2 1980, 819-
824. (b) Davies, A. G.; Griller, D.; Ingold, K. U.; Lindsay, D. A.; Walton,
J. C. J. Chem. Soc., Perkin Trans. 2 1981, 633-641.
9
14502 J. AM. CHEM. SOC. VOL. 127, NO. 41, 2005

Methylenecyclopropane, Distal vs Proximal Bond Cleavage
A R T I C L E S
Scheme 6
Figure 7. Shape and “width” of E-5, showing that it is compatible with
the HZSM-5 dimension.
Figure 8. Shape and “width” of Z-5; the bulk of the Z-isomer prevents
complete incorporation; either a phenylcyclopropane (left) or a styrene
fragment (right) may be incorporated.
The second mechanism involves protonation of E-5, either
inside HZSM-5 or on its surface, forming 5H+, which would
generate cis- and trans-17+ by a cyclopropylcarbinyl rearrange-
ment. Products ct-8 and 9 are then readily explained via trans-
17+ or different rotamers of cis-17+. Deprotonation of trans-
17+ would yield either tt- or ct-8; deprotonation of cis-17+
would yield either ct- or cc-8, whereas cyclization followed by
a hydrogen shift would generate 9 (Scheme 6). Interestingly,
the cyclized product 9 was not observed when adsorbing either
tt- or ct-8 onto HZSM-5. This finding suggests that product 9
is formed via a reactive intermediate, which is not accessible
from tt- or ct-8.
In trying to evaluate the merit of the two mechanisms, we
note that the conversion of E-5 to 9 cannot be explained without
invoking “special zeolite forces”, such as strong electrostatic
fields25 and/or electronic confinement26 in the zeolite interior;
the bifunctional radical cation Z-18•+ does not account readily
for the formation of 9. On the other hand, cis-17+ appears to
be a “natural” intermediate for the formation of 9. These
considerations lead us to conclude that the protonation-induced
mechanism is more plausible. This assignment is supported
indirectly by the failure of tt-8 to form 9. Because tt-8 is known
to undergo oxidation rather than protonation in HZSM-5,6a
pathways induced by protonation, i.e., formation of cis-17+, are
precluded.
Additionally, the ESR and DR spectra of tt-8•+@HZSM-5
essentially agree with the ESR (Figure 3, bottom) and absorption
spectra (Figure 4, top) obtained by γ-irradiation of tt-8 in a
glassy matrix of n-BuCl at 77 K. The fact that the ESR signals
of tt-8•+@HZSM-5 (Figure 3, top) are noticeably sharper than
those of tt-8•+ in the frozen matrix (Figure 3, bottom) may be
due to a less restricted mobility of the guest in the zeolite or to
multiple sites in the glassy matrix. Interestingly, the electronic
spectra of tt-8•+ in the two matrices have different maxima:
the DR spectrum in the zeolite (540, 817 nm) is slightly blue-
shifted compared to the absorption in the matrix (548, 822 nm).
This observation may be due to the strong electrostatic fields25
and/or the electronic confinement26 in the restrictive zeolite
interior.
These experimental results leave little doubt that the proximal
bond of E-5 is cleaved inside HZSM-5 or on its surface.
Considering the redox activity as well as the Brønsted and
Lewis acidities of the external and internal surfaces of HZSM-
5, we consider two mechanisms for the conversion of E-5 to
tt-8•+@HZSM-5, with two different species as key intermediates
(Scheme 6).
The first pathway to be considered is initiated by the oxidation
of E-5 molecules that are at least partially immersed in the
zeolite interior (cf., Figure 7). The resulting E-5•+@HZSM-5
might cleave the (phenyl-activated) proximal bond, yielding a
bifunctional 1,3-radical cation, Z-18•+@HZSM-5, as the
crucial intermediate. A subsequent inversion at the vinylic
radical center and a 1,2-H shift might convert Z-18•+@HZSM-5
to tt-8•+@HZSM-5. The isomeric product ct-8 can be formed
from Z-18•+ by a 1,2-H shift without inversion; however, the
formation of 9 cannot be explained readily.
The protonation-initiated mechanism is further supported by
the reactions of two bulkier substrates, the isomer, Z-5, and the
dixylyl derivative, E-11. The molecular diameters of Z-5 or E-11
(∼7 Å) are larger than the pore size of HZSM-5; the shape of
Z-5 allows partial incorporation (either the phenylcyclopropane
or the styrene fragment) (Figure 8), whereas E-11 is precluded
altogether from entering the HZSM-5 channels. Because the
oxidation-initiated mechanism is limited very likely to molecules
at least partially immersed into the zeolite interior, E-11 cannot
be expected to form any products.
