Fluorimetric detection of Mg2+ and DNA with 9-(alkoxyphenyl) benzo[b]quinolizinium derivatives

Article abstract of DOI:10.1039/c2ob06948b

A benzo[b]quinolizinium-benzo-15-crown-5 ether conjugate 2a is presented that enables the fluorimetric detection of Mg²⁺ and DNA by a significant light-up effect, along with a change of the emission wavelength with different analytes (Mg²⁺

Full text of DOI:10.1039/c2ob06948b

View Article Online / Journal Homepage / Table of Contents for this issue
Dynamic Article Links
Organic &
Biomolecular
Chemistry
Cite this: Org. Biomol. Chem., 2012, 10, 3010
www.rsc.org/obc
PAPER
Fluorimetric detection of Mg2+ and DNAwith 9-(alkoxyphenyl)benzo[b]-
quinolizinium derivatives†
Maoqun Tian, Heiko Ihmels* and Shite Ye
Received 19th November 2011, Accepted 8th February 2012
DOI: 10.1039/c2ob06948b
A benzo[b]quinolizinium-benzo-15-crown-5 ether conjugate 2a is presented that enables the fluorimetric
detection of Mg2+ and DNA by a significant light-up effect, along with a change of the emission
wavelength with different analytes (Mg2+: 495 nm; DNA: 550 nm). The mechanism of the excited-state
deactivation of 2a was investigated by steady-state fluorescence spectroscopy in media of varied viscosity
and compared with the photophysical properties of methoxyphenyl-substituted benzo[b]quinolizinium 2b
(m,p-diOMe), 2c (m-OMe), and 2d (p-OMe) as reference compounds. Compounds 2a–c, which share the
m-alkoxyphenyl substituent as the common feature, have low emission quantum yields (ΦF < 10−2 in
water) but exhibit a significant increase of their fluorescence intensity in viscous glycerol solutions. In
contrast, the viscosity of the medium does not influence the emission properties of the parent phenyl-
substituted benzo[b]quinolizinium 2e and of the p-methoxyphenyl-substituted derivative 2d. Based on
these observations it is concluded that the excited-state deactivation in 2a–c is mainly due to the rotation
of the m-alkoxy group about the Car–O bond. The interaction of 2a–c with DNA or Mg2+ ions was
studied by spectrophotometric titrations and CD spectroscopy. Notably, the association of 2a or 2b with
DNA or 2a with Mg2+ ions induces a strong fluorescence enhancement (15- and 40-fold for DNA,
450-fold for Mg2+), which is rationalized by the suppression of the torsional-relaxation of the alkoxy-
substituent in the excited state. Additionally, the cation-induced light-up effect of 2a is selective towards
+
Mg2+ ions as compared with other cations such as NH4 , Li+, Na+, K+ and Ba2+.
whereas probes that allow the optical detection of different types
Introduction
of analytes are rather rare.3
The detection of analytes with fluorescent probes is among the
most useful and versatile techniques in different areas of chem-
istry,1 because it offers the advantages of fluorescence spec-
troscopy combined with the high potential of chemical synthesis
to provide almost any fluorophore with the required substitution
pattern. Hence, numerous chemosensors have been developed
for the fluorimetric detection of target analytes.1 In a consequent
development of this strategy, several bi- or multifunctional fluor-
escent probes have been designed that enable the detection of
different analytes, ideally with analyte-specific emission proper-
ties.2 Nevertheless, it is interesting to note that most of the latter
probes focus on analytes from the same class of compounds,
We have recently shown that the combination of the benzo[b]-
quinolizinium fluorophore with appropriate receptor units
enables the optical detection of metal cations, such as Hg2+,
Cu2+, or Mg2+, in most cases even in aqueous solution due to
the high water solubility of the quinolizinium unit.4 Furthermore,
it has been demonstrated that the DNA-binding properties of the
benzo[b]quinolizinium ligand may be employed in combination
with a crown-ether receptor unit to detect Hg2+ and DNA simul-
taneously.5 Encouraged by these results we aimed at the variation
of this concept, i.e. by changing the receptor unit and its inte-
gration in the fluorophoric system in compound 2a. We proposed
that this compound may also be used as dual-pathway probe, that
is, it likely exhibits, such as shown for 1,4c low fluorescence
quantum yield and may light up either upon association with
DNA or by complexation of a fitting metal cation. Moreover,
only the cation complexation should result in a significant
change of the donor–acceptor interaction, such that the shift of
the emission maximum is assumed to be different in the presence
of cations or DNA, hence providing an additional tool for fluori-
metric differentiation. Herein we present the synthesis and
detailed investigation of the properties of 2a, along with the
University of Siegen, Organic Chemistry II, Adolf-Reichwein-Str. 2,
D-57068, Siegen, Germany. E-mail: ihmels@chemie.uni-siegen.de;
Tel: +49 (0) 271 740 3440
†Electronic supplementary information (ESI) available: Fig. 3B, 6A,
and 6B from main manuscript with enlarged insets; fluorimetric titration
of Mg2+ to 2a; Scatchard plots from photometric titrations of DNA to
2a, 2b, and 2c; 1H and 13C NMR spectra of all new compounds. See
DOI: 10.1039/c2ob06948b
3010 | Org. Biomol. Chem., 2012, 10, 3010–3018
This journal is © The Royal Society of Chemistry 2012

View Article Online
comparative studies of reference compounds 2b–e which
revealed surprising results and a design principle for fluorescent
light-up probes.
Fig. 1 Absorption and emission spectra of compounds 2a (black), 2b
(blue) and 2c (red) in CH2Cl2. Continuous lines: absorption spectra, c =
50 μM; dashed lines: normalized fluorescence spectra, c = 10 μM, λex
=
407 nm.
