The Light Emitter of the 2-Coumar-anone Chemiluminescence: Theoretical and Experimental Elucidation of a Possible Model for Bioluminescent Systems

Article abstract of DOI:10.1002/ejoc.201501515

The bioluminescence reaction of many biological subspecies, most notably fireflies, forms dioxetanone derivatives as high-energy intermediates. The thermal instability of dioxetanones complicates understanding of the transition from the ground state to th

Full text of DOI:10.1002/ejoc.201501515

DOI: 10.1002/ejoc.201501515
Communication
Chemiluminescence
The Light Emitter of the 2-Coumaranone Chemiluminescence:
Theoretical and Experimental Elucidation of a Possible Model
for Bioluminescent Systems
[
a]
[b]
[c]
[c]
Stefan Schramm, Isabelle Navizet,* Panče Naumov,* Naba K. Nath,
[b]
[a]
[a]
[a]
Romain Berraud-Pache, Pascal Oesau, Dieter Weiss,* and Rainer Beckert*
Abstract: The bioluminescence reaction of many biological aranones, synthetically accessible strongly chemiluminescent
subspecies, most notably fireflies, forms dioxetanone deriva- materials that mimic the bioluminescence reaction. The path-
tives as high-energy intermediates. The thermal instability of ways of chemiexcitation and photorelaxation are clarified on
dioxetanones complicates understanding of the transition from the basis of synthetic evidence and spectrochemical as well as
the ground state to the first excited state that leads to light computational mechanistic analysis.
emission. Herein, we report the reaction mechanism of 2-coum-
Introduction
stance, although the BL of the Siberian glowworm (Fridericia
heliota) and some fungal species have been known for centu-
Chemiluminescence (CL) and bioluminescence (BL) are photo-
chemical phenomena that lead to emission of light in the visi-
ble region of the electromagnetic spectrum by following a se-
quence of chemical reactions. In the case of BL, this excitation
by chemical transformation occurs inside enzymes (luciferases).
In CL, the reaction normally takes place in solution. The forma-
tion of a 1,2-dioxetanone moiety is common with many of the
[
1,2]
known CL and BL reactions.
This chemical group is formed
if molecular oxygen reacts with the subtract, which lead to the
formation of a peroxide bond (see compound 2 in Scheme 1).
Decomposition of this intermediate leads to a luminophore (ox-
yluciferin in BL or light emitter) in its first singlet excited state
(
S ) (chemiexcitation), which relaxes by emission of light (pho-
1
[
3–5]
torelaxation).
Despite the increase in the use of BL and CL
as probes, the comprehension of these systems remains incom-
plete. Indeed, the difficulty of getting RX structures from bio-
logical systems limits the number of available BL luminophores,
whereas many of them are deemed very unstable in pure form,
and their structures remain uncertain or are unknown. For in-
[
a] Friedrich-Schiller-Universität Jena, Institut für Organische Chemie und
Makromolekulare Chemie,
Humboldtstr. 10, 07743 Jena, Germany
E-mail: dieter.weiss@uni-jena.de
http://www.agbeckert.uni-jena.de/
[
[
b] Université Paris-Est, Laboratoire Modélisation et Simulation, Multi Echelle,
MSME UMR 8208 CNRS,
5
bd Descartes, 77454 Marne-la-Vallée, France
Scheme 1. Two suggested pathways for the decomposition of 1,2-dioxetan-
E-mail: isabelle.navizet@u-pem.fr
one intermediate 2 in the chemiluminescence of 2-coumaranone derivatives
1
2
c] New York University Abu Dhabi,
1. R is a halogen and R is an alkyl or aryl group (examples are described in
ref.[ ). Both closed-ring lactone 4* and salicylamide-like structure 5* are
viable candidates for the excited-state structure of the light-emitting species.
In one path (ii,iii,iv) the cyclization occurs before dioxetanone decomposition,
and in the other path (v,vi,vii) the cyclization occurs after dioxetanone de-
composition and light emission.
