[9,9]-Sigmatropic Shift in a Benzidine-Type Rearrangement

Full text of DOI:10.1021/jo9703966

3794
J . Org. Chem. 1997, 62, 3794-3795
Sch em e 1
[9,9]-Sigm a tr op ic Sh ift in a Ben zid in e-Typ e
Rea r r a n gem en t
Koon Ha Park* and J in Soo Kang
Department of Chemistry, Chungnam National University,
Taejon, Korea 305-764
We have discovered a [9,9]-sigmatropic shift in the
acid-catalyzed benzidine rearrangement of bis[4-(2-furyl)-
phenyl]diazane (1). This [9,9] shift provides the first
example of such a shift not only in the benzidine
rearrangements1 but also in organic reactions. Our
findings also support the new formulation by Shine and
co-workers that the benzidine rearrangements follow the
patterns for sigmatropic processes.2 Thus, hydrazo 1
underwent exceptionally clean rearrangement to give the
HCl salt of 5,5′-bis(4-aminophenyl)-2,2′-bifuryl (2), which
was precipitated from the reaction solution almost free
from impurities, contrary to the ordinary benzidine
rearrangements that require tedious workup procedures
in order to get the pure product. Overall, 75% of the
rearrangement product 2 and 20% of disproportionation
products such as the corresponding azo 3 [bis[4-(2-furyl)-
phenyl]diazene] and fission amine 4 [4-(2-furyl)aniline]
and small amounts of unidentified products accounted
for the rearrangement of 1 in 95% ethanol at 0 °C
(Scheme 1).
Sch em e 2
Characterization of product 2 was achieved with
1
spectroscopic data3 such as MS and H and 13C NMR and
further confirmed by making its diacetyl derivative 5. All
1
the spectroscopic data such as H, 13C, and 2-D NOESY
NMR spectra fit the structure of 5 well.4
The specific rate constant in the rearrangement of 1
was 7.65 M-2 min-1 (at 0 °C in ca. 95% ethanol and 0.10
N in HCl), and the disappearance of 1 was found to be
second order in acid concentration.5 The rate constant
shows that the disappearance of 1 is 5.3-11 times faster
than that of hydrazobenzene 8 and 2.4 times faster than
that of 4,4′-divinylhydrazobenzene, but is slower than
that of p-hydrazobiphenyl by a factor of 1.7. Comparisons
could be made with the data reported previously; i.e., the
specific rate constants for rearrangement of hydrazoben-
zene 8 were 1.44 M-2 min-1 (0.15 °C in 95% ethanol and
0.102 N in acid)6 and 0.70 M-2 min-1 (0 °C in 75% ethanol
and 0.10 N in acid)7 and of 4,4′-divinylhydrazobenzene
and p-hydrazobiphenyl (14) 3.16 M-1 min-1 (0 °C in 95%
ethanol and 0.053 N in acid)6 and 13.05 M-2 min-1 (0 °C
in 95% ethanol containing about 8% water and 0.128 N
in acid),8 respectively.
(1) (a) Park, K. H. Heavy-Atom Kinetic Isotope Effects in Solving
Mechanisms of Benzidine Rearrangements: Hydrazobenzene and 2,2′-
Dimethoxyhydrazobenzene. Ph.D. Dissertation, Texas Tech University,
1983. (b) Dewar, M. J . S. In Molecular Rearrangements; de Mayo, P.,
Ed.; Interscience: New York, 1969; Vol. 1, pp 323-343. (c) Shine, H.
J . In Mechanisms of Molecular Migrations; Thyagarajan, B. S., Ed.;
Interscience: New York, 1969; Vol. 2, pp 191-247. (d) Shine, H. J . In
Aromatic Rearrangements; Elsevier: New York, 1967; pp 126-179. (e)
Cox, R. A.; Buncel, E. In The Chemistry of the Hydrazo, Azo, and Azoxy
Groups; Patai, S., Ed.; Wiley: New York, 1975; pp 775-859. (f)
Banthorpe, D. V. Chem. Rev. 1970, 70, 295-322. (g) Olah, G. A.;
Dunne, K.; Kelly, D. P.; Mo, Y. K. J . Am. Chem. Soc. 1972, 94, 7438-
7447. (h) Bunton, C. A.; Rubin, R. J . J . Am. Chem. Soc. 1976, 98,
4236-4246.
(2) (a) Rhee, E. S.; Shine, H. J . J . Am. Chem. Soc. 1986, 108, 1000-
1006. (b) Shine, H. J . In Isotopes in Organic Chemistry; Buncel, E.,
Saunders, W. H., J r., Eds.; Elsevier: Amsterdam, 1992; Vol. 8, Chapter
1.
The fast and clean conversion of 1 to 2 selectively by a
[9,9]-sigmatropic shift was surprising in view of the fact
that other bondings such as [3,3]- and [7,7]-sigmatropic
(3) 1H NMR (300 MHz, acetone-d6) δ: 7.50 (d, J ) 8.6 Hz, 4H, Ph-
H), 6.72 (d, J ) 8.6 Hz, 4H, Ph-H), 6.68 (d, J ) 3.4 Hz, 2H, furyl-H),
6.62 (d, J ) 3.4 Hz, 2H, furyl-H), 4.88 (s, 4H, NH2). MS m/e 316 (M+).
shifts would be also thermally allowed processes.
A
reasonable explanation for the unique rearrangement of
1 may lie in the structure of the transition state leading
to 2. That is, diprotonated hydrazo 6 (Scheme 2) is
expected to adopt a transition state in which aromatic
rings are parallel to each other and are bent to a certain
degree while maintaining an aromatic character.
In this bent transition state, one may imagine skel-
etons of phenyl and furyl as those of cyclohexadienyl and
2-butenyl, respectively. Then this transition structure
IR: 3400, 3300, 3200, 1630, 1605, 1505, 1460, 1270, 1180, 1020 cm-1
.
(4) 1H NMR (500 MHz, DMSO-d6) δ: 10.08 (s, 2H, NH), 7.71 (d, J
) 8.7 Hz, 4H, Ph-H), 7.66 (d, J ) 8.7 Hz, 4H, Ph-H), 6.98 (d, J ) 3.4
Hz, 2H, furyl-H), 6.87 (d, J ) 3.4 Hz, 2H, furyl-H), 2.05 (s, 6H, C(O)-
CH3). 13C: 24.07, 106.89, 107.92, 119.21, 124.07, 124.67, 138.92,
144.69, 152.63, 168.43. MS: m/e 400 (M+). IR (KBr): 3295 (br s, NH),
1658 (CdO), 1583, 1515, 1401, 1300, 1018, 823, 779 cm-1
.
(5) By measuring [1] at time t with Bindschedler’s Green, specific
rate constants depending on the acid concentration were obtained as
follows. k ) 7.67 M-2 min-1 at [HCl] ) 0.1 M and [1]o ) 0.01 M. k )
7.62 M-2 min-1 at [HCl] ) 0.2 M and [1]o ) 0.01 M. See the following
references for the titration method using Bindschedler’s Green. (a)
Dewar, M. J . S. J . Chem. Soc. 1946, 777-780. (b) Croce, L. J .; Gettler,
J . D. J . Am. Chem. Soc. 1953, 75, 874-879. (c) Cohen, M. D.;
Hammond, G. S. J . Am. Chem. Soc. 1953, 75, 880-883. (d) Shine, H.
J .; Trisler, J . C. J . Am. Chem. Soc. 1960, 82, 4054-4058. (e) Clovis,
J . S.; Hammond, G. S. J . Org. Chem. 1963, 28, 3290-3297.
(6) Shine, H. J .; Chamness, J . T. J . Org. Chem. 1963, 28, 1232-
1236.
(7) Shine, H. J .; Henderson, G. N.; Cu, A.; Schmid, P. J . Am. Chem.
Soc. 1977, 99, 3719-3723.
(8) Shine, H. J .; Stanley, J . P. J . Org. Chem. 1967, 32, 905-910.
S0022-3263(97)00396-4 CCC: $14.00 © 1997 American Chemical Society

Communications
J . Org. Chem., Vol. 62, No. 12, 1997 3795
would allow the distance between the 9,9′-carbons to be
shorter for efffective bonding, while making those be-
tween the 3,3′- and 7,7′-carbons larger. Regarding the
mechanism of the benzidine rearrangement, a new
understanding has emerged in the last 16 years based
on heavy-atom kinetic isotope effects (KIE).
The essence of the new understanding lies in the
successful demonstration of sigmatropic shifts in benzi-
dine rearrangements. Thus, hydrazobenzene 8 has been
shown with nitrogen and carbon KIE (eq 1)9 to give
p-benzidine 9 in a concerted way and diphenyline 10 in
a nonconcerted way by [5,5]- and [3,5]-sigmatropic shifts,
respectively.
The revisit of p-hydrazobiphenyl (14) in this work was
to seek, hopefully, any rearrangement product formed by
larger sigmatropic shift than [5,5]. However, we could
only confirm o-semidine and disproportionation products
as reported previously.8 The rearrangement of 15 gave
two rearrangement (∼46%) and disproportionation prod-
ucts (∼50%). Though characterization of the two rear-
1
rangement products is not complete, the H NMR spec-
trum of an isolated product indicates the involvement of
a [9,9] shift. On the other hand, each rearangement of
16 and 17 gave a p-benzidine-type compound by a [5,5]
shift.12
Several remarks are noteworthy in these benzidine-
type rearrangements. One question is why 2-furyl
derivative 1 would undergo a [9,9] shift much more
favorably than 15 while others (14, 16, and 17) would
not? A possible reason for this would be some sort of
O‚‚‚H-O-H‚‚‚O hydrogen bonding between the two furyl
oxygens holding the molecule in an appropriate confor-
mation, uniquely in the 2-furyl derivative 1 but not in
the other cases. The other is the possibility of two
successive [5,5]s instead of a [9,9] shift. The only certain
way of ruling the possibility in or out would be by
measuring a carbon KIE at the appropriate phenyl and
furyl positions, successively. Nevertheless, our experi-
mental results seem to provide at least two arguments
that weigh againt two [5,5]s. One argument is the faster
rate constant of 1 than that of 8. When 8 rearranges by
a [5,5] shift, all that remains is deprotonation. One
would expect that to be faster than the second [5,5] shift
needed in the case of 1, so that 8 might be expected to
rearrange faster than 1. A second argument is the good
product yield (75%). Supposing the two successive [5,5]s,
perhaps cleavage products would be expected from the
intermediate, which would be formed in between the two
[5,5]s.
Analogously, rearrangement products that are struc-
turally similar to o-benzidine (11) and p-semidine (13)
have been shown to be formed in a concerted way,
conforming with [3,3]- and [1,5]-sigmatropic shifts, re-
spectively. On the other hand, formation of rearrange-
ment products such as o-semidine (12) have been shown
to be nonconcerted, corresponding with a [1,3]-sigma-
tropic shift. Accepting that the benzidine rearrange-
ments are sigmatropic shifts,10 the possibility of a larger
sigmatropic shift than [5,5] would be anticipated. In this
regard, our example of a [9,9]-sigmatropic shift clearly
supports the new understanding of the benzidine rear-
rangement.
Among the previous reports dealing with the hydrazo-
aromatics, none has described rearrangement by any
sigmatropic shift larger than [5,5]. p-Hydrazobiphenyl
(14) is reported to disproportionate (75% at 0 °C, 88% at
25 °C) and concurrently to rearrange (25% at 0 °C, 12%
at 25 °C), forming an o-semidine-type product,8 though
a larger shift such as [7,7] or [9,9] would theoretically
be possible. 4,4′-Divinylhydrazobenzene is documented
to give only a polymeric compound without any rear-
rangement product.6 The rearrangement of 3,3′-divinyl-
hydrazobenzene is described to give 2,2′-divinylbenzi-
dine.11
In conclusion, a [9,9]-sigmatropic shift has been dis-
covered for the first time in the acid-catalyzed rearrange-
ment of bis[4-(2-furyl)phenyl]diazane (1). Further, this
example supports Shine’s new formulation that the
benzidine reaarrangements follow the patterns of sig-
matropic processes.
Ack n ow led gm en t. This investigation is dedicated
to Professor Henry J . Shine for his lifelong endeavors
in elucidating the mechanism of the benzidine rear-
rangements. This work was supported by grants from
the KOSEF-OCRC (1996) and the Ministry of Education
through the Basic Science Research Institute Program
(BSRI-96-3433).
With this in mind, we further studied the rearrange-
ment of p-hydrazobiphenyl (14), bis[4-(2-thienyl)phenyl]-
diazane (15), bis[3-(2-furyl)phenyl]diazane (16), and bis-
[3-(2-thienyl)phenyl]diazane (17) in order to compare the
results with those of 1.
Su p p or tin g In for m a tion Ava ila ble: Experimental pro-
cedures and characterization data of 5 (9 pages).
J O9703966
(9) In earlier reports, the ratio 70:30 for 9:10 was quoted; see any
reference in ref 1. Recently remeasured ratio of 9 to 10 in 75% aqueous
ethanol is 85:15; see ref 10.
(10) Subotkowski, W.; Kupczyk-Subotkowska, L.; Shine, H. J . J . Am.
Chem. Soc. 1993, 115, 5073-5076.
(11) Wiley, R. H.; Smith, N. R. J . Am. Chem. Soc. 1948, 70, 2295-
2296.
(12) Each rearrangement of 16 and 17 gave p-benzidine-type
compound in 21% and 17% isolated yield, respectively, in addition to
disproportionation products such as corresponding diazene and fission
amine.