Table 1 Absorption and emission properties of compounds 2a–2c
c
Compound
Solvent
λabsa/nm (log εb)
λF /nm
ΦFd/10−2
e
H2O
412 (4.19)
418 (4.15)
418 (4.31)
420 (4.08)
430 (4.22)
424 (4.17)
413 (4.20)
—
—
—
—
0.63
1.0
0.50
0.40
7.9
5.8
13
e
MeCN
MeOH
DMF
CH2Cl2
CHCl3
Glycerol
e
e
2a
559
544
564
Scheme 1 Synthesis of benzo[b]quinolizinium derivatives 2a–e.
e
H2O
410 (4.13)
417 (4.17)
415 (4.19)
421 (4.03)
427 (4.17)
426 (4.17)
411 (4.13)
—
—
—
—
0.14
0.51
0.50
0.52
7.9
7.1
7.7
e
Results
MeCN
MeOH
DMF
CH2Cl2
CHCl3
Glycerol
e
e
Synthesis
2b
2c
585
557
560
Benzo[b]quinolizinium derivatives 2a–e were synthesized by
Suzuki–Miyaura coupling reactions between 9-benzo[b]quinoli-
zinium boronic acid (3) and bromoarenes (Scheme 1).6 After
counter-anion metathesis, the products were isolated in 15–79%
yield as hexafluorophosphate, perchlorate or tetrafluoroborate
salts. The structures of the new compounds 2a–c were confirmed
by 1H- and 13C-NMR spectroscopic analysis, mass-spectrometric
and elemental analysis data. The known compounds 2d and 2e
were identified by comparison of the NMR spectroscopic data
with the literature data.6
H2O
409 (4.13)
415 (4.01)
411 (4.13)
414 (3.85)
415 (4.19)
415 (4.01)
410 (4.13)
453
459
460
461
506
466
459
0.66
3.6
2.1
1.2
16
18
16
MeCN
MeOH
DMF
CH2Cl2
CHCl3
Glycerol
a Long-wavelength absorption maximum, c = 50 μM. b Extinction
coefficient, in cm−1 M−1 c Fluorescence emission maximum, c =
.
10 μM, λex = 407 nm. d Fluorescence quantum yield relative to
Coumarin 153 (ref. 7). e Too low to be determined.
Photophysical properties
A. Absorption and emission properties. The data of the
absorption and emission properties of 9-arylbenzo[b]quinolizi-
nium derivatives 2a–c are presented in Fig. 1 and Table 1. All
investigated compounds show weak solvatochromic properties,
i.e., depending on the solvent the maxima of the long-
wavelength absorption are located between 409 nm and 430 nm.
The dialkoxy-substituted derivatives 2a and 2b show broad,
structureless and red-shifted absorption bands. The absorption
bands of the meta-methoxy-substituted derivative 2c are slightly
more structured and closely resemble the ones of unsubstituted
derivative 2e.6 Compounds 2a and 2b show very low fluor-
escence quantum yields (ΦF < 0.01), whereas a slightly higher
quantum yield (ΦF = 0.01–0.04) of 2c was observed in highly
polar solvents, such as water, acetonitrile, methanol or DMF.
Notably, the quantum yields of 2a and 2b increase significantly
(ΦF = 0.06–0.18) in chlorinated solvents such as dichloro-
methane or chloroform. The meta-methoxy substituted derivative
2c exhibits a significantly blue-shifted emission spectra (Δλ =
50–90 nm) as compared with the dialkoxy-substituted deriva-
tives 2a and 2b.
B. Viscosity-dependent emission properties. The dependence
of the fluorescence properties of 2a–c on the viscosity of the
medium was studied in glycerol–water mixtures with different
This journal is © The Royal Society of Chemistry 2012
Org. Biomol. Chem., 2012, 10, 3010–3018 | 3011

View Article Online
Fig. 3 Spectrophotometric (A) and spectrofluorimetric titration (B) of
Mg2+ to compound 2a (A: c = 50 μM; B: c = 10 μM, λex = 395 nm) in
MeCN. The arrows indicate the changes of absorption and emission
upon addition of Mg2+. Insets: Plot of the absorption at 430 nm (A) and
emission intensity at 495 nm (B) versus Mg2+ concentration; numerical
fits calculated for K = 1.4 × 105 M−1 from photometric titration.
Fig. 2 A: Fluorescence spectra of compound 2a in water–glycerol
mixtures with different viscosity. The arrow indicates the development
of the emission intensity with increasing the glycerol–water ratio
(0–100%). B: Plot of fluorescence quantum yields of compounds 2a
(●), 2b (▲) and 2c (■) versus viscosity of the solution.
contents of the two components. These mixtures cover a broad
range of viscosities; from η = 1.005 cP in water to 1499 cP in
glycerol at 20 °C. In general, the fluorescence of compounds
2a–c increases significantly, i.e. by a factor of 20 (2a), 55 (2b)
and 24 (2c), with increasing glycerol content of the solution
(Fig. 2, Table 1). In contrast, the unsubstituted derivative 2e and
the p-methoxy-substituted derivative 2d displayed only insignifi-
cant variations of the emission properties upon changing the vis-
cosity of the medium.
same time, the fluorescence intensity of 2a increased upon
addition of Mg2+ by a factor of 450 with an emission maximum
at 495 nm. The binding isotherm from the spectrophotometric
titration was fitted to a 1 : 1 stoichiometry and the resulting
binding constant of 2a-Mg2+ was determined to be K = 1.4 ×
105 M−1. In aqueous solutions an association of Mg2+ ions with
the crown ether unit in 2a was not detected because of the com-
peting hydration.8 Additionally, in aqueous media the lone elec-
tron pairs of the oxygen atoms of the crown ether point towards
the exterior of the cyclic structure, such that a conformational
change of the crown ether to establish a guest coordination is
energetically less favourable.9
Interaction of the crown ether–quinolizinium conjugate 2a
with Mg2+
The interaction of the crown ether derivative 2a with Mg2+ was
investigated by photometric and fluorimetric titrations with Mg
(ClO4)2 in MeCN (Fig. 3). The addition of Mg2+ to 2a induced a
significant blue shift of the broad structureless absorption band
and the development of a well-structured band, which resembles
the one of the 9-phenylbenzo[b]quinolizinium (2e). Simul-
taneously, several isosbestic points (335 nm, 382 nm, 395 nm)
were observed, which slightly faded during the titration. At the
The selectivity of Mg2+-induced light-up effect in 2a was
investigated by the determination of the fluorescence properties
of 2a in the presence of potentially competing cations such as
+
NH4 , Li+, Na+, K+, Ca2+, or Ba2+ (10 equiv., resp.). No signifi-
cant fluorescence enhancement was observed in the presence of
+
NH4 , Na+, or K+, whereas the addition of Li+, Ba2+, and Ca2+
led to an increase of the fluorescence intensity by a factor of 8,
10, and 40 (Fig. 4).
