the corresponding dioxetane (ortho-3b). The thermal decom-
position of ortho-3b exhibited features completely different
from those for the dioxetanes described above. On standing for
several hours at room temperature in toluene or CDCl3, ortho-
3b changed into an unusual product (6) (pale yellow granules
melted at 59.0 °C)† without any detectable amount of the
normal product (ortho-4b) expected initially. Dioxetane ortho-
3b decomposed, however, into ortho-4b in high yield on
heating in refluxing toluene. It should be noted that both
products 6 and ortho-4b are thermally stable and do not change
into each other upon heating. Thus, we carried out thermolysis
of ortho-3b at various temperatures in toluene-d8 and measured
the product ratio of 6:ortho-4b) by 1H NMR spectroscopy:
6:ortho-4b = 93:7 at 50 °C, 72:28 at 70 °C, 35:65 at 90 °C,
11:89 at 110 °C. These results suggest that decomposition to
ortho-4b (mode A) and unusual decomposition to 6 (mode B)
take place concurrently in a temperature-dependant manner for
dioxetane ortho-3b.
The decomposition of mode B is most likely rationalized by
a mechanism similar to the Adam reaction,9¶ comprising the
intramolecular nucleophilic attack of an N-methylamino group
at the O–O moiety of the dioxetane and successive O–O bond
fission accompanying a proton exchange in an intermediary
twitterion (8, R1 = H, R2 = Me), as illustrated in Scheme 2.
Although the proposed mechanism includes multi-step reac-
tions, the decomposition of ortho-3b to 6 should be essentially
a unimolecular reaction as in the pathway to ortho-4b (mode A).
Consequently, the product ratio (6/ortho-4b) described above
should be equal to the ratio of rate constants (kB/kA) for the
corresponding modes at a given temperature. By plotting log(6/
ortho-4b) vs. 1/T, we estimated differences in activation energy
Ea and log A between modes A and B as DEa = Ea (A) 2 Ea (B)
= 18.9 kcal mol21 and DlogA = logA (A) 2 logA (B) = 11.7.
The result suggests that the mode B requires far a lower
activation energy and proceeds through a transition state far
more highly ordered than mode A. As such in the transition
state, one can image a structure where the o-aminophenyl
moiety lies in or near the plane comprising O–C–C shown as T-
1.
O–O as in the case of ortho-3b to 6; the initially formed
zwitterionic intermediate (8, R1, R2 = Me) may undergo
Stevens-like rearrangement10 to afford 7 as shown in Scheme
2.¶∑
In conclusion, the present results show that, for dioxetanes
bearing a substituted phenyl moiety, a p-amino group accel-
erates significantly decomposition of the dioxetane in the order
of H < OMe < < NH2 < NHMe < NMe2, while meta-
analogues are insensitive to this substituent effect. On the other
hand, o-methylamino and o-dimethylamino groups cause
preferentially unusual decomposition of dioxetane by their
intramolecular nucleophilic attack at O–O of the dioxetane,
though their unsubstituted amino analogue decomposes to give
the normal carbonyl product.
Notes and references
† Structures of all products obtained here were characterized by 1H NMR ,
13C NMR, IR, and mass spectral analysis. Selected data for 6: dH(400 MHz,
CDCl3) 0.92 (s, 9H), 1.22 (s, 3H), 1.54 (s, 3H), 3.14 (s, 3H), 3.72 (qAB, J
7.3, 2H), 4.01 (s, 1H), 6.78 (d, J 7.8, 1H), 7.05 (ddd, J 7.8, 7.3, 1.0, 1H), 7.33
(ddd, J 7.8, 7.3, 1.0, 1H), 7.54 (d, J 7.8, 1H); dC(100 MHz, CDCl3) 21.4,
25.7, 28.2, 39.4, 46.7, 47.6, 80.7, 88.4, 110.8, 119.5, 122.3, 127.5, 127.5,
129.8, 151.8. For 7: dH(400 MHz, CDCl3) 0.70 (br s, 9H), 1.19 (s, 3H), 1.50
(s, 3H), 2.86 (s, 3H), 3.73 (qAB, J 7.8, 2H), 4.40 (s, 1H), 4.42 (qAB, J 7.3,
2H), 6.76 (d, J 8.3, 1H), 6.86 (m, 1H), 7.26 (m, 1H), 7.75 (dd, J 7.8, 1.5, 1H);
dC(100 MHz, CDCl3) 22.7, 26.2, 28.2, 35.1, 40.2, 48.0, 78.5, 79.5, 90.6,
108.0, 112.7, 118.6, 125.2, 129.3, 132.0, 148.9.
‡ Dioxetanes para-3b,c were unstable under the chromatographic condi-
tions, so that the crude para-3b,c including little other than a trace amount
of keto ester (para-4b,c) was used without purification for thermolysis.