10]
P. O. Box 129188, Abu Dhabi, United Arab Emirates
E-mail: pance.naumov@nyu.edu
nyuad.nyu.edu/naumov-group
Supporting information for this article is available on the WWW under
http://dx.doi.org/10.1002/ejoc.201501515.
Eur. J. Org. Chem. 2016, 678–681
678
© 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Communication
ries, the structures of their substrates were characterized only alternative pathway, 1,2-dioxetanone 2 can decompose directly
[
6–8]
recently.
The protocols for cases in which synthetic proce- by releasing CO (path v) to afford the excited state (i.e., 5*) of
2
[
9,10]
dures are available
are laborious and cost ineffective, which compound 5, which subsequently relaxes and cyclizes (path vii)
greatly limits prospects for systematic studies of the related CL to product 6.
and BL mechanisms. The recent interest in new fluorescent
[
11–13]
probes
calls for the design and convenient access to di-
verse, chemically robust, stable, bright luminophores with tun-
able emission energies. All properties are crucial for their utility
as labels in bioanalytical applications, typically with microscopic
imaging techniques.
Results and Discussion
To fulfil the demand for chemically variable luminophores, we
recently prepared new derivatives of 2-coumaranones. This
class of chemiluminescent compounds (1 Scheme 1) first syn-
[
14–18]
thesized in 1979
are easily accessible and are strongly
The structures of the ethyl and methyl
[
19–21]
chemiluminescent.
derivatives were confirmed by X-ray diffraction analysis (Fig-
ure 1; for details see the Supporting Information, Figures S6–
S9). Upon treatment with base [i.e., 1,8-diazabicyclo[5.4.0]un-
dec-7-ene (DBU)] in polar aprotic solvents (e.g., MeCN, DMF, or
acetone), these compounds emit bright blue light that is readily
observable even with the naked eye in daylight (Figure 2, a).
The CL also occurs after oxidation with a minimum amount of
horse radish peroxidase (Figure 2, b). The energy released in
this reaction can be further transferred to a fluorescence dye;
the color of the secondary emission can be tuned across a wide
range of the visible spectrum by selecting the dye (Figure 2, c–
n). The 2-coumaranones own a 1,2-dioxetanone structure as an
Figure 2. Chemiluminescence and fluorescence of 2-coumaranones. (a) The
chemiluminescence is visible in daylight. (b) Chemiluminescence is induced
by horseradish peroxidase (HRP). (c–n) Fluorescence of organic dyes induced
by chemiluminescent light from the 2-coumaranones in DMF before (c–h)
and after (i–n) the addition of DBU as the base. Organic dyes: (c,i) a thiazole
dye (for the structure of the dye, see the Supporting Information), (d,j) Rhod-
amine 6G, (e,k) sodium fluoresceinate, (f,l) 9,10-bis(phenylethynyl)anthracene,
(
g,m) pyranine. Panels h and n show control experiments (no dye added).
To decipher the reaction pathways for the CL of 2-coumara-
nones and to identify the light emitter in the CL reaction, we
1
2
focused on compound 7 (Scheme 2, or 1 with R = F and R =
Et in Scheme 1). The chemical product of the CL of 7, com-
pound 8, was isolated from the reaction mixture and its struc-
ture was confirmed by X-ray diffraction (for details of the struc-
ture analysis, see the Supporting Information). This result con-
firmed that 8 was indeed the product of CL, at least after isola-
[
21]
intermediate (see compound 2 in Scheme 1)
with many of the known CL and BL reactions.
in common
Thus, they
[
1,2,22]
can be convenient analogues of BL precursors that provide in-
stant access to this unstable high-energy intermediate, the
structure of which has not yet been directly observed. Herein,
theCLmechanismandthestructureofthelightemitterof2-coum-
aranones were delineated to obtain fundamental insight into
the molecular mechanism of their light emission. This required
a combination of quantum chemical calculations and spectro-
chemical and structural analysis. The reaction pathways that
have been advanced for CL are shown in Scheme 1. Until now,
studies on these compounds have not distinguished between
[
15]
tion as described earlier.