3012 | Org. Biomol. Chem., 2012, 10, 3010–3018
This journal is © The Royal Society of Chemistry 2012

View Article Online
Fig. 4 The fluorescence spectra of 2a (c = 10 μM) upon addition of
Mg2+ (3 equiv.), Li+, Na+, K+, Ca2+, or Ba2+ (10 equiv., resp.).
Interaction of 2a, 2b and 2c with double stranded DNA
Derivatives 2a–c show a similar behavior during photometric
titrations with calf thymus DNA (ct DNA) in BPE buffer
(Fig. 5). In each case, a decrease of the absorbance and a devel-
opment of long-wavelength band were observed. At the same
time, several isosbestic points (348 nm, 364 nm and 426 nm for
2a, 318 nm, 351 nm and 417 nm for 2c) were detected during
the titration of DNA to 2a and 2c. In the case of 2b, the isos-
bestic point at 439 nm faded when the concentration of DNA
was larger than 30 μM (cligand/cDNA < 1.6). In addition, the
development of a long-wavelength absorption with a maximum
at 422 nm was observed. For the evaluation of the DNA-binding
constants and binding-site sizes, the data from spectrophoto-
metric titrations were represented as Scatchard plot and fitted to
the model of McGhee and von Hippel10 to give the binding con-
stant of 2a and 2c with DNA, K = 8.2 × 104 M−1 (2a), 4.0 × 105
M−1 (2c) and the binding site size n = 1.5 (2a), 2.8 (2c). In the
case of 2b, only with cDNA > 0.1 mM (cligand/cDNA < 0.5) the
corresponding Scatchard plot could be fitted satisfactorily with
a binding constant K = 1.6 × 105 M−1 and the binding site size
n = 2.5. Attempts to fit the complete titration data to the model
of McGhee and von Hippel or a two-side model11 failed.
During the spectrofluorimetric titration of ct DNA to 2a or 2b,
an enhancement of the fluorescence intensity at 550 nm or
575 nm by a factor of 15 or 40 was observed (Fig. 6A and 6B).
In contrast, only a slight fluorescence enhancement (2-fold) was
observed upon addition of DNA to 2c (Fig. 6C).
The binding properties of 2a, 2b and 2c with DNA were
further investigated by circular dichroism (CD) spectroscopy
(Fig. 7). Whereas aqueous solutions of the achiral compounds
2a, 2b and 2c alone did not show any CD signals (Fig. 7A, line
a), a positive induced CD (ICD) signal was observed in the
long-wavelength absorption range of the ligand upon addition of
DNA to 2a (Fig. 7A, line b and c). In the case of 2b a bisignate
signal pattern was observed at very high ligand-to-DNA ratio of
5.0 (Fig. 7B, line b), whereas the CD spectrum of 2b gave a
strong positive ICD signal at lower ligand-to-DNA ratio of 0.1
(line c in Fig. 7B). In contrast, the ligand 2c developed a nega-
tive ICD signal upon addition of ct DNA (Fig. 7C).
Fig. 5 Spectrophotometric titrations of 2a (A), 2b (B) and 2c with ct
DNA in BPE buffer (c = 50 μM). The arrows indicate the changes of the
bands upon addition of ct DNA. Insets: Scatchard plots, r/c versus r; r =
ligand-to-DNA ratio, fitted to the model of McGhee and von Hippel.
Discussion
Photophysical properties
The structureless absorption spectra of 2a and 2b and the expan-
sion of the absorption band to >500 nm may be attributed to the
This journal is © The Royal Society of Chemistry 2012
Org. Biomol. Chem., 2012, 10, 3010–3018 | 3013

View Article Online
Fig. 7 A: CD spectra of 2a (A), 2b (B), and 2c (C) in the absence (a)
and presence of DNA; A: c2a/cDNA = 2.0 (b), c2a/cDNA = 0.2 (c). B: c2b
DNA = 5.0 (b), c2b/cDNA = 0.1 (c). C: DNA c2c/cDNA = 1.0 (b), c2c/cDNA
= 0.1 (c).
/
Fig. 6 Spectrofluorimetric titrations of 2a (A, λex = 425 nm), 2b (B,
λex = 438 nm) and 2c (C, λex = 417 nm) with ct DNA in BPE buffer (c
= 10 μM). The arrows indicate the changes of the bands upon addition
c
of ct DNA. Insets: Plot of the relative emission intensity versus cDNA
.
resembling one of the unsubstituted derivative 2e.6 These photo-
physical properties of 2c may be rationalized by the acceptor
property of the m-methoxy group due to its −I effect, as indi-
cated by the Hammett substituent constant (σ = 0.12). Other
than the electron donating p-methoxy group (σ = −0.27)13 a +M
effect is not operative in meta position. The different extent of
donor–acceptor interplay between the electron-rich methoxy-
substituted phenyl substituent and the benzo[b]quinolizinium
unit,6 as commonly observed for donor–acceptor substituted
chromophores.12 In contrast, the absorption spectrum of 2c has a
blue-shifted zero onset (ca. 450 nm) and is more structured, thus
3014 | Org. Biomol. Chem., 2012, 10, 3010–3018
This journal is © The Royal Society of Chemistry 2012

View Article Online
Scheme 2 Resonance structures of p-alkoxy-substituted 9-phenyl-
benzo[b]quinolizinium derivatives.
Presumably the rotation is hindered because a partial double
bond character of the O–phenyl bond is developed due to
excited-state charge transfer, as indicated by a quinoid resonance
structure that represents the electron density distribution in the
excited state (Scheme 2). As shown by the photophysical data,
the m-alkoxy substituent does not act as electron donor, but
rather as electron acceptor, mainly because it is not able to estab-
lish a linear conjugation with the positively charged quinolizi-
nium nitrogen atom, so that the corresponding O–phenyl bond
does not have any double bond character and can freely rotate.
Based on these considerations, we conclude that the rotation of
the m-alkoxy substituent constitutes a major contribution to the
radiationless deactivation of the excited states of 2a–c.