(25) (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.; Lara, E. C. d. J. Phys. Chem. 1966, 100, 233-237.
(26) Ma´rquez, F.; Garc´ıa, H.; Palomares, E.; Ferna´ndez, L.; Corma, A. J. Am.
Chem. Soc. 2000, 122, 6520-6521.
The Z-isomer (Z-5) gave tt- and ct-8, dihydronaphthalene 9,
and the colored tt-8•+@HZSM-5 upon reaction with HZSM-5.
These results support the existence of active sites near the
9
J. AM. CHEM. SOC. VOL. 127, NO. 41, 2005 14503

A R T I C L E S
Scheme 7
Ikeda et al.
Conclusion
We have observed regiodivergent bond cleavage reactions
of E-5 by PET reaction in solution and via spontaneous
protonation-oxidation by a redox-active acidic zeolite. The PET
reaction in solution proceeds by cleavage of the distal cyclo-
propane bond, whereas the zeolite-induced reaction causes
cleavage of the phenyl-activated proximal cyclopropane bond.
The results observed for E-5 and for a bulkier analogue, E-11,
provide valuable insight into the mechanism of the zeolite-
induced reactions on both the external and internal zeolite
surfaces and about the distribution of Lewis and Brønsted acid
sites on the zeolite exterior and interior.
channel entrance or on the external zeolite surface; the formation
of 9 must occur outside the zeolite channels, whereas the
eventual oxidation to tt-8•+@HZSM-5 requires the complete
incorporation of tt-8 into the zeolite channels.
The bulky E-11 gave rise to the respective 1,3-butadienes
(tt- and ct-12) and 1,2-dihydronaphthalene derivatives (13) but
failed to produce coloration on the zeolite surface or any
evidence for tt-12•+@HZSM-5. These results support protona-
tion of E-11 on the zeolite exterior and argue against a
significant density of redox active sites on the external surface
of HZSM-5.
The results observed for Z-5 or E-11 require that the reaction
is triggered near the channel entrance or on the external surface
of HZSM-5, respectively, most likely via protonation at Brønsted
sites such as terminal silanol groups.27 Although we have no
evidence for details, recent studies28 suggest that the initial
protonation may proceed via a π-complex and an alkyl oxonium
species (Scheme 7). Product molecules of a suitable shape, viz.,
tt-8, can enter the zeolite interior for the oxidation step forming
tt-8•+@HZSM-5, either at a Brønsted or a Lewis site (e.g.,
nonframework AlO+).29
Acknowledgment. H.I. and S.T.-K. thank the Ministry of
Education, Culture, Sports, Science, and Technology (MEXT)
of Japan for financial support in the form of a Grant-in-Aid for
Scientific Research in Priority Areas (Area No. 417) and also
Professor M. Ueda (Tohoku University) for his generous
considerations. K.A. acknowledges financial support from the
MEXT of Japan (15350074 and 17651054) and the Core
Research for Evolutional Science and Technology of Japan
Science and Technology Agency. H.D.R. acknowledges finan-
cial support from the National Science Foundation (CHE-
9714850). We also acknowledge TOSOH Co. Ltd. for the zeolite
samples.
Supporting Information Available: General methods of the
experiments, preparation, and isolation of organic materials,
evaluation of the electron-accepting ability of zeolites, com-
parison of ESR spectra of tt-8•+@HZSM-5, and Cartesian
coordinates, partial spin, and charge densities of E-10•+
calculated by UB3LYP/6-31G** and UAM1 (PDF). This
material is available free of charge via the Internet at
http://pubs.acs.org.
(27) (a) Trombetta, M.; Busca, G. J. Catal. 1999, 187, 521-523. (b) Trombetta,
M.; Armaroli, T.; Alejandre, A. G.; Solis, J. R.; Busca, G. Appl. Catal., A
2000, 192, 125-136.
(28) (a) Haw, J. F.; Nicholas, J. B.; Xu, T.; Beck, L. W.; Ferguson, D. B. Acc.
Chem. Res. 1996, 29, 259-267. (b) Yang, S.; Kondo, J. N.; Domen, K. J.
Phys. Chem. B 2001, 105, 7878-7881. (c) Kondo, J. N.; Shao, L.;
Wakabayashi, F.; Domen, K. J. Phys. Chem. B 1997, 101, 9314-9320.
(29) (a) Jacobs, P. A.; Beyer, H. K. J. Phys. Chem. 1979, 83, 1174-1177. (b)
Zhang, W.; Han, X.; Liu, X.; Lei, H.; Liu, X.; Bao, X. Microporous
Mesoporous Mater. 2002, 53, 145-152.
JA045930N
9
14504 J. AM. CHEM. SOC. VOL. 127, NO. 41, 2005