Intramolecular Charge-Transfer-Induced Decomposition Promoted by an Aprotic Polar Solvent for Bicyclic Dioxetanes Bearing a 4-(Benzothiazol-2-yl)-3- hydroxyphenyl Moiety

Article abstract of DOI:10.1021/jo1021822

Bicyclic dioxetanes 3a-d bearing a 4-(benzothiazol-2-yl)-3-hydroxyphenyl group decomposed to give the corresponding keto esters 4a-d accompanied by the emission of bright light when simply dissolved in an aprotic polar solvent such as N-methylpyrrolidone

Full text of DOI:10.1021/jo1021822

pubs.acs.org/joc
Intramolecular Charge-Transfer-Induced Decomposition Promoted by an
Aprotic Polar Solvent for Bicyclic Dioxetanes Bearing
a 4-(Benzothiazol-2-yl)-3-hydroxyphenyl Moiety
Masatoshi Tanimura, Nobuko Watanabe, Hisako K. Ijuin, and Masakatsu Matsumoto*
Department of Chemistry, Kanagawa University, Tsuchiya, Hiratsuka, Kanagawa 259-1293, Japan
matsumo-chem@kanagawa-u.ac.jp
Received November 3, 2010
Bicyclic dioxetanes 3a-d bearing a 4-(benzothiazol-2-yl)-3-hydroxyphenyl group decomposed to
give the corresponding keto esters 4a-d accompanied by the emission of bright light when simply
dissolved in an aprotic polar solvent such as N-methylpyrrolidone (NMP) or DMF at 50-100 °C.
This solvent-promoted decomposition (SPD) was effectively a chemiluminescence process caused by
the hydrogen bonding of a phenolic OH with a solvent molecule(s). The characteristics of the
chemiluminescence in SPD resembled those in base-induced decomposition (BID), which occurs
through an oxidoaryl-substituted dioxetane 5 by an intramolecular charge-transfer-induced decom-
position (CTID) mechanism. Both free energies of activation, ΔGqSPD and ΔGqBID, increased in the
order 3a < 3b < 3c < 3d, and were linearly correlated with each other. However, SPD showed features
different from those of BID in terms of enthalpy of activation and entropy of activation. SPD had large
negative values for ΔSq (ca. -71 J mol-1 K-1) regardless of the substituent R at the 5-position for 3a-d,
while the ΔSq values for BID changed from 0.5 to -22 J mol-1 K-1 as R became smaller. The enthalpy
of activation ΔHq for SPD was 14-21 kJ mol-1 smaller than that for BID.
Introduction
which effectively gives a singlet-excited species.1,3-7 For
example, the deprotonation of hydroxyaryl-substituted
dioxetane 1 with a base produces unstable dioxetane 2 bearing
an oxidoaryl anion as an excellent electron donor that
undergoes CTID accompanied by the emission of bright
light (Scheme 1).3,8 In addition to such base-induced CTID
(BID), we have identified a new type of CTID, which we call
solvent-promoted decomposition (SPD). For SPD, dioxetane 1
The high-energy molecule 1,2-dioxetane typically under-
goes two types of decomposition leading to chemiexcitation.
The first is uncatalyzed thermal decomposition (TD), which
mainly gives a triplet-excited carbonyl fragment, and the
emission of bright light is hardly expected.1,2 The other is
the intramolecular charge-transfer-induced decomposition
(CTID) of a dioxetane bearing an aromatic electron donor,
(2) (a) Dewar, M. J. S.; Kirschner, S. J. Am. Chem. Soc. 1974, 96, 7578–
7589. (b) Turro, N. J.; Lechtken, P.; Shore, N. E.; Schuster, G.; Steinmetzer,
H.-C.; Yekta, A. Acc. Chem. Res. 1974, 7, 105–106. (c) Adam, W.; Baader,
W. J. J. Am. Chem. Soc. 1985, 107, 410–416. (d) Reguero, M.; Bernardi, F.;
Bottoni, A.; Olivucci, M.; Robb, M. J. Am. Chem. Soc. 1991, 113, 1566–1572.
(e) Wilson, T.; Halpern, A. M. J. Phys. Org. Chem. 1995, 8, 359–363.
(f) Vasil’ev, R. F. J. Biolumin. Chemilumin. 1998, 13, 69–74. (g) Tanaka,
C.; Tanaka, J. J. Phys. Chem. A 2000, 104, 2078–2090.
