Evaluation Only. Created with Aspose.PDF. Copyright 2002-2021 Aspose Pty Ltd.
Electron Transfer Reactions of Bicyclo[2.1.0]pentanes
J. Am. Chem. Soc., Vol. 118, No. 10, 1996 2383
3c′ 23 as rearrangement products, with the higher substituted 3c
as the major regioisomer. The tricyclic housane 2i gave for all
three electron transfer modes (entries 26-28) cyclopentene 3i
as the only rearrangement product.
Y[TP] mode even in the presence of the hindered base. Only
for this extremely acid-sensitive housane does the zeolite
framework8 seem to be acidic enough to catalyze rearrangement
in competition with electron transfer.
Control experiments on the cyclopentenes 3b,b′ 3e and 3c,c′
confirmed that these olefins did not interconvert into each other
under the employed electron transfer conditions. Furthermore,
by monitoring the product distribution of the photolyses as a
function of time, it has been determined that no secondary
oxidation of the cyclopentenes 3 took place up to the conversions
stated in Table 1.
Location of the 2,4,6-Triphenylpyrylium Cation (TP+) in
the Y-Type Faujasite. This question was assessed by absorp-
tion studies with DBH derivatives of different size on protonated
Y zeolite (HY). For the small azoalkanes 1a,h, a high
percentage (91-100%) of material was absorbed in contrast to
the large and bulky azoalkane 1j. Thus, the azoalkanes 1a,h
have the proper size to penetrate into the interior of the zeolite,
whereas the 1,4-diphenyl-substituted azoalkane 1j remains
outside, possibly absorbed on the external surface. Indeed, when
the PET reaction of azoalkane 1j was performed in the
heterogeneous phase, hardly any oxidation was observed. This
contrasts sharply with the extensive oxidation observed in
solution.3c Thus, these results suggest that most TP+ cations
are located internally in the cavities.
For the symmetrical 1,4-dialkylated housanes 2d,e, the CET
mode (TBA•+) proved to be again more suitable than the
photochemical ones in regard to conversion and mass balance
(compare entries 12, 15, and 16 Versus 11, 13, and 14). Thus,
complete conversion of the substrates to cyclopentenes 3d24
(entry 12) and 3e25 (entries 15 and 16) was achieved with
catalytic amounts of TBA•+. Monitoring of the reaction
1
progress by H NMR spectroscopy revealed that the initial
deuterium distribution in the housane 2e(D) was retained (eq
4) quantitatively (entries 15 and 16) in the cyclopentene product
Discussion
Chemical Electron Transfer of Bicyclopentanes 2. Table
1 reveals convincingly that the CET mode is advantageous for
the oxidative rearrangement of the bicyclopentanes 2. The same
chemistry as that for the PET mode is observed, except that
the oxidations proceed catalytically in a clean manner to afford
the rearranged cyclopentenes 3 in high yields. The lack of
alternative reaction channels (see Scheme 1) for the 1,3-radical
cation (and the 1,2-radical cation) presumably accounts for the
observed clean reactions and the high yields (75-99%) of the
CET mode, while in the PET mode, reaction with the reduced
photooxidant occurs (low mass balance, side products, etc.).3
This underscores the advantage of the chemical electron transfer
(CET) over the photosensitized (PET) modes.
However, as shown in the Results, at least one bridgehead
alkyl or aryl substituent on the bicyclopentane framework is
necessary for oxidation by TBA•+. Its oxidation power is not
sufficient to convert the parent unsubstituted housane 2a (Eox
) 1.91 V).29 Although the stronger oxidant TDA•+ smoothly
oxidizes even housane 2a, its application as electron transfer
oxidant is restricted to substrates for which the rearrangement
product, in this case the cyclopentene 3, has an unsubstituted
double bond, e.g. the cyclopentenes 3a,b′,c′, which are not
oxidized further to an intractable higher-molecular-weight
material. Fortunately, tri- and tetrasubstituted olefins persist
toward TBA•+. For this reason, only TDA•+ was used to oxidize
bicyclopentane 2a (entry 3).