Fig. 8 Viscosity dependence of fluorescence quantum yields for 2a
(●), 2b (▲) and 2c (■). Slopes: k = 0.42 (r2 = 0.99, 2a), 0.54
(r2 = 0.98, 2b) and 0.45 (r2 = 0.99, 2c).
the donor–acceptor interaction in derivatives 2a–c is also
reflected in the emission spectra, i.e. the derivatives 2a and 2b
with the stronger donor substituents exhibit a larger red-shift of
the fluorescence maximum as compared with 2c.
It should be noted, however, that the fluorescence of the 9-(4-
The benzo[b]quinolizinium derivatives 2a–c have very low
emission quantum yields with a remarkable increase in chlori-
nated solvents, such as dichloromethane and chloroform. Such
an enhanced fluorescence quantum yield has been occasionally
observed for cationic dyes and has been attributed to the high
polarizability of these chlorinated solvents.14 Notably, the fluor-
escence quantum yield of 2a–c also increases with increasing
viscosity of the medium, which indicates a radiationless deacti-
vation of the excited state by conformational changes, as has
been demonstrated already for 9-aminobenzo[b]quinolizinium
derivatives, which undergo torsional relaxation about the N–C
(aryl) bond.14a According to the Förster–Hoffmann equation
such a relaxation process in the excited state is usually described
by a linear relationship between ΦF and ηk (eqn (1); with T =
const., ΦF ≪ 1).15
N,N-dimethylamino)phenylaminobenzo[b]quinolizinium
(2f)
has been shown to be quenched by an intramolecular photo-
induced electron transfer (PET) between the excited quinolizi-
nium fluorophore and the aminophenyl donor unit.6 Therefore it
cannot be excluded that a similar PET takes place also in deriva-
tives 2a–c.
Interaction of benzocrown ether-substituted derivative 2a
with Mg2+
ΦF ¼ Cηk
ð1Þ
The formation of a well-structured and blue-shifted absorption
band upon addition of Mg2+ to the solution of 2a in MeCN indi-
cates a significant suppression of the charge transfer from the
electron rich phenyl ring to the benzo[b]quinolizinium chromo-
phore (Fig. 3A). The isosbestic points that were observed during
the photometric titration indicate that only one complex is
formed between 2a and Mg2+. Furthermore, a 1 : 1 stoichiometry
of the complex was supported by the good fit of the titration iso-
therm to the theoretical binding curve (r2 = 0.999). The binding
constant (K = 1.4 × 105 M−1) is comparable to the one of the
known benzo-15-crown-5-Mg2+ complex (K = 4.8 × 104 M−1).17
Notably, the addition of Mg2+ to 2a in MeCN induces a 450-
fold enhancement of the fluorescence intensity (Fig. 3B), which
is significantly blue-shifted (ca. 70 nm) compared with the fluor-
escence of 2a in viscous medium such as glycerol. Thus, the flu-
orescence enhancement most likely originates from the
suppression of the rotation of the m-alkoxy substituent by com-
plexation of the benzocrown ether with Mg2+. As a consequence,
the fluorescence of 2a is blue-shifted because of the disruption
Indeed, in the case of 2a–c, the double-logarithmic plot of the
fluorescence quantum yields versus the viscosity of the medium
is almost linear (Fig. 8). However, the slopes of the plots are sig-
nificantly smaller (2a: k = 0.42; 2b: k = 0.54; and 2c: k = 0.45)
than the ones derived theoretically for the rotation of the phenyl
group (k = 2/3) and obtained experimentally e.g. for di- and
triphenylmethane dyes.16 This deviation from the theoretical
value for the rotation of the phenyl indicates a different or an
additional deactivation pathway for the excited state. In this
regard it is to be noted that the viscosity of the medium does not
influence the emission properties of compounds 2d and 2e.
These observations indicate that the rotation of the phenyl group
about the phenyl–aryl bond alone does not contribute signifi-
cantly to the overall fluorescence quenching. At the same time, a
rotation of the alkoxy substituents about the O–phenyl bond may
be considered as deactivating process; but the marginal influence
of the viscosity on the emission intensity of 2d shows that at
least for para-alkoxy substituents such an effect is not operative.
This journal is © The Royal Society of Chemistry 2012
Org. Biomol. Chem., 2012, 10, 3010–3018 | 3015

View Article Online
of the donor–acceptor system by complexation with Mg2+ ions.
The fluorescence response of 2a towards Mg2+ ions is especially
remarkable considering the tolerance of potentially competing
alkali, alkaline earth cations or NH4+ (Fig. 4). This known selec-
tivity for Mg2+ may be attributed to its relatively larger surface
charge density,18 which was shown in gas phase studies to be the
dominant factor that determines the intrinsic metal ion affinity of
crown ether receptors.19 A similar effect has been found in bi-
arylpyridine derivative.20 In the latter case, however, the addition
of acid efficiently quenches the fluorescence due to the formation
of a TICT state, and the fluorescent enhancement caused by sub-
sequent addition of metal cations is selective towards Mg2+ ions.
form aggregates along the DNA backbone, as frequently
observed for donor–acceptor substituted DNA binders.
During the photometric titration, the behaviour of 2c is similar
to that of 2a (Fig. 5C), i.e. bathochromic and hypochromic
effects, as well as formation of several isosbestic points. More-
over, compound 2c exhibits a stronger interaction with DNA (K
= 4.0 × 105 M−1). The CD spectroscopic analysis, namely the
development of a negative ICD signal, indicates the parallel
alignment of the transition dipole moment of the ligand relative
to the long axis of the binding pocket. The absorption and emis-
sion data of 2c show that the m-phenyl substituent has only a
marginal influence on the electronic situation of benzo[b]quinoli-
zinium core, so that it may be concluded that the transition
dipole moment of 2c resembles the one of the parent benzo[b]-
quinolizinium, i.e. slightly tilted from the short molecule axes.26
Hence, the negative ICD indicates that 2c is also bound to DNA
by a perpendicular intercalation of the molecule into the binding
site.
