Notable Substituent Electronic Effects on the Regioselectivity of an Oxygen-Atom Abstraction in the Reaction of Unsymmetrically Substituted Monocyclic 1,2-Dioxolanes with Triphenylphosphine

Full text of DOI:10.1021/jo961444+

752
J . Org. Chem. 1997, 62, 752-754
Nota ble Su bstitu en t Electr on ic Effects on
th e Regioselectivity of a n Oxygen -Atom
Abstr a ction in th e Rea ction of
Un sym m etr ica lly Su bstitu ted Mon ocyclic
1,2-Dioxola n es w ith Tr ip h en ylp h osp h in e
F igu r e 1.
Manabu Abe,* Yohko Sumida, and Masatomo Nojima
tuted 3,3-di-p-anisyl-5,5-diphenyl 1,2-dioxolane (1b) was
used for the reduction to see if the electronic effect
influences the regioselectivity on the oxygen-atom ab-
straction (entry 2). Di-p-anisylethylene (2b) and benzo-
phenone (3b) were mainly produced, together with small
amounts of di-p-anisyl ketone (4b) and diphenylethylene
(5b). This result suggests that the electronic effect of the
p-anisyl group, which can stabilize the adjacent carboca-
tion center, is an important factor to control the product
selectivity, such that the cleavage of the C4-C5 and C3-
O2 bonds predominates. The similar electronic effect was
also observed for the reactions of trisubstituted diox-
olanes 1c,d (entries 3-5). Namely, disubstituted ethyl-
ene 2c,d and the corresponding aldehydes 3c,d were
preferentially obtained from 1c,d . In other words, of two
peroxidic oxygens the sterically more congested one is
predominantly abstracted. In the reaction of 1d , alcohol
7 was also obtained (vide supra, Scheme 2). From these
results, it is clear that for the unsymmetrically substi-
tuted 1,2-dioxolanes 1b-d the oxygen atom (O2) attached
to C3 is predominantly abstracted by Ph3P.
Department of Materials Chemistry, Faculty of Engineering,
Osaka University, Suita 565, J apan
Received J uly 30, 1996
The reactions of triphenylphosphine (Ph3P) with cyclic
peroxides, such as ozonides,1 endoperoxides,2,3 and di-
oxetanes,4,5 have been frequently reported. A reasonable
mechanism for the reduction is proposed that the first
biphilic insertion of Ph3P into the peroxide bond gener-
ates the phospholane intermediate A, which is followed
by the subsequent P-O bond scission and the elimination
of triphenylphosphine oxide (Ph3PO) via zwitterions B
to give the corresponding deoxygenated products (eq 1).
To clarify the intervention of the zwitterionic interme-
diate, such as B in eq 1, the reactions of (E)- and (Z)-1d
with Ph3P were performed in the presence of H2O,
respectively (eqs 3, 4 and entries 6, 7).7 The stereochem-
However, the selectivity of which oxygen atom in the
peroxy bond is abstracted by Ph3P has not been firmly
established (Figure 1). We now report herein that in the
reduction of monocyclic 1,2-dioxolanes 1 with Ph3P, the
electronic effect of the substituents attached to the
peroxide linkage controls the regioselectivity of the
oxygen-atom abstraction.
To understand the mode of the reaction of monocyclic
1,2-dioxolanes with Ph3P, the symmetrically-substituted
3,3,5,5-tetraphenyl-1,2-dioxolane 1a was selected first (eq
2 and Table 1). Treatment with an equimolar amount
istry of 1d was determined by the NOE measurements
(Figure 2). The diol 8 was produced at the expense of
the formation of 2-5 and 7. Interestingly, the formation
of the diol 8 was stereospecific. The syn-8 was exclusively
produced from (Z)-1d , while anti-8 was the sole product
from (E)-1d . These results strongly support the inter-
vention of the zwitterionic intermediates 10d and 11d
(Schemes 1, 2). The stereospecific formation of the diols
8 may be rationalized in terms of the nucleophilic attack
of water on the cationic P atom (Scheme 1). The 18O-
tracer study using H218O (5% 18O) seems to support this.