CL emission of 7 was λ
The maximum wavelength of the
= 433 nm with a quantum yield
em,CL
–1
of (4.5 ± 0.2) × 10ꢀ E mol . The fluorescence maximum (λ
=
em,fl
4
34 nm) of the reaction mixture after any observable CL had
ceased (“spent solution”), excited at λ = 366 nm, was (within
spectral resolution) identical to that of CL emission. However, if
product 8 was excited in MeCN at the maximum absorption in
the absence of base, isolated N-protonated product 8 showed
[
14,19,20]
these two pathways.
Reaction of 1 with molecular oxy-
gen (path i) initially leads to high-energy 1,2-dioxetanone inter-
mediate 2. Decomposition of 2 can occur by cyclization
only very weak, hardly detectable fluorescence with λem,fl
=
3
28 nm. This observation indicated that the protonated form of
(
path ii), whereby spirocyclic 1,2-dioxetanone 3 is generated.
8
is neither the emitter in the CL reaction nor emitting in the
Spirocycle 3 decomposes by releasing CO (path iii) to afford
2
spent solution. This prompted us to consider other protonated
states as candidates, including tautomeric form 10 and its con-
jugate base 9, the resonance structures of which, 9a, 9b, and
product lactone 4 in its excited state (4*). After light emission
(
path iv), the lactone relaxes to its ground state (GS) 6. As an
9c, could be generated by deprotonation of 8 with a strong
base such as DBU (Scheme 2). α-Azacoumarinolate anion 9
could hydrolyze in strong basic solution to open-chain carb-
aminic acid 11. Indeed, free carbaminic acids such as 11 are
[
23]
well known to be thermally labile, even at room temperature;
decomposition by release of CO would lead to the phenolate
2
1
of 4-fluorosalicylamide 12. All of these reactions are viable
pathways and, thus, were examined as possible routes for the
Figure 1. Crystal structures of compound 7 in Scheme 2 (1 with R = F and
2
1
2
R = Et in Scheme 1), 1 (with R = F and R = Me in Scheme 1) and 8.
Eur. J. Org. Chem. 2016, 678–681 www.eurjoc.org
679
© 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Communication
ring opening before (light emitter) and after the emission (in the charged (–1) compound resulted in most stable anion form
the spent solution). We first focused our attention on the sim- 12m (Table S2), with λabs = 340 nm and λem = 405 nm, both
plest product, 4-fluorosalicylamide 12. Salicylamides are known values deviating by only 20 to 10 nm from the respective exper-
[
24]
to exhibit intense blue fluorescence.
We prepared 12 and imental values. The small magnitude of the blueshift (<0.1 eV)
analyzed its absorption and fluorescence spectra (Figure 3; for confirmed the reliability of the used DFT method for this sys-
details, see the Supporting Information and Table S1).
tem. The components in the solution after the CL decay were
established by a combined experimental and theoretical ap-
proach. O-Protonated tautomer 10 was excluded on the basis
of 2D-NOESY NMR spectroscopy (see the Supporting Informa-
tion). The calculations were consistent with this result on the
basis of the prohibitively high energy gap between 8 and 10
ΔE = 73.4 kJ mol– in vacuo). Deprotonation of 8, facilitated by
1
(
base, results in 9. Electronic calculations on deprotonated forms
showed that the negative charge is completely delocalized
9
over the heteroatoms of the molecule (Figure S3). Three reso-
nance structures are drawn in Scheme 2. Compound 10 and
deprotonated basic form 9 are hetero-analogous to 4-hydroxy-
coumarin. These, in turn, are well known for their intense blue
[
25]
fluorescence, especially at basic pH values.