Interaction of derivatives 2a–c with ds DNA
The photometric titration of compound 2a to DNA exhibits the
characteristics of a DNA-binding process of cationic hetarenes,
i.e. a hypochromic effect of the long-wavelength absorption
band and the formation of red-shifted bands (Fig. 5A).21 The iso-
sbestic points are conserved during the titration, which indicates
one almost exclusive binding mode. The binding constant of 2a
(K = 8.2 × 104 M−1) is comparable to the ones of 9-donor substi-
tuted benzo[b]quinolizinium derivatives with DNA.14a At the
same time, the positive ICD signal (Fig. 7A) is the result of a
perpendicular orientation of the transition dipole moments of the
ligand relative to the DNA binding site.22 In analogy to the
results obtained for a series of annelated quinolizinium deriva-
tives23 this ICD denotes a binding mode with the long molecular
axes of the ligand perpendicular to the long axis of the intercala-
tion binding site, since the transition dipole moment in 2a is
aligned almost parallel to the long molecule axis due to the
donor–acceptor interplay within the chromophore. Overall the
DNA-binding properties of 2a resemble the ones observed for
other cationic hetarenes that exhibit a high affinity towards
nucleic acids.24
The fluorescence intensity of 2a and 2b increases significantly
upon the interaction with DNA (Fig. 6A and 6B). The wave-
lengths of the emission maxima of 2a and 2b in the presence of
DNA are close to the ones of 2a and 2b in glycerol (ΔλF
≈
15 nm). Thus, it may be assumed that the intercalation of 2a and
2b with DNA restricts the conformational freedom of the ligands
and thereby suppresses the main excited-state deactivation
pathway. As the rotation of the 3-alkoxy substituent contributes
significantly to the fluorescence quenching in the unbound
ligand (see above), the enhancement of the fluorescence intensity
of 2a and 2b is proposed to be mainly governed by the suppres-
sion of this torsional relaxation. The enhancement of the fluor-
escence is larger for 2a (40-fold) than for 2b (15-fold). This
difference is consistent with the experimental results obtained
for the viscosity-dependent fluorescence of these compounds
(Table 1). However, only a slight fluorescence enhancement (2-
fold) was observed upon addition of DNA to 2c. Presumably, a
photoinduced electron-transfer (PET) reaction between 2c and
nucleic base pairs quenches the fluorescence in the binding site.
Presumably the PET is more pronounced in the case of 2c,
because the lack of donor–acceptor interplay as compared with
2a and 2b (see above) leads to a higher reduction potential in the
excited state.27
In the case of ligand 2b, the DNA-binding process is more
complicated (Fig. 5B). At lower DNA concentrations (cDNA
≤
30 μM), the change of the absorption bands is essentially identi-
cal to the one of the derivative 2a. In particular, in this range of
ligand-to-DNA ratios one isosbestic point at 439 nm reveals
mainly one binding mode between the DNA and 2b. Neverthe-
less, the loss of the isosbestic point and the development of the
new long-wavelength absorption bands at higher DNA concen-
tration (cDNA > 30 μM), i.e. lower ligand-to DNA ratio, indicate
an additional binding mode between the ligand and DNA.
The two different binding modes were also shown by the
CD-spectroscopic experiments. Thus, at low ligand–DNA ratio
(c2b/cDNA = 0.1) a positive ICD signal was observed which is
similar to the one of 2a, so that a resembling intercalative
binding mode may be deduced. This assumption is supported by
the observation that the binding constants derived from the data
of photometric titration at cLigand/cDNA < 0.5 is comparable to
the one of 2b. At high ligand concentration (c2b/cDNA = 5.0),
however, the CD band develops into a bisignate ICD signal
(Fig. 7B, line b) which is usually the result of exciton coupling
from aggregates of the ligand that associate with the DNA back-
bone.25 Overall, the data from spectrometric titrations of 2b
reveal that this compound intercalates into DNA at low ligand–
DNA ratios r, whereas at higher r values this compound tends to
Conclusions
Although fluorescent probes are known that selectively detect
nucleic acids28 or Mg2+ ions,29 to the best of our knowledge,
there are no reports on a probe molecule that is able to detect
both analytes. Herein, we presented a benzo[b]quinolizinium-
benzo-15-crown-5 ether conjugate 2a that enables the fluori-
metric detection of Mg2+ and DNA by a significant light-up
effect, along with an analyte-specific change of the emission
wavelength. At the same time it is certainly a drawback of the
presented system that due to the commonly known incompatibil-
ity of crown-ether receptors with aqueous media the detection of
the separate analytes requires different solvents. Thus, it remains
a challenge to modify the probe such that both analytes may be
detected in the same medium. Nevertheless, with the present
study it is shown in principle that the combination of the
3016 | Org. Biomol. Chem., 2012, 10, 3010–3018
This journal is © The Royal Society of Chemistry 2012

View Article Online
benzo[b]quinolizinium fluorophore with appropriate receptor
units may be used for the design of dual pathway probes. More-
over, the examination of the light-up mechanism by comparison
with reference compounds 2b and 2c revealed a significant
influence of the position of the methoxy functionality (meta
versus para) on the emission properties of donor–acceptor
systems. Overall, some principles are presented that may be
employed generally for the development of efficient fluorescent
probes.
(m, 1 H, H-2), 8.39–8.63 (m, 4 H, H-8, H-7, H-1, H-10), 9.07
(s, 1 H, H-11), 9.21 (d, J = 6 Hz, 1 H, H-4), 10.33 (s, 1 H,
3
H-6); 13C-NMR (150 MHz, [D6]DMSO): δ = 68.4 (CH2), 68.6
(CH2), 68.7 (CH2), 68.8 (CH2), 69.6 (CH2), 69.7 (CH2), 70.4
(CH2), 70.5 (CH2), 112.8 (CHar), 113.8 (CHar), 120.9 (CHar),
121.9 (Cq), 122.2 (CHar), 124.0 (Cq), 124.7 (CHar), 126.7
(CHar), 128.6 (Cq), 130.1 (CHar), 130.4 (CHar), 130.9 (CHar),
134.2 (CHar), 135.9 (CHar), 137.6 (Cq), 139.6 (CHar), 145.0
(Cq), 149.0 (Cq), 150.2 (Cq); MS (ESI+): m/z (%) = 446 (100)
[M]+; elemental analysis calcd (%) for C27H28F6NO5P (591.48):
C, 54.83; H, 4.77; N, 2.37; found: C, 54.55; H, 4.32; N, 2.48.
Experimental
9-(3,4-Dimethoxyphenyl)benzo[b]quinolizinium
hexafluoro-
General instrumentations and materials
phosphate (2b). Yellow powder, yield 175 mg (38%); m.
1
p. 263–265 °C (dec.); H-NMR (600 MHz, [D6]DMSO): δ =
All commercially available chemicals were reagent grade and
used without further purification. The melting point was deter-
mined with a Büchi 510K melting point apparatus and was not
corrected. Mass spectra (ESI in the positive-ion mode, source
voltage 6 kV) were recorded with a Finnigan LCQ Deca instru-
ment; only m/z values in the range of 100–2000 units were ana-
lyzed. NMR spectra were measured on Varian NMR System 600
(1H: 600 MHz, 13C: 150 MHz) spectrometer at 20 °C; chemical
shifts are given in ppm (δ) relative to TMS (δ = 0.00 ppm).
Unambiguous proton NMR assignments were established by
{1H, 1H}-COSY, HSQC and HMBC experiments. Elemental
microanalysis was performed with a HEKAtech EuroEA com-
bustion analyzer by Mr. H. Bodenstedt (Organische Chemie I,
Universität Siegen). Purified water with resistivity ≥18 MΩ
cm−1 was used for spectrometric measurements. 9-Benzo[b]-
quinolizinium boronic acid (3), 9-(p-methoxyphenyl)benzo[b]-
quinolizinium (2d) and 9-phenylbenzo[b]quinolizinium (2e)
were prepared according to the literature procedure.6
3.86 (s, 3 H, CH3), 3.93 (s, 3 H, CH3), 7.17–7.18 (m, 1 H,
H-Ph), 7.56 (s, 1 H, H-Ph), 7.58–7.79 (m, 1 H, H-Ph), 7.89 (dd,
3
3J = 7 Hz, 3J = 7 Hz, 1 H, H-3), 8.02 (dd, J = 7 Hz, 3J = 7 Hz,
3
3
1 H, H-2), 8.41 (d, J = 9 Hz, 1 H, H-8), 8.50 (d, J = 9 Hz, 1
H, H-7), 8.54 (d, 3J = 8 Hz, 1 H, H-1), 8.64 (s, 1 H, H-10), 9.10
(s, 1 H, H-11), 9.21 (d, J = 7 Hz, 1 H, H-4), 10.34 (s, 1 H,
3
H-6); 13C-NMR (150 MHz, [D6]DMSO): δ = 55.6 (CH3), 55.8
(CH3), 110.9 (CHar), 112.3 (CHar), 120.6 (CHar), 121.8 (CHar),
122.2 (CHar), 124.0 (CHar), 124.7 (Cq), 126.6 (CHar), 128.5
(CHar), 129.9 (Cq), 130.3 (CHar), 130.9 (CHar), 134.2 (CHar),
135.9 (Cq), 137.6 (Cq), 139.6 (CHar), 145.1 (Cq), 149.4 (Cq),
150.4 (Cq); MS (ESI+): m/z (%) = 316 (100) [M]+; elemental
analysis calcd (%) for C21H18F6NO2P (461.34): C, 54.67; H,
3.93; N, 3.04; found: C, 54.66; H, 3.69; N, 3.24.
9-(m-Methoxyphenyl)benzo[b]quinolizinium tetrafluoroborate
(2c). Yellow powder, yield 116 mg (31%); m.p. 243–245 °C
(dec.); 1H-NMR (600 MHz, [D6]DMSO): δ = 3.95 (s, 3 H,
CH3), 7.17–7.19 (m, 1 H, H-Ph), 7.57–7.60 (m, 3 H, H-Ph),
3
3
7.97–8.00 (m, 1 H, H-3), 8.10 (dd, J = 8 Hz, J = 8 Hz, 1 H,
H-2), 8.42 (dd, 3J = 9 Hz, 4J = 2 Hz, 1 H, H-8), 8.59–8.64 (m, 2
H, H-7, H-1), 8.76 (s, 1 H, H-10), 9.22 (s, 1 H, H-11), 9.30 (d,
3J = 7 Hz, 1 H, H-4), 10.45 (s, 1 H, H-6); 13C-NMR (150 MHz,
[D6]DMSO): δ = 55.8 (CH3), 113.5 (CHar), 115.6 (CHar), 120.4
(CHar), 122.6 (CHar), 124.2 (Cq), 125.0 (CHar), 125.5 (CHar),
127.2 (CHar), 129.2 (CHar), 131.0 (2 CHar), 131.5 (CHar), 134.7
(CHar), 136.1 (Cq), 138.1 (Cq), 139.6 (Cq), 140.3 (CHar), 145.5
(Cq), 160.4 (Cq); MS (ESI+): m/z (%) = 286 (100) [M]+; elemen-
tal analysis calcd (%) for C20H16BF4NO (373.15): C, 64.37; H,
4.32; N, 3.75; found: C, 63.79; H, 4.13; N, 3.94.
Synthesis
General procedure for the synthesis of 9-arylbenzo[b]quinoli-
zinium derivatives. Under inert-gas atmosphere, a solution of 36
(303 mg, 1.00 mmol), the corresponding aryl bromide
(1.50 mmol), Pd(dppf)Cl2·CH2Cl2 (40.8 mg, 0.05 mmol) and
KF (232 mg, 4.00 mmol) in DME–MeOH–H2O (12 ml; 2 : 1 : 1)
was stirred under reflux for 24 h (2a), 7 h (2b) or 12 h (2c).
After cooling to r.t., MeOH (10 ml) was added to the reaction
mixture and the formed precipitate was removed by filtration. A
saturated aqueous solution of NaPF6 or NaClO4 (5 ml) was
added to the filtrate until no more precipitation was observed.
The yellow precipitate was collected, washed with water, EtOAc,
and diethyl ether. The analytically pure product was separated by
recrystallization of the precipitate from acetonitrile–ethyl acetate
or by column chromatography (Al2O3, Activity I, eluent:
MeCN) and subsequent recrystallization from acetonitrile–ethyl
acetate.
Spectrophotometric measurements
Absorption spectra were recorded on a Varian Cary 100 double-
beam spectrophotometer; emission spectra were recorded on a
Varian Cary Eclipse fluorescence spectrophotometer. All spectro-
photometric measurements were performed in thermostated
quartz sample cells at 20 °C. Solutions for analysis were pre-
pared by dilution of stock solutions (1.0 × 10−3 M in water)
immediately before the experiments. The solution concentrations
were 50 μM for absorption and CD spectroscopy and 10 μM for
fluorescence spectroscopy. Spectrophotometer slit widths were
kept at 2 nm for absorption spectroscopy and 5/5 nm for emis-
sion spectroscopy. Titrations with Mg2+ or DNA: The solution
9-(2,3,5,6,8,9,11,12-Octahydro-1,4,7,10,13-benzopentaoxacyclo-
pentadecin-15-yl)-aminobenzo[b]quinolizinium
hexafluoropho-
sphate (2a). Yellow powder, yield 76 mg (15%); m.
p. 113–114 °C (dec.); 1H-NMR (600 MHz, [D6]DMSO): δ =
3.64–3.67 (m, 8 H, CH2), 3.81–3.85 (m, 4 H, CH2), 4.15–4.25
3
(m, 4 H, CH2), 7.15 (d, J = 6 Hz, 1 H, H-Ph), 7.56–7.58 (m, 2
3
3
H, H-Ph), 7.89 (dd, J = 6 Hz, J = 6 Hz, 1 H, H-3), 8.02–8.03
This journal is © The Royal Society of Chemistry 2012
Org. Biomol. Chem., 2012, 10, 3010–3018 | 3017

View Article Online
12 (a) J. N. Murrell, The Theory of the Electronic Spectra of Organic Mol-
ecules, John Wiley and Sons, New York, 1963; (b) R. Gompper and
H.-U. Wagner, Angew. Chem., Int. Ed. Engl., 1988, 27, 1437.
13 C. Hansch, C. Leo and R. W. Taft, Chem. Rev., 1991, 91, 165.
14 (a) A. Granzhan, H. Ihmels and G. Viola, J. Am. Chem. Soc., 2007, 129,
1254; (b) O. van den Berg, W. F. Jager and S. J. Picken, J. Org. Chem.,
2006, 71, 2666; (c) W. F. Jager, T. S. Hammink, O. van den Berg and
F. C. Grozema, J. Org. Chem., 2010, 75, 2169; (d) L. Wu and
K. Burgess, J. Org. Chem., 2008, 73, 8711.
15 T. Förster and G. Hoffmann, Z. Phys. Chem., 1971, 75, 63.
16 (a) G. Oster and Y. Nishijima, J. Am. Chem. Soc., 1956, 78, 1581;
(b) M. Vogel and W. Rettig, Ber. Bunsen-Ges. Phys. Chem., 1985, 89,
962; (c) P. Gautam and A. Harriman, J. Chem. Soc., Faraday Trans.,
1994, 90, 697.
of the titrants Mg(ClO4)2 or ct DNA contained the appropriate
concentration of the compound (to avoid dilution effects) and
were added to a cuvette containing a solution of the compound
in an appropriate buffer. The titration was monitored by absorp-
tion or emission spectroscopy. The titration was continued until
no further changes in the spectrum were observed. All spectro-
photometric measurements were performed at least three times to
ensure reproducibility. Binding constants were determined by
fitting the experimental binding isotherms to theoretical models
according established procedures.21a,30 to Circular dichroism
(CD) spectra were recorded on a Chirascan (Applied Photophy-
sics Limited, UK) apparatus.
17 C. Erk and A. Gocmen, Talanta, 2000, 53, 137.
18 J. P. Geue, N. J. Head, A. D. Ward and S. F. Lincoln, Dalton Trans.,
2003, 521.
19 M. B. More, D. Ray and P. B. Armentrout, J. Am. Chem. Soc., 1999, 121,
417.
20 Y. Liu, M. Han, H.-Y. Zhang, L.-X. Yang and W. Jiang, Org. Lett., 2008,
10, 2873.
21 (a) D. E. Graves, Drug–DNA interactions, in Methods in Molecular
Biology, Vol. 95: DNA Topoisomerase Protocols: Volume II: Enzymology
and Drugs, ed. N. Osheroff and M. A. Bjornsti, Humana Press, Totowa,
NJ, 2001, pp. 161–169; (b) H. Ihmels and D. Otto, Top. Curr. Chem.,
2005, 258, 161.
Acknowledgements
Generous support by the Deutsche Forschungsgemeinschaft
(FOR516) is gratefully acknowledged. We thank Ms. Laura
Thomas for assistance with the determination of CD spectra.
22 B. Nordén and T. Kurucsev, J. Mol. Recognit., 1994, 7, 141.
23 H. Ihmels, K. Faulhaber, D. Vedaldi, F. Dall’Acqua and G. Viola, Photo-
chem. Photobiol., 2005, 81, 1107.
Notes and references
24 For most recent examples, see e.g.: (a) R. B. P. Elmes, M. Erby,
S. M. Cloonan, S. J. Quinn, D. C. Williams and T. Gunnlaugsson, Chem.
Commun., 2011, 47, 686; (b) A. Mazzoli, B. Carlotti, C. Bonaccorso,
C. G. Fortuna, U. Mazzucato, G. Miolo and A. Spalletti, Photochem.
Photobiol. Sci., 2011, 10, 1830; (c) Z. Miskolczy, M. Megyesi, L. Biczók
and Helmut Görner, Photochem. Photobiol. Sci., 2011, 10, 592;
(d) T. Phillips, I. Haq and J. A. Thomas, Org. Biomol. Chem., 2011, 9,
3462.
25 H. Ihmels, K. Faulhaber, C. Sturm, G. Bringmann, K. Messer,
N. Gabellini, D. Vedaldi and G. Viola, Photochem. Photobiol., 2001, 74,
505.
26 H. Ihmels, K. Faulhaber, B. Engels and C. Lennartz, Chem.–Eur. J.,
2000, 6, 2854.
1 (a) J. F. Zhang, Y. Zhou, J. Yoon and J. S. Kim, Chem. Soc. Rev., 2011,
40, 3416; (b) J. Wu, W. Liu, J. Ge, H. Zhang and P. Wang, Chem. Soc.
Rev., 2011, 40, 3483; (c) M. E. Moragues, R. Martinez-Manez and
F. Sancenon, Chem. Soc. Rev., 2011, 40, 2593; (d) R. M. Duke,
E. B. Veale, F. M. Pfeffer, P. E. Kruger and T. Gunnlaugsson, Chem. Soc.
Rev., 2010, 39, 3936; (e) M. Y. Berezin and S. Achilefu, Chem. Rev.,
2010, 110, 2641; (f) J. Han and K. Burgess, Chem. Rev., 2010, 110,
2709; (g) H. Kobayashi, M. Ogawa, R. Alford, P. L. Choyke and
Y. Urano, Chem. Rev., 2010, 110, 2620; (h) X. Qian, Y. Xiao, Y. Xu,
X. Guo, J. Qiana and W. Zhua, Chem. Commun., 2010, 46, 6418;
(i) R. W. Sinkeldam, N. J. Greco and Y. Tor, Chem. Rev., 2010, 110,
2579; ( j) A. P. Demchenko, Introduction to Fluorescence Sensing,
Springer, Berlin, 2009; (k) C. McDonagh, C. S. Burke and
B. D. MacCraith, Chem. Rev., 2008, 108, 400; (l) A. P. de Silva,
T. P. Vance, M. E. S. West and Glenn D. Wright, Org. Biomol. Chem.,
2008, 6, 2468; (m) S. R. Adams, M. H. Ellisman and R. Y. Tsien,
Science, 2006, 312, 217; (n) L. Zhu and E. V. Anslyn, Angew. Chem., Int.
Ed., 2006, 45, 1190; (o) L. Basabe-Desmonts, D. N. Reinhoudt and
M. Crego-Calama, Chem. Soc. Rev., 2007, 36, 993.
27 C. Bohne, K. Faulhaber, B. Giese, A. Häfner, A. Hofmann, H. Ihmels,
A.-K. Köhler, S. Perä, F. Schneider and M. A. L. Sheepwash, J. Am.
Chem. Soc., 2005, 127, 76.
28 For recent exemplary examples, see: (a) B. A. D. Neto and
A. A. M. Lapis, Molecules, 2009, 14, 1725; (b) T. Mahmood, A. Paul
and S. Ladame, J. Org. Chem., 2010, 75, 204; (c) M. R. Gill, H. Derrat,
C. G. W. Smythe, G. Battaglia and J. A. Thomas, ChemBioChem, 2011,
12, 877; (d) J. Huang, Y. Wu, Y. Chen, Z. Zhu, X. Yang, C. J. Yang,
K. Wang and W. Tan, Angew. Chem., Int. Ed., 2011, 50, 401;
(e) L. J. Kricka and P. Fortina, Clin. Chem., 2009, 55, 670;
(f) O. Vázquez, M. I. Sánchez, J. Martínez-Costas, M. E. Vázquez and
J. L. Mascareñas, Org. Lett., 2010, 12, 216; (g) E. B. Veale and
T. Gunnlaugsson, J. Org. Chem., 2010, 75, 5513; and references cited
therein.
29 For recent exemplary examples, see: (a) Y. Dong, J. Li, X. Jiang, F. Song,
Y. Cheng and C. Zhu, Org. Lett., 2011, 13, 2252; (b) J. Brandel,
M. Sairenji, K. Ichikawa and T. Nabeshima, Chem. Commun., 2010, 46,
3958; (c) M. Ishida, Y. Naruta and Fumito Tani, Angew. Chem., Int. Ed.,
2010, 122, 95; (d) Y. Liu, M. Han, H.-Y. Zhang, L.-X. Yang and
Wei Jiang, Org. Lett., 2008, 10, 2873; (e) H. M. Kim, P. R. Yang,
M. S. Seo, J. S. Yi, J. H. Hong, S.-J. Jeon, Y.-G. Ko, K. J. Lee and
B. R. Cho, J. Org. Chem., 2007, 72, 2088; (f) D. Ray and
P. K. Bharadwaj, Inorg. Chem., 2008, 47, 2252; and references cited
therein.
2 (a) L. Zhu, L. Zhang and A. H. Younes, Supramol. Chem., 2009, 21,
268; (b) H.-H. Liu and Y. Chen, Eur. J. Org. Chem., 2009, 5261;
(c) M. Schmittel and H.-W. Lin, Angew. Chem., 2007, 119, 911;
M. Schmittel and H.-W. Lin, Angew. Chem., Int. Ed., 2007, 46, 893
(d) Y. Q. Li, J. L. Bricks, U. Resch-Genger, M. Spieles and W. Rettig,
J. Phys. Chem. A, 2006, 110, 10972.
3 X. He and V. W.-W. Yam, Org. Lett., 2011, 13, 2172.
4 (a) M. Tian and H. Ihmels, Chem. Commun., 2009, 3175; (b) M. Tian,
H. Ihmels and E. Brötz, Dalton Trans., 2010, 39, 8195; (c) M. Tian and
H. Ihmels, Eur. J. Org. Chem., 2011, 4145.
5 M. Tian, H. Ihmels and K. Benner, Chem. Commun., 2010, 46, 5719.
6 M. Tian and H. Ihmels, Synthesis, 2009, 4226.
7 G. Jones, W. R. Jackson, C. Y. Choi and W. R. Bergmark, J. Phys.
Chem., 1985, 89, 294.
8 C. J. Pedersen and H. K. Frensdorff, Angew. Chem., Int. Ed. Engl., 1972,
11, 16.
9 A. Späth and B. König, Beilstein J. Org.Chem., 2010, 6, 1.
10 J. D. McGhee and P. H. Von Hippel, J. Mol. Biol., 1974, 86, 469.
11 K. Bonjean, M. C. De Pauw-Gillet, M. P. Defresne, P. Colson,
C. Houssier, L. Dassonneville, C. Bailly, R. Greimers, C. Wright,
J. Quetin-Leclercq, M. Tits and L. Angenot, Biochemistry, 1998, 37,
5136.
30 (a) B. Valeur, Molecular Fluorescence: Principles and Applications,
Wiley-VCH, Weinheim, 2002; (b) J. Polster and H. Lachmann, Spectro-
metric Titrations: Analysis of Chemical Equilibria, VCH, Weinheim,
1989.
3018 | Org. Biomol. Chem., 2012, 10, 3010–3018
This journal is © The Royal Society of Chemistry 2012