Namely, in the reaction of (Z)-1d the isolated Ph3PO
(83%) was labeled in 3% 18O, in contrast to the formation
of the nonlabeled diol syn-8 (61%).
of Ph3P in dry benzene at 60 °C resulted in the formation
of the O-O, C-C, and C-O bond cleavage products,
diphenylethylene 2a and benzophenone 3a together with
Ph3PO 6 (entry 1).6 Next, the unsymmetrically substi-
(1) Carles, J .; Fliszer, S. Can. J . Chem. 1969, 47, 1113.
(2) (a) Burns, P. A.; Foote, C. S.; Mazur, S. J . Org. Chem. 1976, 41,
899. (b) Burns, P. A.; Foote, C. S. J . Org. Chem. 1976, 41, 908. (c)
Ganem, B.; Holbert, G. W.; Weiss, L. B.; Ishizumi, K. J . Am. Chem.
Soc. 1978, 100, 6483. (d) Clennan, E. L.; Heah, P. C. J . Org. Chem.
1981, 46, 4105. (e) Clennan, E. L.; Heah, P. C. J . Org. Chem. 1982,
47, 3329. (f) Clennan, E. L.; Heah, P. C. J . Org. Chem. 1983, 48, 2621.
(g) Clennan, E. L.; Heah, P. C. J . Org. Chem. 1984, 49, 2284.
(3) (a) Balci, M. Chem. Rev. 1981, 81, 91. (b) Clennan, E. L.; Foot,
C. S. In Organic Peroxides; Ando, W., Ed.; J ohn Wiley & Sons Ltd.:
New York, 1992; Chapter 6, p 255.
(4) (a) Bartlett, P. D.; Baumstark, A. L.; Landis, M. E.; J . Am. Chem.
Soc. 1973, 95, 5267. (b) Bartlett, P. D.; Baumstark, A. L.; Landis, M.
E.; Lerman, C. L. J . Am. Chem. Soc. 1974, 96, 5267. (c) Baumstark,
A. L.; McCloskey, C. J .; Williams, J . E.; Chrisope, D. R. J . Org. Chem.
1980, 45, 3593. (d) Adam, W.; Ahrweiler, M.; Reinhardt, D.; Sauter,
M. Tetrahedron Lett. 1994, 35, 6063.
(5) Sauter, M.; Adam, W. Acc. Chem. Res. 1995, 28, 289.
(6) The 1,2-dioxolanes 1 used in this study were stable at 60 °C for
10 h.
(7) The similar H2O-trapping experiment of the zwitterionic inter-
mediate was reported by Clennan et al., see ref 2d.
S0022-3263(96)01444-2 CCC: $14.00 © 1997 American Chemical Society

Notes
J . Org. Chem., Vol. 62, No. 3, 1997 753
Ta ble 1. Red u ction of Dioxola n es 1 w ith P P h 3 in Dr y
Ben zen e a t 60 °C
Sch em e 1
entry
1a
time (h) additive
products and yields (%)b
1
2
1a
1b
10
10
2a (83), 3a (88), 6 (92)
2b (68), 3b (69), 6 (91), 4b (18),
5b (14)
3
1c
8
8
8
8
8
2c (48), 3c (52),c 6 (91), 4c (8),
5c (11)c
4
Z-1d
E-1d
Z-1d
E-1d
2d (44),c 3d (48),c 6 (89), 7 (20),
4d (8),c 5d (12)c
Sch em e 2
5
2d (42),c 3d (46),c 6 (88), 7 (18),
4d (9),c 5d (11)c
6d
7d
H2O
H2O
2d (14),c 3d (16),c 6 (88), 7 (5),
syn-8 (63), 4d (2),c 5d (3),c
2d (15),c 3d (17),c 6 (89), 7 (3),
anti-8 (51), 4d (3),c 5d (4)c
a
The used dioxolanes 1 were dried over P2O5 under reduced
b
pressure due to their hygroscopicities. Isolated yields unless
otherwise noted. c The yields were determined by GC analyses by
d
using n-decane as an internal standard. The reactions were
performed in the presence of 2 equiv of H2O.