In all cases examined, catalytic amounts (2-10 mol %) of
Ar3N•+ oxidant were enough to achieve complete conversion
of the substrate within a few seconds, although the aminium
salts possess lower oxidation potentials than the housanes 2.
The driving force presumably derives from the irreversible
rearrangement step which follows the endothermic electron
transfer process.30 Therefore, a catalytic cycle is proposed
(Scheme 2) in which electron transfer from housane 2 to the
aminium salt Ar3N•+ serves for initiation, followed by subse-
quent 1,2 Wagner-Meerwein rearrangement of the housane
radical cation 2•+. The cycle is completed by BET from the
3e(D). The well-separated 5-Ha,s resonances for housane 2e(D)
and the 2,3-H resonances for 3e(D) provided reproducible results
in the NMR analysis.
For the unsymmetrical 1,4-dialkylated housanes 2f-h, the
expected regioisomeric cyclopentenes 3f-h25,26 and 3f-h′ 27
were obtained in all cases (entries 17-25), with 3f-h as the
major regioisomers. Unexpectedly, oxidation of the trialkylated
housane 2h yielded only the two cyclopentenes 3h,h′ (entries
21-23) from H shifts, instead of the four possible ones derived
from H as well as methyl shifts.
The superiority of the chemical electron transfer mode
(complete conversion and higher mass balance) over the
photochemical ones was confirmed also for housane 2j (entries
29 and 31). Both PET and CET yielded olefin 3j3c and the
cycloreversion products 4j3c,28 and cyclopentadiene. The lack
of rearrangement products for the Y[TP] mode (entry 30)
confirmed that this bulky housane 2j did not enter the interior
of the zeolite and that outside the zeolite framework no electron
transfer process occurs.
The possible involvement of acid-catalyzed rearrangement
of the housanes 2 was suppressed by the addition of the hindered
base 2,6-di-tert-butylpyridine (not shown in Table 1). The
unaltered product distributions confirmed that this hindered
pyridine interferes with neither the PET nor CET reaction.
Contrary to the PET and CET reactions in solution, acid-
catalyzed rearrangement was observed with housane 2b for the
(23) (a) von Braun, J.; Ku¨hn, M. Ber. Dtsch. Chem. Ges. 1927, 60, 2551-
2557. (b) von Braun, J.; Kamp, E.; Kopp, J. Ibid. 1937, 70, 1750-1760.
(24) Rei, M. H. J. Org. Chem. 1978, 43, 2173-2178.
(25) Sisido, K.; Kurozumi, S.; Utimoto, K.; Isida, T. J. Org. Chem. 1966,
31, 2795-2802.
(26) (a) Vaveck, D.; Jaques, J. Bull. Soc. Chim. Fr. 1969, 10, 3505-
3515. (b) Richer, J. C.; Belanger, P. Can. J . Chem. 1966, 44, 2057-2066.
(27) Gajewski, J. J.; Squicciarini, M. P. J. Am. Chem. Soc. 1989, 111,
6717-6728.
(28) (a) Adam, W.; Reinhard, G.; Platsch, H.; Wirz, J. J. Am. Chem.
Soc. 1990, 112, 4570-4571. (b) Paquette, L. A.; Leichter, L. M. J. Org.
Chem. 1974, 39, 461-467.
(29) (a) Gassman, P. G.; Yamaguchi, R. Tetrahedron 1982, 38, 1113-
1122. (b) Gassman, P. G.; Yamaguchi, R.; Koser, G. F. J. Org. Chem. 1978,
43, 4392-4393.
(30) (a) The (irreversible) oxidation potentials for 2a and 2i are 1.91
and 1.42 V and the (reversible) ones for TDA and TBA are 1.50 and 1.06
V. The cyclovoltammetric measurements were performed in Prof. M.
Schmittel’s research group at the University of Wu¨rzburg, and we are
grateful to H. Trenkle for technical assistance. (b) The (irreversible)
oxidation potentials for 3a and 3i are 2.03 and 1.58 V.