Therefore, 9 is a
viable candidate as a fluorescent species. However, DFT calcula-
tions were consistent with the fact that the experimental
fluorescence of anion 9 is nondetectable (Table S5, Figure 3, e).
Furthermore, we considered the possibility that the fluores-
cence was generated by open-chain carbaminic acid 11. As
mentioned above, this compound is not very stable and de-
composes rapidly to 12 or cyclizes to 9. The computed Gibbs
free energies of these two reactions are in favor of decomposi-
Scheme 2. Tautomerization of chemiluminescent product 8 to α-azacoumarin
10 and deprotonation of chemiluminescent end-product 8 to α-azacoumarin-
olate anion 9 (resonance structures 9a, 9b, and 9c). Hydrolysis of lactone 9b
leads to open-chain carbaminic acid 11, which decomposes to 4-fluorosalicyl-
amide 12. Structures are the most stable calculated ones (see Table S3).
–1
–1
tion to 12 (Δ G = –50 kJ mol ) and 9 (Δ G = –25 kJ mol ). The
r
r
DFT- and CASPT2-calculated absorption and emission wave-
lengths of 11 are remarkably close to the experimental results
obtained on the spent solution (Table 1). Therefore, open-form
11, which coexists with nondetectable fluorescent (Figure 2, e,f)
cyclic form 9, is capable of fluorescence.
Table 1. Experimental and calculated wavelength maxima for the most stable
conformations of compounds 5 (conformer 5a) and 11.
λ [nm]
Spent solution[a] 5 (DFT)[b]
5 (CASPT2)[c]
11 (DFT)[b]
Absorption
Fluorescence
378
434
367
429
374
423
369
434
Figure 3. Fluorescence of potential emitters in DMF. (a) Compound 5 with
DBU, (b) compound 5 without DBU, (c) fluorosalicylamide 12 with DBU,
[a] For experimental details, see the Experimental Section in the Supporting
Information. [b] B3LYP/6-31+G(d,p) with polarized continuum model (PCM)
MeCN solvation. [c] CASPT2/ANO-RCC-VTZP in vacuo.
(
(
d) fluorosalicylamide 12 without DBU, (e) isolated product 8 with DBU, and
f) isolated product 8 without DBU. (h) Absorption and (i) fluorescence spec-
1
2
tra of 4-fluorosalicylamide 12 with DBU (black), compound 5 with DBU (blue),
and the spent solution after decay of the chemiluminescence (red). The solu-
tions were excited at their specific maximum absorption wavelength.
Compound 5* (with R = F and R = Et in Scheme 1, Table S5,
and excited ester form of 11) appears as the most viable emitter
in the CL reaction. The O-deprotonated form 5a is stabilized by
about 9.65 kJ mol– relative to N-deprotonated form 5b
1
Except for a spectral shift, the absorption and fluorescence
spectra of solution of 12 and the spent solution had similar
appearance and identical Stokes shifts. These results were used
to establish and calibrate a quantum calculation protocol for
compound 12 (see Computations Section in the Supporting In-
formation). The absorption and emission transitions of different
isomers of 12 in MeCN were computed at the DFT B3LYP/6-
(Scheme 3, Table S6, 5a is the protonated structure drawn in
Scheme 1). The spectral wavelength of this structure at the DFT
and CASPT2 levels is reported in Table 1. We then focused on
31+G(d,p) level of theory. The strongly basic conditions in the
experiments imply that the luminophore should exist as an an-
ion. Besides, the theoretical results for the neutral conformers
do not match the experimental data (Table S2). Calculations of
Scheme 3. Different structures and isomers of compound 5.