Syn th eses of 1,2-Dioxola n es 1a -d . Preparation of 1,2-
dioxolanes 1a ,c,d were reported previously.11 The dioxolane 1b
was prepared from 1,1-di-p-anisyl-2,2-diphenylcyclopropane by
using the electron-transfer photooxygenation reported by Gollnick
et al.12 These hygroscopic dioxolanes 1a -d were dried over P2O5
under reduced pressure, prior to use. Spectroscopic data for (E)-
and (Z)-1d are as follows:
3,5-Dip h en yl-3-m eth yl-1,2-d ioxola n e ((E)-1d ): 1H NMR δ
1.71 (s, 3 H), 2.76 (dd, J ) 7.8 and 12.2 Hz, 1 H), 3.35 (dd, J )
7.3 and 12.2 Hz, 1 H), 5.19 (dd, J ) 7.3 and 7.8 Hz, 1 H), 7.22-
7.51 (m, 10 H). (Z)-1d : 1H NMR δ 1.72 (s, 3 H), 2.94 (dd, J )
7.8 and 12.2 Hz, 1 H), 3.16 (dd, J ) 7.8 and 12.2 Hz, 1 H), 5.44
(dd, J ) 7.8 and 7.8 Hz, 1 H), 7.20-7.54 (m, 10 H).
R ed u ct ion of 1,2-Dioxola n e 1 w it h P h 3P . A Gen er a l
P r oced u r e. A reaction flask (50 mL) was flushed with dry
argon. A mixture of 1 (1 mmol) and Ph3P (1 mmol) was heated
in dry benzene (10 mL) at 60 °C. After the reaction was finished,
the crude mixture of products was subjected to the GC analyses
by using n-decane as an internal standard. Then, the solvent
was removed under reduced pressure. The residue was sub-
jected to a flush column chromatographic separation on silica
gel [elution with a mixture of ethyl acetate-hexane (5:95-30:
70)]. Products and yields are listed in Table 1. The isolated 7
was identified by the comparison of the NMR spectrum with that
of the reported authentic sample.13
F igu r e 2.
The mechanism for the regioselective oxygen-atom
abstraction from 1b-d is summarized in Scheme 2.
Biphilic insertion of Ph3P into the peroxy bond occurs first
to generate the labile phospholane intermediate 9. Sub-
sequent P-O bond scission leads to the zwitterionic
intermediates 10 and 11 which may be in equilibrium
with 9. Then, the ionic cleavages of the C-C and C-O
bonds occur to give the observed products. As mentioned
in the experimental results, it is clear that the electronic
effects are important factor to determine the product
distributions. The selective formation of 2 and 3 can be
explained by the relative rates in the rate-determining
step from the zwitterionic intermediates 10,11 to the final
products. Namely, the rates for the formation of 2,3 from
10 may be faster than that for the formation of 4,5, since
the transition states for the formation of 2,3 from 10
should be more stabilized by the electron-donating sub-
stituents at C3, compared with the transition states for
the formation of 4,5.8 The regioselective formation of the
unsaturated alcohol would be rationalized by the in-
tramolecular 1,5-H+ shift in the intermediate 10d .
1,3-Dip h en yl-3-bu ten -1-ol (7): 1H NMR δ 2.12 (s, 1 H), 2.84
(dd, J ) 8.9 and 14.2 Hz, 1 H), 2.99 (dd, J ) 4.3 and 14.2 Hz, 1
H), 4.71 (dd, J ) 4.3 and 8.9 Hz, 1 H), 5.15 (s, 1 H), 5.40 (s, 1
H), 7.20-7.65 (m, 10 H).