(1) (a) Bartlett, P. D.; Landis, M. E. In Singlet Oxygen; Wasserman,
H. H., Murray, R. W., Eds.; Academic Press: New York, 1979; pp 243-286.
(b) Adam, W. In The Chemistry of Peroxide; Patai, S., Ed.; Wiley: New York,
1983; pp 829-920. (c) Adam, W. In Small Ring Heterocycles; Hassner, A.,
Ed.; Wiley: New York, 1986; pp 351-429. (d) Adam, W.; Heil, M.; Mosandl,
€
T.; Saha-Moller, C. R. In Organic Peroxides; Ando, W., Ed.; Wiley:
€
New York, 1992; pp 221-254. (e) Saha-Moller, C. R.; Adam, W. In
€
(3) (a) Beck, S.; Koster, H. Anal. Chem. 1990, 62, 2258–2270. (b) Adam,
Comprehensive Heterocyclic Chemistry II. A Review of the Literature
1982-1995; Padwa, A., Ed.; Pergamon: New York, 1996; pp 1041-1082.
(f) Adam, W.; Trofimov, A. V. In The Chemistry of Peroxides; Rappoport,
Z., Ed.; Wiley: New York, 2006; Vol. 2, pp 1171-1209. (g) Baader, W. J.;
Stevani, C. V.; Bastos, E. L. In The Chemistry of Peroxides; Rappoport, Z.,
Ed.; Wiley: New York, 2006; Vol. 2, pp 1211-1278.
€
W.; Reihardt, D.; Saha-Moller, C. R. Analyst 1996, 121, 1527–1531.
(c) Matsumoto, M. J. Photochem. Photobiol., C 2004, 5, 27–53. (d) Matsumoto,
M.; Watanabe, N. Bull. Chem. Soc. Jpn. 2005, 78, 1899–1920. (e) Adam, W.;
Trofimov, A. V. In The Chemistry of Peroxides; Rappoport, Z., Ed.; Wiley:
New York, 2006; Vol. 2, pp 1171-1209.
902 J. Org. Chem. 2011, 76, 902–908
Published on Web 12/30/2010
DOI: 10.1021/jo1021822
r
2010 American Chemical Society

Tanimura et al.
JOCArticle
SCHEME 1. Thermal Decomposition, Base-Induced Decomposition, and Solvent-Promoted Decomposition by a CTID Mechanism for a
Hydroxyaryl-Substituted Dioxetane
CHART 1. Bicyclic Dioxetanes 3a-d Bearing a 4-(Benzo-
thiazol-2-yl)-3-hydroxyphenyl Group
decomposed to give bright light without the aid of a strong
base in an aprotic polar solvent.9,10 We report here the detailed
features of SPD and the differences in its thermodynamic
aspects compared to those of BID.
In this work, we investigated SPD and BID for a series of
bicyclic dioxetanes possessing an alkyl group on the same
skeleton to understand their characteristics from the perspective
of a substituent effect. The bicyclic dioxetanes selected for
the present investigation were 1-[4-(benzothiazol-2-yl)-3-hydro-
xyphenyl]-4,4-dimethyl-2,6,7-trioxabicyclo[3.2.0]heptanes
3a-d with an alkyl group R at the 5-position: 3a, R=methyl;
3b, R=ethyl; 3c, R = isopropyl; and 3d, R=tert-butyl, as
illustrated in Chart 1. For dioxetanes 3a-d, the thermody-
namic aspects of BID in acetonitrile and TD in p-xylene have
very recently been disclosed.11 Therefore, we thought that
these dioxetanes would be suitable for easily understanding
the characteristic features of SPD.
(4) The chemiluminescent (CL) decomposition of dioxetanes bearing an
aromatic electron donor has been proposed to proceed via the intramolecular
CIEEL (chemically initiated electron exchange luminescence) mecha-
nism,1,3,5 where an initially formed radical ion pair is annihilated by back
electron transfer (BET) to give an excited aromatic carbonyl compound.
However, the question of whether such a CL reaction includes BET as a
fundamental process is still being argued and remains unclear.6,7 Therefore,
we have recently been using the term CTID, which includes CIEEL and other
CT-induced mechanisms.
(5) (a) Koo, J.-Y.; Schuster, G. B. J. Am. Chem. Soc. 1977, 99, 6107–6109.
(b) Koo, J.-Y.; Schuster, G. B. J. Am. Chem. Soc. 1978, 100, 4496–4503.
(c) Zaklika, K. A.; Kissel, T.; Thayer, A. L.; Burns, P. A.; Schaap, A. P.
Photochem. Photobiol. 1979, 30, 35–44. (d) Catalani, L. H.; Wilson, T.
J. Am. Chem. Soc. 1989, 111, 2633–2639. (e) McCapra, F. J. Photochem.
Photobiol., A 1990, 51, 21–28. (f) McCapra, F. In Chemiluminescence and
Bioluminescence; Hastings, J. W., Kricka, L. J., Stanley, P. E., Eds., Wiley:
New York, 1996; pp 7-15. (g) Adam, W.; Bronstein, I.; Trofimov, T.;
Vasil’ev, R. F. J. Am. Chem. Soc. 1999, 121, 958–961. (h) Adam, W.;
Matsumoto, M.; Trofimov, T. J. Am. Chem. Soc. 2000, 122, 8631–8634.
(i) Nery, A. L. P.; Weiss, D.; Catalani, L. H.; Baader, W. J. Tetrahedron 2000,
56, 5317–5327.
Results and Discussion
1. Solvent-Promoted Decomposition and Base-Induced De-
composition for Bicyclic Dioxetanes Bearing a 4-(Benzothi-
azol-2-yl)-3-hydroxyphenyl Group. Dioxetanes 3a-d were
thermally stable and essentially did not decompose in a
nonpolar solvent such as p-xylene at room temperature: their
half-lives have been estimated to be 1-43 y at 25 °C in
p-xylene.11 However, 3a-d decomposed to emit weak green
light even at room temperature, when they were simply
dissolved in NMP (N-methylpyrrolidone). Thus, we investi-
gated the properties of the chemiluminescence for this solvent-
promoted decomposition (SPD). On heating at 60 °C in
NMP, 3a underwent SPD according to first-order kinetics to
SPD
give bright light with a maximum wavelength of λmax
=
(6) (a) Catalani, L. H.; Wilson, T. J. Am. Chem. Soc. 1989, 111, 2633–
2639. (b) McCapra, F. J. Photochem. Photobiol., A 1990, 51, 21–28.
(c) Wilson, T. Photochem. Photobiol. 1995, 62, 601–606.
499 nm, rate of SPD, kSPD=1.6 ꢀ 10-2 s-1, and chemilumi-
nescence efficiency, ΦSPD=0.25.12,13 Dioxetanes 3b-d simi-
larly decomposed at 60 °C in NMP to give bright light. Their
(7) (a) Takano, Y.; Tsunesada, T.; Isobe, H.; Yoshioka, Y.; Yamaguchi,
K.; Saito, I. Bull. Chem. Soc. Jpn. 1999, 72, 213–225. (b) Tanaka, J.; Tanaka,
C.; Matsumoto, M. In Bioluminescence and Chemiluminescence; Tsuji, A.,
Matsumoto, M., Maeda, M., Kricka, L. J., Stanley, P. E., Eds.; World Scientific:
Singapore, 2004; pp 205-208. (c) Tanaka, C.; Tanaka, J.; Matsumoto, M. In
Bioluminescence and Chemiluminescence; Tsuji, A., Matsumoto, M., Maeda,
M., Kricka, L. J., Stanley, P. E., Eds.; World Scientific: Singapore, 2004;
pp 209-212. (d) Isobe, H.; Takano, Y.; Okumura, M.; Kuramitsu, S.; Yamaguchi,
K. J. Am. Chem. Soc. 2005, 127, 8667–8679.
(8) (a) Schaap, A. P.; Gagnon, S. D. J. Am. Chem. Soc. 1982, 104, 3504–
3506. (b) Schaap, A. P.; Handley, R. S.; Giri, B. P. Tetrahedron Lett. 1987, 28,
935–938. (c) Schaap, A. P.; Chen, T. S; Handley, R. S.; DeSilva, R.; Giri, B. P.
Tetrahedron Lett. 1987, 28, 1155–1158.
(9) Part of this work has been reported as a preliminary communication.10
(10) Matsumoto, M.; Tanimura, M.; Akimoto, T.; Watanabe, N.; Ijuin,
H. K. Tetrahedron Lett. 2008, 49, 4170–4173.
(11) Tanimura, M.; Watanabe, N.; Ijuin, H. K.; Matsumoto, M. J. Org.
Chem. 2010, 75, 3678–3684.
(12) ΦCL was estimated based on the value 0.29 for the chemiluminescent
decomposition of 3-adamantylidene-4-(3-tert-butyldimethylsiloxyphenyl)-
4-methoxy-1,2-dioxetane in a TBAF/DMSO system.13
(13) Trofimov, A. V.; Mielke, K.; Vasil’ev, R. F.; Adam, W. Photochem.
Photobiol. 1996, 63, 463–467.
J. Org. Chem. Vol. 76, No. 3, 2011 903

Tanimura et al.
JOCArticle
FIGURE 1. Chemiluminescence spectra of dioxetanes 3a-d for SPD in NMP and BID in a TBAF/NMP system.
TABLE 1. Solvent-Promoted Decomposition (SPD) of Dioxetanes
3a-d
dioxetane solventa temp/°C λmaxSPD/nm
ΦSPD b
kSPD/s-1
3a
3b
3c
3d
3c
3d
3d
3c
3d
3c
3d
3c
3d
3d
NMP
NMP
NMP
NMP
NMP
NMP
DMSO
DMF
DMF
DMPU
DMPU
PPC
60
60
60
60
499
499
498
498
498
498
497
496
496
498
498
493
493
498
0.25
0.25
0.25
0.23
1.6 ꢀ 10-2
1.1 ꢀ 10-2
2.8 ꢀ 10-3
6.3 ꢀ 10-4
3.0 ꢀ 10-2
1.5 ꢀ 10-2 c
100
100
100
100
100
100
100
100
100
100
0.23
0.23c
1.3 ꢀ 10-3
0.22
d
-
3.6 ꢀ 10-2
1.2 ꢀ 10-2
1.0 ꢀ 10-2
2.6 ꢀ 10-3
8.2 ꢀ 10-3
2.4 ꢀ 10-3
0.23
0.25
0.25
0.25
PPC
EGM
0.24
FIGURE 2. Chemiluminescence spectra for the SPD of dioxetane
3d in (a) NMP, (b) DMSO, (c) DMF, (d) DMPU, and (e) PPC.
7.3 ꢀ 10-5 8.0 ꢀ 10-4
aNMP: N-methylpyrrolidone. DMPU: N,N0-dimethylpropyleneurea.
PPC: propylene carbonate. EGM: ethylene glycol monomethyl ether.
bBased on a value reported for the BID of 3-adamantylidene-4-(3-tert-
butyldimethylsiloxyphenyl)-4-methoxy-1,2-dioxetane in TBAF/DMSO.12,13
cValues reported in ref 10 are revised. dNot estimated.
TABLE 2. Base-Induced Decomposition (BID) of Dioxetanes 3a-d in a
TBAF/NMP Systema
dioxetane
λmaxBID/nm
ΦBID b
kBID/s-1
t1/2BID/sc
3a
3b
3c
3d
498
498
498
497
0.25
0.24
0.24
0.24
5.3 ꢀ 10-3
3.9 ꢀ 10-3
8.0 ꢀ 10-4
1.5 ꢀ 10-4
130
180
870
4500
SCHEME 2. Solvent-Promoted Chemiluminescent Decomposi-
tion of Bicyclic Dioxetanes 3a-d in NMP
aBID was carried out at 40 °C. bBased on a value reported for the BID
of 3-adamantylidene-4-(3-tert-butyldimethylsiloxyphenyl)-4-methoxy-
1,2-dioxetane in TBAF/DMSO.12,13 cHalf-life = ln 2/kBID
.
decomposition of 3d occurred, although ΦSPD was quite low,
as shown in Table 1. Dioxetane 3c as another representative
also effectively underwent CTID in DMF, DMPU, or PPC
as shown in Table 1.
Notably, acetonitrile hardly functioned as a solvent for
SPD of 3a-d, although it is a typical aprotic polar solvent
with a dielectric constant that is as large as that of NMP.14
This finding and the results summarized in Table 1 show that
chemiluminescent SPD appeared to effectively take place
only in an aprotic polar solvent that was a strong proton
acceptor and a weak proton donor.15-17 Thus, SPD should
proceed under the strong participation of hydrogen bonding
between a phenolic hydroxyl group of 3 and solvent
molecule(s).
chemiluminescence spectra and properties are summarized
together with those for 3a in Figure 1A and Table 1, respec-
tively. Table 1 shows that SPD was an effective chemilumines-
cence process for 3a-d. Note that the reaction mixture of
3a-d after SPD exclusively gave the corresponding keto esters
4a-d (Scheme 2).
One of our initial interests was to determine what kind of
solvent promoted the chemiluminescent decomposition of
dioxetanes 3. Thus, SPD was examined with 3d as a repre-
sentative compound at 100 °C in various solvents. As shown
in Figure 2 and Table 1, aprotic polar solvents such as DMF,
DMPU (N,N0-dimethylpropyleneurea), and PPC (propylene
carbonate) were as equally effective as NMP for SPD of 3d.
DMSO was not suitable for SPD, since 3d was effectively
reduced to an epoxide (Supporting Information): ΦSPD was
only 1/180 of that in NMP. When the reaction mixture was
heated in ethylene glycol monomethyl ether (EGM) as a
representative protic solvent with a high boiling point, the
(14) Reichardt, C. Solvents and Solvent Effects in Organic Chemistry;
Wiley-VCH: Weinheim, Germany, 2003.
(15) Gutmann, V. The Donor-Acceptor Approach to Molecular Interac-
tion; Plenum Press: New York, 1978.
(16) Chastrette, M. Tetrahedron 1979, 35, 1441–1448.
(17) Maria, P.-C.; Gal, J.-F.; de Franceschi, J.; Fargin, E. J. Am. Chem.
Soc. 1987, 109, 483–492.
904 J. Org. Chem. Vol. 76, No. 3, 2011

Tanimura et al.
JOCArticle
SCHEME 3. TBAF-Induced Decomposition of Dioxetanes 3a-d
FIGURE 3. Electronic absorption spectra and fluorescence spectra
of 4c in (a) NMP, (b) acetonitrle, and (c, d) NMP/acetonitrile mixed
solvent.