Eur. J. Org. Chem. 2016, 678–681
www.eurjoc.org
680
© 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Communication
the synthesis of compound 5. Initial attempts indicated that scopy experiments. S.S. acknowledges the Friedrich-Ebert Stif-
synthetic 5 is unstable in solution and decomposes quickly to tung for financial support and Prof. L. F. M. L. Ciscato, S. Haben-
final product 8. Therefore, all subsequent procedures were per- icht, Prof. S. Gräfe, and Dr. S. Kupfer for their advice. This re-
formed below 5 °C and in the absence of light. The recorded search was in part performed by using Core Technology Plat-
absorption and emission spectra of deprotonated synthetic 5 form resources at New York University Abu Dhabi.
correspond to those recorded from a spent solution (Figure 2)
and are consistent with the predicted theoretical results. This Keywords: Luminescence · Bioluminescence ·
finding is in agreement with the reaction path: v→vi→vii.
Photochemistry · Reaction mechanisms · Oxygen heterocycles ·
On the basis of these two arguments, we conclude that the Natural products
emitter of the CL reaction of 2-coumaranone 7 is form 5a of
compound 5.
[
1] I. Navizet, Y.-J. Liu, N. Ferré, D. Roca-Sanjuán, R. Lindh, ChemPhysChem
How can this study be a model for bioluminescent systems?
First, the present chemiluminescent system presents the forma-
tion and decomposition of the dioxetanone moiety, as in most
bioluminescent systems, and all insight into such a decomposi-
tion can be relevant for bioluminescent systems. Second, emit-
ter 5 may also be structurally related to the newly discovered
2011, 12, 3064–3076.
2] B.-W. Ding, P. Naumov, Y.-J. Liu, J. Chem. Theory Comput. 2015, 11, 591–
[
5
99.
[
[
3] T. Wilson, J. W. Hastings, Annu. Rev. Cell Dev. Biol. 1998, 14, 197–230.
4] J. Vieira, L. Pinto da Silva, J. C. G. Esteves da Silva, J. Photochem. Photobiol.
B 2012, 117, 33–39.
[5] L. Pinto da Silva, J. C. G. Esteves da Silva, ChemPhysChem 2012, 13, 2257–
[
7]
2262.5a.
Fridericia heliota oxyluciferin. The latter has two moieties in-
volved in bioluminescence: the lysine part, at which formation
[
6] V. N. Petushkov, M. A. Dubinnyi, A. S. Tsarkova, N. S. Rodionova, M. S.
Baranov, V. S. Kublitski, O. Shimomura, I. V. Yampolsky, Angew. Chem. Int.
Ed. 2014, 53, 5566–5568; Angew. Chem. 2014, 126, 5672–5674.
of dioxetanone and release of CO occur, and a chromophore
2
(
CompX). The conjugated system of CompX is very similar to
[7] M. A. Dubinnyi, Z. M. Kaskova, N. S. Rodionova, M. S. Baranov, A. Y. Gorok-
hovatsky, A. Kotlobay, K. M. Solntsev, A. S. Tsarkova, V. N. Petushkov, I. V.
Yampolsky, Angew. Chem. Int. Ed. 2015, 54, 7065–7067; Angew. Chem.
the herein-studied 2-coumaranone systems (see Figure S10).
Thus, the present chemiluminescence system may be used to
understand the role of the bioluminescent CompX moiety.
2015, 127, 7171–7173.
[
8] K. V. Purtov, V. N. Petushkov, M. S. Baranov, K. S. Mineev, N. S. Rodionova,
Z. M. Kaskova, A. S. Tsarkova, A. I. Petunin, V. S. Bondar, E. K. Rodicheva,
S. E. Medvedeva, Y. Oba, Y. Oba, A. S. Arseniev, S. Lukyanov, J. I. Gitelson,
I. V. Yampolsky, Angew. Chem. Int. Ed. 2015, 54, 8124–8128; Angew. Chem.
Conclusions
2015, 127, 8242–8246.
[
9] H. Würfel, D. Weiss, R. Beckert, A. Güther, J. Sulfur Chem. 2012, 33, 9–16.