Red u ction of (E)- a n d (Z)-1d w ith P h 3P in th e P r esen ce
of H2O. A reaction mixture of (E)-1d or (Z)-1d (1 mmol) and
Ph3P (1 mmol) in dry benzene (10 mL) was heated in the
presence of H2O (2 mmol) at 60 °C, respectively. After the
reaction was finished, the products were analyzed by GC as
shown above. The reaction mixture was subjected to a flush
column chromatographic separation on silica gel [elution with
a mixture of ethyl acetate-hexane (5:95-30:70)]. The isolated
Exp er im en ta l Section
Gen er a l. 1H and 13C NMR spectra were recorded on J EOL
J NM-EX-270 spectrometer at 270 MHz and 67.8 MHz, respec-
tively. 1H NMR chemical shifts were reported in ppm (δ) using
residual CHCl3 (δ 7.26) in CDCl3. 13C NMR chemical shifts were
reported in ppm (δ) relative to the internal standard CDCl3 (δ
77.00). IR spectra were recorded on a Hitachi 260-30 spectro-
photometer. Mass spectrometric data were obtained by using
a J EOL J NS-BX 303-HF mass spectrometer. GC analyses were
performed with a Shimadzu GC-9A utilizing a flame ionization
detector on SE-30 capillary column. Flush column chromatog-
raphy was performed using silica gel (Wakogel C-300) as
absorbent. Benzene was dried and distilled from sodium ben-
zophenone ketyl prior to use.
(9) Molecular orbital calculations by the PM3 method10 were per-
formed with the MOPAC program (ver 6.0) using Iris Indigo R4000
computer. The structural output was recorded by using the MOL-
GRAPH program (ver 2.8) by Daikin Industries, Ltd.
(10) Stewart, J . J . P. J . Comput. Chem. 1989, 10, 209.
(11) Yoshida, M.; Miura, M.; Nojima, M.; Kusabayashi, S. J . Am.
Chem. Soc. 1983, 105, 6279.
(12) Gollnick, K.; Xiao, X.-L.; Paulmann, U. J . Org. Chem. 1990,
55, 5945.
(8) For your information, the heat of formation of 10d was 0.6 kcal/
mol lower than that of 11d , calculated by PM3 method.9
(13) Sidduri, A.; Rozema, M. J .; Knochel, P. J . Org. Chem. 1993,
58, 2694.

754 J . Org. Chem., Vol. 62, No. 3, 1997
Notes
anti- or syn-8 were identified by the comparison of the NMR
spectra with those of the reported authentic samples.14
2,4-Dip h en ylbu ta n e-2,4-d iol (a n ti-8): 1H NMR δ 1.51 (s, 3
H), 2.12 (dd, J ) 2.7 and 14.9 Hz, 1 H), 2.81 (dd, J ) 10.3 and
14.9 Hz, 1 H), 4.36 (s, 1 H), 4.36 (s, 1 H), 4.46 (dd, J ) 2.7 and
10.3 Hz, 1 H), 7.18-7.54 (m, 10 H). syn -8: 1H NMR δ 1.80 (s, 3
H), 1.97 (dd, J ) 2.4 and 14.9 Hz, 1 H), 2.18 (dd, J ) 10.8 and
14.9 Hz, 1 H), 3.44 (s, 1 H), 4.41 (s, 1 H), 5.18 (dd, J ) 2.4 and
10.8 Hz, 1 H), 7.20-7.53 (m, 10 H).
determined by the comparison of the mass spectroscopic data of
the products with those of the authentic samples by the reported
method.15 Mass spectroscopic data for labeled Ph3PO and
natural Ph3PO are as follows:
La beled P h 3P O: EIMS m/z (relative intensity) 278 (M+, 52),
279 (M+1, 18), 280 (M+2, 5). Na tu r a l P h 3P O: EIMS m/z
(relative intensity) 278 (M+, 53), 279 (M + 1, 10), 280 (M + 2,
1).
18O-Tr a cer Stu d y. A 18O-tracer study was performed in the
same conditions depicted above except for using H218O (5%
labeled). The contents of 18O atom in the products were
Ack n ow led gm en t. We thank all of the referees as
well as the editors very much for their kind suggestions.
J O961444+
(14) Ukaji, Y.; Kanda, H.; Yamamoto, K.; Fujisawa, T. Chem. Lett.
1990, 597.
(15) Sawaki, Y.; Foote, C. S. J . Am. Chem. Soc. 1979, 101, 6292.