FIGURE 4. Time course for SPD of dioxetane 3c monitored by
electronic absorption spectrum in NMP.
360 nm, which was considerably different than that in NMP,
and a fluorescence spectrum at λmaxfl = 530 nm, rather than
at λmaxfl =498 nm, due to ESIPT (excited state intramole-
cular proton transfer).19 Figure 3 shows these differences in
the electronic spectra of 4c between NMP and acetonitrile.
Figure 3 shows also that the electronic absorption and
fluorescence spectra of 4c changed depending on the con-
centration of NMP in acetonitrile, and only 3 w/w % NMP
was needed to generate 6c or its closely related species to
some extent.
We next attempted to understand the features of dioxetane
3c dissolved in NMP. We expected that 3c would exist in a
concentration that was sufficient to show characteristic
electronic absorption spectrum,20 since 3c was estimated to
have a moderate lifetime at 25 °C (t1/2 ≈ 2 h, vide infra) in
NMP, though a solution of 3c in NMP always included
decomposition product 4c. A freshly prepared solution of 3c
in NMP showed an absorption spectrum with peak A at
λmaxab = 420 nm and peak B at λmaxab = 438 nm. The time-
course of the electronic absorption spectrum monitored for
SPD of 3c in NMP showed that peak A gradually decreased
and finally disappeared, while peak B increased with accom-
panying slight red-shift, and a new peak C due to 6c appeared
and increased (Figure 4).
We next investigated the base-induced decomposition
(BID) of 3a-d in NMP to understand how it compared to
the less-common SPD. Upon treatment with a large excess of
tetrabutylammonium fluoride (TBAF) in NMP at 40 °C,
3a-d decomposed according to pseudo-first-order kinetics
to effectively give light, the spectrum of which is shown in
Figure 1B.11,18 As summarized in Table 2, the results show
that both the efficiency and maximum wavelength of chemi-
luminescence for BID in NMP coincided with those for SPD
(Table 1). The spent reaction mixtures of 3a-d gave the
corresponding keto esters 4a-d in high yields after care-
ful neutralization. When authentic keto esters 4a-d were
dissolved in a TBAF/NMP solution, they showed fluores-
cence, the spectra of which coincided with the chemilumines-
cence spectra for BID of 3a-d in NMP. These results
strongly suggest that BID of 3a-d produced an oxido anion
of keto esters 6a-d in an excited state through dioxetanes
5a-d bearing an oxidophenyl anion (Scheme 3).
Thus, the electronic absorption and fluorescence spectra
of keto ester 4 were investigated to clarify the situation of 4
dissolved without any base in NMP (SPD conditions). When
4c as a representative was simply dissolved in NMP at 25 °C,
ab
an electronic absorption spectrum was observed at λmax
=
As a reference, we measured an electronic absorption
spectrum of anionic dioxetane 5c, which was generated
from 3c by using a large excess of TBAF in NMP: 5c was
444 and 467 nm and a fluorescence spectrum was observed at
λmaxfl = 498 nm (Figure 3). Both spectra coincided with
those for 6c generated from 4c in a TBAF/NMP system.
These results strongly suggested that 4c existed as anion 6c or
its closely related species even when dissolved simply in
NMP. On the other hand, when dissolved in acetonitrile,
4c showed an absorption spectrum at λmaxab = 347 and
(19) (a) Ikegami, M.; Arai, T. J. Chem. Soc., Perkin Trans. 2 2002, 1296–
1301. (b) Rodembusch, F. S.; Campo, L. F.; Stefani, V.; Rigacci, A. J. Mater.
Chem. 2005, 15, 1537–1541. See also references cited therein.
(20) The fluorescence spectrum due to 3c itself dissolved in NMP could
not be observed: even an NMP solution with freshly dissolved 3c showed a
fluorescence and/or chemiluminesnce spectrum with λmaxCL = 498 nm due to
6c and/or SPD of 3c.
(18) BID of 3a-d has been reported to show λmaxCL = 492 nm and ΦCL
0.27-0.28 in a TBAF/acetonitrile system.11
=
(21) When 3 is treated with a large excess of TBAF in acetonitrile or
NMP, 5 would be exclusively produced.
J. Org. Chem. Vol. 76, No. 3, 2011 905

Tanimura et al.
JOCArticle
TABLE 3. Activation Parameters of SPD, BID, and TD for Dioxetanes 3a-d
activation parameter
3a
3b
3c
3de
SPDa
BIDb
TDc
ΔHqSPD/kJ mol-1
ΔSqSPD/J mol-1K-1
ΔGqSPD/kJ mol-1
ΔHqBID/kJ mol-1
ΔSqBID/J mol-1K-1
ΔGqBID/kJ mol-1
ΔHqTD/kJ mol-1
ΔSqTD/J mol-1K-1
ΔGqTD/kJ mol-1
69.4 ( 0.5
-71.8 ( 0.8
90.8 ( 0.6
83.5 ( 0.5
-22.0 ( 0.2
90.0 ( 0.6
115 ( 0
-6.0 ( 0.0
117.0 ( 0.1
0
69.8 ( 0.5
-73.6 ( 0.8
91.8 ( 0.6
86.5 ( 0.4
-15.2 ( 0.1
91.0 ( 0.4
114 ( 1
-10.8 ( 0.1
117.4 ( 0.8
0.10
74.7 ( 0.3
-70.7 ( 0.5
95.8 ( 0.4
93.2 ( 0.2
-6.9 ( 0.0
95.3 ( 0.2
115 ( 1
-21.3 ( 0.3
121 ( 1
0.19
79.4 ( 0.5
-69.1 ( 0.7
100 ( 0.6
99.9 ( 0.3
0.5 ( 0.0
99.7 ( 0.3
124 ( 1
-8.4 ( 0.1
126 ( 1
0.30
-σ* d
-Esd
0
0.08
0.48
1.43
aAt 50-100 °C in NMP. bAt 35-60 °C in TBAF/NMP. cAt 80-120 °C in p-xylene.11 dTaft’s steric parameters for R. -σ*: polar substituent constants.
-Es: steric substituent constant.22 R is Me for 3a, Et for 3b, i-Pr for 3c, and t-Bu for 3d. eValues for SPD and TD of 3d reported in ref 10 are revised.
estimated to have a half-life of ca. 1.5 h at 25 °C (vide infra).
ab
The thus-measured absorption spectrum of 5c showed λmax
at 415 nm (Figure 4). On the other hand, when dissolved
in acetonitrile, 3c showed an absorption spectrum with
peaks at λmaxab = 291 and 338 nm, which was most likely
due to neutral dioxetane 3c itself (Figure 4). These results
suggested that the features of 3c dissolved in NMP were
significantly different from those of both anion 5c and
neutral 3c.
2. Thermodynamic Aspects of Solvent-Promoted Decom-
position and Base-Induced Decomposition of Bicyclic Dioxe-
tanes Bearing a 4-(Benzothiazol-2-yl)-3-hydroxyphenyl Group
in NMP. We previously stated that SPD resembled BID in
terms of its chemiluminescence spectra and efficiencies and
that SPD, like BID, proceeded according to (pseudo-)first-
order kinetics. On the other hand, experiments of electronic
absorption spectra suggested that dioxetane 3c dissolved in
NMP was apparently different from oxido-anion 5c generated
from 3c in TBAF/NMP or from neutral 3c in acetonitrile.
FIGURE 5. Relationship of ΔGq
dioxetanes 3a-d.
to ΔGq
and to ΔGq for
SPD
BID TD
Thus, we attempted to elucidate how SPD was mechanistically
similar to or different from BID.
We investigated the thermodynamic aspects of SPD as
well as BID in NMP with regard to the substituent effect for
3a-d. The time-course of decomposition was monitored at
35-60 °C for BID and at 50-100 °C for SPD. Notably, the
reaction temperature did not affect the maximum wave-
length or the efficiency of chemiluminescence for either
SPD or BID of any of the dioxetanes 3a-d. By using the
measured rate constants, we estimated activation parameters
from Eyring plots, i.e., enthalpy of activation ΔHq, entropy
of activation ΔSq, and free energy of activation ΔGq. The
results are summarized in Table 3, which also shows activa-
tion parameters for TD of 3a-d in p-xylene.11
log (k/k0) = F*σ* þ δEs, where σ* is a polar substituent con-
stant and Es is a steric substituent constant.22 DSP can be
applied to ΔGq, since log k at a given temperature is propor-
tional to the free energy of activation ΔGq. Thus, by using
the values for ΔGqs, σ*, and Es in Table 3, we estimated that
F*=-16 ( 11 kJ mol-1 and δ=-3.4 ( 2.2 kJ mol-1 for
ΔGqSPD (R2 = 0.98), andF*=-17 (13kJmol-1 and δ=-3.6
( 2.5 kJ mol-1 for ΔGqBID (R2 = 0.98). Although these values
have ratherlargestandarddeviations, wecansee thatSPDand
BID resembled each other and were sensitive to the polarity of
the alkyl substituents R at the 5-position in terms of the
response of ΔGq.
Table 3 shows that all of the free energies of activation, i.e.,
ΔGqSPD, ΔGqBID, and ΔGqTD, increased in the order 3a <
3b< 3c < 3d. Thus, the dependency ofΔGq on the substituent
R apparently resembled that among TD, BID, and SPD for
dioxetanes 3a-d. Figure 5 shows the relationships between
However, SPD showed features different from those of
BID in terms of ΔHq and ΔSq (Table 3). SPD proceeded with
ΔHq far smaller than that for BID: ΔHq
- ΔHq
=
SPD
BID
14-21 kJ mol-1, though both ΔHqSPD and ΔHqBID increased
in the order 3a<3b < 3c < 3d. The substituent effect on ΔHq
for SPD was more similar to that of TD than to that of BID.
Thus, the difference in ΔHqSPD, i.e. ΔΔHqSPD, between 3d
and 3a, which represents the extremes of bulkiness of the
substituent R among 3, was considerably smaller than
ΔΔHqBID and was nearly equal to ΔΔHqTD: ΔΔHq=9, 10,
and 16 kJ mol-1 for ΔHqTD, ΔHqSPD, and ΔHqBID, respectively.
The greatest difference in the thermodynamic aspects of
SPD from those of BID and TD was in the entropy of
activation ΔSq. For BID of 3a-d in TBAF/NMP, ΔSq varied
ΔGq
and ΔGq
and between ΔGq
and ΔGq for
SPD
BID
SPD TD
3a-d. We can see from Figure 5 that SPD was linearly
related to BID rather than to TD in terms of ΔGq. To better
understand the thermodynamic aspects of these decomposi-
tions, we analyzed the substituent effect in these decomposi-
tions by the use of Taft’s dual substituent parameter (DSP):
(22) (a) Taft, W., Jr. In Steric Effects in Organic Chemistry; Newman,
M. S., Ed.; Wiley: New York, 1956; pp 556-675. (b) MacPhee, J. A.; Panaye,
A.; Dubois, J.-E. Tetrahedron 1978, 34, 3553–3562.
906 J. Org. Chem. Vol. 76, No. 3, 2011

Tanimura et al.
SCHEME 4. Solvent-Promoted Decomposition of Dioxetane 3 through Hydrogen Bonding with a Solvent Molecule, NMP
JOCArticle
from 0.5 to -22 J mol-1 K-1 as the substituent R became
smaller. For TD of 3a-d, ΔSqs did not show such a tendency.
On the other hand, ΔSq for SPD had a markedly large
negative value (-69 to -74 J mol-1 K-1) regardless of the
substituent R for 3a-d. These results show that (a) SPD for
3a-d proceeded through a transition state that was far less
disordered than that for BID and (b) the structure of the
transition state was presumably controlled not by substituent
R but rather was regulated by an external factor(s), i.e.,
hydrogen bonding with NMP molecule(s).
A plausible mechanism (Scheme 4, mechanism A) consis-
tent with these results is that hydrogen bonding of 3 with the
NMP molecule induces negative charge on phenolic oxygen
as illustrated in 7 (Scheme 4), which undergoes chemilumi-
nescent decomposition by CTID mechanism: CTID occurred
with low enthalpy of activation as shown in Table 3, when a
phenolic ring lies in a particular conformation (vide infra).
Another candidate that we first thought is mechanism B in
which SPD of 3 proceeds consecutively through rapid pre-
equilbrium between 3 (with hydrogen bonding) and inter-
mediary species such as a contact ion pair 8 between anionic
dioxetane 5 and protonated NMP (Scheme 4). However,
considering that pKa values are 12-17 for phenols in NMP,23
mechanism B would be less plausible except in the case that
unknown factors cause predominant shift to 8in pre-equilibrium
between 3 in NMP.
For an oxidoaryl-substituted dioxetane, CTID has been
suggested to proceed preferentially when the oxidoaryl
group lies in a particular conformation(s).3c,11 According
to this idea, the result that SPD possessed low ΔHq and large
negative ΔSq suggests that species 7 activated through
hydrogen bonding with NMP could undergo CTID only
when an aromatic ring lay in a conformation(s) that was far
more regulated than in the case of BID: in contrast, for BID,
the intermediary anion 5 should be slightly solvated or, as an
extreme, naked in an aprotic polar solvent system such as
TBAF/NMP.
was an effective chemiluminescence reaction, in which diox-
etane 3 was presumably activated through hydrogen bond-
ing with aprotic polar solvent such as NMP so that it
underwent CTID. The chemiluminescence properties for
SPD resembled those for BID. Both free energies of activation,
ΔGqSPD and ΔGqBID, increased depending on the bulkiness of
the substituent R in 3, and were linearly correlated with each
other. However, there were marked differences in ΔSq and
ΔHq between SPD and BID. SPD had ΔSq with a large
negative value (-69 to -74 J mol-1 K-1) regardless of the
substituent R at the 5-position for 3a-d, while for BID it
changed from 0.5 to -22 J mol-1 K-1 as R became smaller.
ΔHq for SPD was 14-21 kJ mol-1 smaller than that for
BID. These results suggest that SPD proceeded only when
dioxetanes 3 with hydrogen bonding lay in a considerably
restricted conformation(s) with respect to an aromatic electron
donor.
Experimental Section
1-[4-(Benzothiazol-2-yl)-3-hydroxyphenyl]-5-tert-butyl-4,4-
dimethyl-2,6,7-trioxabicyclo[3.2.0]heptanes (3a-d) and 2,2-
Dimethyl-3-oxobutyl 4-(benzothiazol-2-yl)-3-hydroxybenzene-
carboxylates (4a-d). The bicyclic dioxetanes 3a-d and keto
esters 4a-d used here were previously synthesized and fully
6characterized,11 and were stored in a refrigerator. Their purities
were checked by 1H NMR before use in the present investigations.
Solvent-Promoted Decomposition of 1-[4-(Benzothiazol-2-yl)-
3-hydroxyphenyl]-5-tert-butyl-4,4-dimethyl-2,6,7-trioxabicyclo-
[3.2.0]heptane (3d): Typical Procedure. A solution of 1-[4-
(benzothiazol-2-yl)-3-hydroxyphenyl]-5-tert-butyl-4,4-dimeth-
yl-2,6,7-trioxabicyclo[3.2.0]heptane (3d) (34.8 mg) in NMP (2 mL)
was heated at 100 °C under a nitrogen atmosphere for 1 h. After
being cooled the reaction mixture was poured into water and
extracted with AcOEt. The AcOEt solution was dried over MgSO4
and concentrated in vacuo. The residue was chromatographed on
silica gel and eluted with AcOEt-hexane (1:4) to give 2,2-dimethyl-
3-oxobutyl 4-(benzothiazol-2-yl)-3-hydroxybenzenecarboxylate
4d as a colorless solid (34.6 mg, 99% yield). The thus-obtained 4d
showed spectroscopic properties identical with those of authentic
4d.11 Solvent-promoted decomposition of dioxetanes 3a-c was
similarly carried out in NMP and the corresponding decomposi-
tion products 4a-c were isolated in >99% yields.
Conclusion
Bicyclic dioxetanes 3a-d were found to undergo SPD
accompanied by the emission of bright light in an aprotic
polar solvent that was a strong proton acceptor. SPD of 3
(23) Izutsu, K. In Acid-Base Dissociation Constants in Dipolar Aprotic
Solvents; Blackwell Sci.: Oxford, UK, 1990.
J. Org. Chem. Vol. 76, No. 3, 2011 907

Tanimura et al.
JOCArticle
Measurement of Chemiluminescence and Time-Course of the
Solvent-Promoted Decomposition of Dioxetanes 3: General
Procedure. Chemiluminescence was measured with a JASCO
FP-750, and/or FP-6500 spectrometer, and/or Hamamatsu
Photonics PMA-11 multichannel detector.
A freshly prepared solution (2.90 mL) of TBAF (1.0 ꢀ 10-2
mol/L) in NMP was transferred to a quartz cell (10 ꢀ 10 ꢀ 50 mm3),
which was placed in a spectrometer that was thermostated with
stirring at an appropriate temperature range of 35-60 °C. After
3-5 min, a solution of dioxetane 3 in p-xylene (1.0 ꢀ 10-4 mol/L,
0.10 mL) was added by means of a syringe, and measurement was
started immediately. The time-course of the intensity of light
emission was recorded and processed according to first-order
kinetics. The total light emission was estimated as in the case of
solvent-promoted decomposition described above.
NMP (2.90 mL) that had been preincubated at an appropriate
temperature was transferred to a quartz cell (10 ꢀ 10 ꢀ 50 mm3),
which was placed in a spectrometer that was thermostated with
stirring at an appropriate temperature range of 50-100 °C.
After 3-5 min, a solution of the dioxetane 3 in p-xylene (1.0 ꢀ
10-3 mol/L, 0.10 mL) was added by means of a syringe, and
measurement was started immediately. The time-course inten-
sity of light emission was recorded and processed according
to first-order kinetics. The total light emission was estimated
by comparing it with that of an adamantylidene dioxetane,
the chemiluminescent efficiency ΦCL of which has been re-
ported to be 0.29 and which was used here as a standard.12,13
Similar experiments were carried out with DMF, DMPU, PPC,
and EGM.
Acknowledgment. The authors gratefully acknowledge
financial assistance provided by Grants-in-aid Nos. 17550050
and 21550052 for Scientific Research from the Ministry of Edu-
cation, Culture, Sports, Science, and Technology, Japan.
Supporting Information Available: General method for the
Experimental Section, 1H NMR/13C NMR spectra of an epoxide
A produced from 3d and ORTEP view and crystallographic
information file for epoxide A, as well as observed rate constants
at various temperatures for SPD and BID, and Eyring plots for
3a-d. This material is available free of charge via the Internet at
http://pubs.acs.org.
Measurement of Chemiluminescence and Time-Course of the
Base-Induced Decomposition of Dioxetanes 3: General Proce-
dure. Chemiluminescence was measured as in the case of solvent-
promoted decomposition described above.
908 J. Org. Chem. Vol. 76, No. 3, 2011