§ The rate of the CT-induced decomposition of a dioxetane bearing a
phenoxide anion as an electron donor has been reported to decrease via
restriction of rotation of the aromatic ring (ref. 8).
¶ Nucleophilic cleavage of a dioxetane with an aromatic amine is
unprecedented, although a sec-alkylamine has been reported to cause
decomposition of a dioxetane to yield N,N-dialkyl-O-(2-hydroxyethyl)hy-
droxylamine (Adam reaction) (ref. 9). The possibility cannot be ruled out
that the reaction of ortho-3b to give 6 proceeds by a mechanism including
attack of a diradical formed initially by homolytic O–O bond cleavage on an
amino group, although ortho-3a should also give an analogue of 6 by this
mechanism. The marked difference in decomposition mode between ortho-
3a and ortho-3b is most likely attributed to a difference in nucleophilicity
Finally, we attempted to synthesize a dioxetane (ortho-3c)
bearing an o-(N,N-dimethylamino)phenyl moiety. When singlet
oxygenation of a dihydrofuran (ortho-2c) was carried out
similarly to the case of ortho-2a in CH2Cl2 at 0 °C, ortho-2c
gave none of the expected dioxetane (ortho-3c) but gave instead
an unprecedented oxygenation product 7 (colorless granules,
mp 140.0 °C, 93%)† and a small amount of a keto ester (ortho-
4c) (7%). The low-temperature singlet oxygenation of ortho-2c
gave similar results, so that we could obtain little direct
evidence for formation of a dioxetane (ortho-3c). However, the
reaction is reasonably thought to proceed through an unstable
dioxetane (ortho-3c), because both 7 and ortho-4c are products
in which both carbons in the CNC moiety of the starting
dihydrofuran (ortho-2c) are oxygenated. Formation of the
unique cyclic aminal 7 is probably attributed to an intra-
molecular nucleophilic reaction of a dimethylamino group with
between NH2 and NHMe: the order of nucleophilicity would be NH2
NHMe NMe2. The thermal instability of ortho-3c might be also
rationalized by the high nucleophilicity of the NMe2 group.
<
<
∑ Nucleophilic attack of a tert-alkylamine on a dioxetane has been reported
to lead only to normal carbonyl products through Grob fragmentation (ref.
11) of an intermediary zwitterion (ref. 9). For an intramolecular reaction as
presented here, a zwitterion such as 8 might, however, cause predominantly
Stevens-like rearrangement, because an oxy anion would lie so close to a
methyl of the ammonium ion (ON+Me2) that the oxy anion is able to easily
abstract a methyl proton. The formation of a minor product (ortho-4c) may
be due to Grob fragmentation of 8 and/or direct decomposition of ortho-3c
as in the case of para- and meta-3.
1 A review: C. R. Saha-Moler and W. Adam, Four-membered Rings with
Two Oxygen Atoms in Comprehensive Heterocyclic Chemistry II, ed. A.
Padwa, Pergamon, NY, 1996, vol. 1B, pp. 1041–1082.
2 G. B. Schuster, Acc. Chem. Res., 1979, 12, 366.
3 L. H. Catalani and T. Wilson, J. Am. Chem. Soc., 1989, 111, 2633.
4 F. McCapra, Mechanism in Chemiluminescence and Bioluminescence-
Unfinished Business, in Bioluminescence and Chemiluminescence, ed.
J. W. Hastings, L. J. Kricka and P. E. Stanley, Wiley, NY, 1996, pp.
7–15.
5 K. A. Zaklika, T. Kissel, A. L. Thayer, P. A. Burns and A. P. Schaap,
Photochem. Photobiol., 1979, 30, 35; A. P. Schaap, S. D. Gagnon and
K. A. Zaklika, Tetrahedron Lett., 1982, 23, 2943.
R1
R2
N
O
O
N
R1 = H, R2 = Me
ortho-3b
ortho-3c
O
O
6
But
O
T-1
8
6 H. Siegerman, Oxidation and Reduction Half-Wave Potentials of
Organic Compounds, in Techniques of Electroorganic Synthesis, ed.
N. L. Weinberg, Wiley, NY, 1975, pp. 667–826.
R1 = R2 = Me
7 M. Matsumoto, N. Watanabe, N. C. Kasuga, F. Hamada and K.
Tadokoro, Tetrahedron Lett., 1997, 38, 2863.
8 M. Matsumoto, N. Watanabe, T. Shiono, H. Suganuma and J.
Matsubara, Tetrahedron Lett., 1997, 38, 5825.
9 W. Adam and M. Heil, J. Am. Chem. Soc., 1992, 114, 5591.
10 S. H. Pine, Org. React., 1970, 18, 403.
11 C. A. Grob, Angew. Chem., Int. Ed. Engl., 1969, 8, 535.
CH2
H
–
Me
+
N
O
O
7
But
O
Scheme 2
Communication 8/06823B
2320
Chem Commun., 1998