In summary, by using a combined experimental and computa-
tional approach, a fundamental step in the mechanism of a
class of strongly chemiluminescent 2-coumaranones was herein
clarified. The key intermediates were identified on the basis of
direct comparison with their synthetic analogues, and conclu-
sions were aided by computational assessment of the viable
isomers. Both the experimental and computational results were
consistent and showed that the open form was the light emit-
ter, not the closed form, in the chemiluminescence of 2-coum-
aranones. The open form may be structurally related to the
chromophore (CompX) of the Fridericia heliota bioluminescence
system and provides a convenient model for the latter, which
is difficult to access by experimental methods. The results pre-
sented herein highlight the successful utility of complementary
approaches in resolving nontrivial mechanistic pathways. Fur-
ther calculations on the formation and opening of the diox-
etanone will provide additional information on this fundamen-
tal reaction in bio- and chemiluminescent systems.
[
[
10] T. B. Shrestha, D. L. Troyer, S. H. Bossmann, Synthesis 2014, 46, 646–652.
11] A. Roda, M. Mirasoli, E. Michelini, M. Di Fusco, M. Zangheri, L. Cevenini,
B. Roda, P. Simoni, Biosens. Bioelectron. 2016, 76, 164–179.
[12] M. Zangheri, L. Cevenini, L. Anfossi, C. Baggiani, P. Simoni, F. Di Nardo,
A. Roda, Biosens. Bioelectron. 2015, 64, 63–68.
[
[
13] L. J. Kricka, Anal. Chim. Acta 2003, 500, 279–286.
14] G. J. Lofthouse, H. Suschitzky, B. J. Wakefield, R. A. Whittaker, B. Tuck, J.
Chem. Soc. Perkin Trans. 1 1979, 1634–1639.
[15] B. Matuszczak, Monatsh. Chem. 1996, 127, 1291–1303.
16] B. Matuszczak, Die Pharmazie 1996, 51, 862–865.
17] B. Matuszczak, Monatsh. Chem. 1997, 128, 945–951.
18] B. Matuszczak, J. Prakt. Chem./Chem.-Ztg. 1998, 340, 20–25.
19] S. Schramm, L. F. M. L. Ciscato, P. Oesau, R. Krieg, J. F. Richter, I. Navizet, D.
Roca-Sanjuána, D. Weiß, R. Beckert, ARKIVOC (Gainesville, FL, U.S.) 2015,
5, 44–59.
20] S. Schramm, D. Weiß, H. Brandl, R. Beckert, H. Görls, D. Roca-Sanjuána, I.
Navizet, ARKIVOC (Gainesville, FL, U.S.) 2013, 3, 174–188.
21] L. F. M. L. Ciscato, F. H. Bartoloni, A. S. Colavite, D. Weiss, R. Beckert, S.
Schramm, Photochem. Photobiol. Sci. 2014, 13, 32–37.
[
[
[
[
[
[
[22] F. H. Bartoloni, M. A. de Oliveira, L. F. M. L. Ciscato, F. A. Augusto, E. L.
Bastos, W. J. Baader, J. Org. Chem. 2015, 80, 3745–3751.
[
23] M. M. Sidky, A. A. El-kateb, M. R. Mahran, I. T. Hennawy, H. A. A. El-Malek,
Phosphorus Sulfur Silicon Relat. Elem. 1987, 29, 11–15.
[
[
24] G. J. Woolfe, P. J. Thistlethwaite, J. Am. Chem. Soc. 1980, 102, 6917–6923.
25] R. Poláček, P. Májek, K. Hroboňová, J. Sádecká, J. Fluoresc. 2015, 25, 297–
Acknowledgments
2
93.
This work was financially supported by New York University Abu
Dhabi. The authors thank the service team of the IOMC and
IAAC in Jena and in particular Dr. W. Günther for NMR spectro-
Received: December 2, 2015
Published Online: January 12, 2016
Eur. J. Org. Chem. 2016, 678–681
www.eurjoc.org
681
© 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim