Fluorescent sensors for diamines

Article abstract of DOI:10.1039/b503269e

A series of seven diamine sensors was prepared using dimers of a quinolone aldehyde chromophore. Binding to six different diamine guests was explored by a combination of NMR, absorption and fluorescence spectroscopy. It was shown that the dimeric sensors bound the diamine guests by formation of a bis-iminium ion which produced large changes in the fluorescence of the quinolone core. Issues of selectivity between guests are discussed. The Royal Society of Chemistry 2005.

Full text of DOI:10.1039/b503269e

View Article Online / Journal Homepage / Table of Contents for this issue
PAPER
www.rsc.org/materials | Journal of Materials Chemistry
Fluorescent sensors for diamines{
Kristen Secor, Jeffrey Plante, Christopher Avetta and Timothy Glass*
Received 9th March 2005, Accepted 18th July 2005
First published as an Advance Article on the web 12th August 2005
DOI: 10.1039/b503269e
A series of seven diamine sensors was prepared using dimers of a quinolone aldehyde
chromophore. Binding to six different diamine guests was explored by a combination of NMR,
absorption and fluorescence spectroscopy. It was shown that the dimeric sensors bound the
diamine guests by formation of a bis-iminium ion which produced large changes in the
fluorescence of the quinolone core. Issues of selectivity between guests are discussed.
synthesis and characterization of sensors such as 2 and their
Introduction
use in the fluorescent detection of diamines.
Fluorescent sensors for intracellular analyte detection are
rapidly becoming common biochemical tools. Sensors for
cellular ions such as Ca+2 and Zn+2 are particularly well
studied.1 Sensors for the intracellular detection of organic
analytes are less well developed due to the difficulty in
obtaining selective and high affinity recognition of organic
compounds in an aqueous environment.2,3 We have been
Experimental
Absorption and fluorescence experiments
Fluorescence spectra were recorded on a Shimadzu RF-5301
PC spectrofluorimeter at 37 uC. The excitation wavelength was
495 nm with excitation with emission slit widths of 5 nm each.
Absorption spectra were recorded on a Cary 1E spectro-
photometer at 37 uC. Solutions of each sensor were prepared at
10 mM in 50% MeOH–buffered water (50 mM HEPES,
240 mM NaCl, pH 5 7.4 prior to dilution in MeOH). All guest
solutions were prepared as above and contained 10 mM senor
to prevent dilution. Binding isotherms were constructed by
recording fluorescence intensity at 537 nm as a function of
added analyte and fitting the data to a one-site model using
Graphpad Prism software. A 1 : 1 binding model was assumed
(verified by NMR) and the data fit this model well for all
investigating
a class of coumarin-containing fluorescent
sensors for the specific sensing of amine-containing analytes
(compound 1, Scheme 1).4 Compounds like 1 respond to the
presence of amines by the formation of an iminium ion which,
due to fortuitous hydrogen bonding, alters the fluorescent
properties of the coumarin core. Importantly, this recognition
and sensing method works well in an aqueous environment
and gives a good fluorescent response to analyte binding.
By attaching auxiliary recognition elements at the R group,
selective sensors for amine containing analytes such as
dopamine have been prepared.5
analytes tested. The fluorescence intensity at saturation (Isat
)
Zimmerman et al. have prepared receptors for diamines by
incorporating trifluoromethyl ketones into a dendrimer.6 As
part of our program in amine sensing, we began a study in
which dimeric versions of 1 could be used to bind diamines
such as lysine.7 During this study, we discovered that the
coumarin core of 1 could be replaced by a quinolone group
and that quinolone dimers such as 2 (Scheme 2) were quite
synthetically accessible. Thus, a series of sensors could be
prepared and the linking group (X) could be varied in order to
achieve optimal binding properties. Herein, we report the
was taken from the theoretical fit to the data.
Synthesis
Compound 5. Compound 4 (4.39 g, 19.4 mmol), bis-(2,4,6-
trichlorophenyl) malonate (8.98 g, 19.4 mmol), and toluene
(150 cm3) were added to a flame-dried flask. A reflux
condenser and drying tube were attached, and the mixture
was refluxed for 20 h. The reaction mixture was cooled and
filtered. The filtrate was washed with hexanes to yield the title
compound as a tan powder (5.8 g, 100%). mp 284–286 uC,
Scheme 1
125 Chemistry, 601 S. College Ave., University of Missouri Columbia,
MO, USA. E-mail: glasst@missouri.edu; Fax: (+1) 573 882 2754;
Tel: (+1) 573 882 3813
{ This paper was invited for the guest editor issue of Journal of
Materials Chemistry on Fluorescent Sensors. See J. Mater. Chem.,
2005, 15(27).
Scheme 2
This journal is ß The Royal Society of Chemistry 2005
J. Mater. Chem., 2005, 15, 4073–4077 | 4073

View Article Online
decomp; nmax(film)/cm21 1625, 1569, 1544, 1537, 1488, 1384
and 1354; dH/ppm (250 MHz, DMSO-d6) 11.0 (1 H, s), 7.65
(1 H, d, J 9.0), 7.19–7.28 (5 H, m), 6.59 (1 H, dd, J 2.0 and 9.0),
6.29 (1 H, d, J 1.9), 5.69 (1 H, s), 5.40 (2 H, s) and 2.86 (6 H, s);
dC/ppm (62.5 MHz, DMSO-d6) 163.5, 161.9, 152.0, 140.8,
137.9, 128.4, 126.8, 126.7, 124.1, 107.1, 105.9, 96.2, 93.4, 43.9
and 39.6; m/z (FAB) 301.1529 (M + Li+. C18H18LiN2O2
requires 301.1528).
An analytically pure sample was obtained by recrystallization
from toluene. mp 92 uC; nmax(film)/cm21 2924, 1695, 1597,
1497, 1390 and 1170; dH/ppm (300 MHz, CDCl3, Me4Si) 10.55
(2H, s), 8.27 (2H, d, J 9.3 Hz), 7.33–7.23 (10H, m), 7.04 (4H,
m), 6.62 (2H, dd, J 2.3, 9.4 Hz), 6.21 (2H, d, J 2.3 Hz), 5.50
(4H, s), 4.09 (4H, s) and 2.98 (12H, s); dC/ppm (75 MHz
CDCl3; Me4Si) 190.5, 162.3, 153.6, 153.4, 142.1, 137.3, 136.4,
131.1, 129.7, 128.8, 128.6, 128.2, 127.3, 126.8, 120.7, 112.6,
109.2, 95.7, 46.2, 42.0 and 40.1; m/z (FAB) 785.2795 (M + Li+.
C46H42N4O4S2Li requires 785.2795).
Compound 6. POCl3 (6.2 cm3, 66.5 mmol) was added to
DMF (13 cm3, 168 mmol) at 0 uC in a dry flask equipped with
a drying tube. The solution was stirred at 0 uC for 15 min then
at ambient temperature for 3.5 h. DMF (250 cm3) then
compound 5 (13.8 g, 46.9 mmol) were added. The mixture was
stirred at ambient temperature for 42 h. The resulting yellow
suspension was poured onto H2O (800 cm3) and filtered. The
filtrate was washed with MeOH (800 cm3). The resulting
yellow solid was 90% pure and was used in subsequent
reactions. A portion of the material was purified via flash
chromatography (Et2O–CH2Cl2, 0 : 100 to 10 : 90) for
characterization. mp 225u–226u; nmax(film)/cm21 1682, 1637,
1611, 1583, 1558, 1514, 1496 and 1454; dH/ppm (300 MHz,
CDCl3; Me4Si) 10.55 (1 H, s), 8.00 (1 H, d, J 9.4), 7.24–7.34
(5 H, m), 6.64 (1 H, dd, J 2.3 and 9.4), 6.20 (1 H, d, J 2.3), 5.49
(2 H, s) and 3.00 (6 H, s); dC/ppm (75 MHz, CDCl3; Me4Si)
189.5, 162.3, 153.8, 148.4, 142.3, 136.0, 134.2, 129.8, 128.9,
127.5, 126.8, 116.0, 109.6, 95.3, 46.2 and 40.1; m/z (FAB)
347.1144 (M + Li+. C19H17ClLiN2O2 requires 347.1139).
Compound 2c. The crude reaction mixture was poured into
H2O (200 cm3) and filtered. The filtrate was washed into a
flask with CH2Cl2 (100 cm3), dried over Na2SO4, and the
solvent was removed in vacuo. The yellow solid was purified
via flash chromatography (THF–hex, 20 : 80 to 50 : 50), and
the product was isolated as a yellow solid (31.4 mg, 45%). An
analytically pure sample was obtained by recrystallization
from toluene. mp 109 uC; nmax(film)/cm21 2924, 1633, 1597,
1498, 1391 and 1170; dH/ppm (300 MHz, CDCl3, Me4Si) 10.46
(2 H, s), 8.26 (2H, d, J 9.2), 7.33–7.24 (10 H, m), 7.07 (4 H, s),
6.58 (2 H, dd, J 2.3 and 9.4), 6.19 (2 H, d, J 2.3), 5.51 (4 H, s),
4.33 (4 H, s) and 2.97 (12 H, s); dC/ppm (75 MHz CDCl3;
Me4Si) 190.4, 162.3, 153.6, 153.1, 142.0, 136.6, 135.5, 130.9,
128.8, 128.0, 127.3, 126.7, 120.5, 112.3, 109.2, 95.7, 46.0, 40.0
and 39.4; m/z (FAB) 785.2807 (M + Li+. C46H42N4O4S2Li
requires 785.2795).
Compound 2d. The solvent from the crude reaction mixture
was removed in vacuo. The resulting yellow solid was dissolved
in H2O (10 cm3) and CHCl3 (30 cm3) and extracted with
CHCl3 (5 6 30 cm3). The combined organic extracts were
dried over MgSO4. The solvent was removed in vacuo, and
the yellow solid was purified via flash chromatography (THF–
hexanes, 20 : 80 to 100 : 0), and the product was isolated as
a yellow solid (91 mg, 91%). mp 224–226 uC, decomp;
General procedure for preparation of dimers 2a–g.
Compound 6 (2.2 equiv.), dithiol (1 equiv.), K2CO3 (4 equiv.),
and DMF (0.015 M in dithiol) were combined in a flame-
dried, N2-filled flask. The reaction mixture was stirred at
ambient temperature overnight, then worked up according to
the procedures below.
Compound 2a. The crude reaction mixture was poured into
H2O (200 cm3) and filtered. The filtrate was washed into a
flask with CH2Cl2 (100 cm3), dried over Na2SO4, and the
solvent was removed in vacuo. The yellow solid was purified
via flash chromatography (THF–hex, 20 : 80 to 50 : 50), and
the product was isolated as a yellow solid (47.6 mg, 76%). An
analytically pure sample was obtained by recrystallization
from toluene. mp 121 uC; nmax(film)/cm21 2924, 1695, 1597,
1498, 1391 and 1170; dH/ppm (300 MHz, CDCl3, Me4Si) 10.54
(2 H, s), 8.28 (2 H, d, J 9.3), 7.33–7.23 (10 H, m), 7.06 (4 H, s),
6.60 (2 H, dd, J 2.2 and 9.3), 6.21 (2 H, d, J 2.1), 5.50 (4 H, s),
4.13 (4 H, s) and 2.99 (12 H, s); dC/ppm (75 MHz CDCl3;
Me4Si) 190.5, 162.3, 153.6, 153.4, 142.0, 136.3, 131.0, 129.1,
128.8, 127.3, 126.8, 120.6, 112.5, 109.1, 95.7, 46.1, 41.9 and
40.1; m/z (FAB) 785.2807 (M + Li+. C46H42N4O4S2Li requires
785.2795).
n
max(film)/cm21 2927, 1694, 1598, 1503, 1390, 1364, 1164 and
1169; dH/ppm (300 MHz, CDCl3; Me4Si) 10.48 (2 H, s), 7.96
(2 H, d, J 9.4), 7.25–7.32 (10 H, m), 6.99–7.12 (4 H, m), 6.52
(2 H, dd, J 2.4 and 9.4), 6.23 (2 H, d, J 2.4), 5.55 (4 H, s) and
2.97 (12 H, s); dC/ppm (75 MHz, CDCl3; Me4Si) 190.1, 161.7,
153.5, 150.6, 142.5, 138.0, 136.3, 131.5, 129.9, 128.8, 128.8,
127.4, 127.2, 126.8, 121.2, 110.9, 109.4, 95.7, 46.4 and 40.0;
m/z (FAB) 357.2516 (M + Li+. C44H38LiN4O4S2 requires
357.2495).
Compound 2e. The crude reaction mixture was poured onto
H2O (20 cm3) and extracted with CH2Cl2 (5 6 30 cm3). The
combined organic extracts were dried on MgSO4, and the
solvent was removed in vacuo. The resulting red–brown solid
was purified via flash chromatography (EtOAc–CH2Cl2, 0 : 100
to 20 : 80), and the product was isolated as a yellow solid
(65.1 mg, 43%). mp 263–264 uC, decomp; nmax(film)/cm21
2924, 1694, 1598, 1495, 1483, 1434, 1389, 1363 and 1171;
dH/ppm (250 MHz, CDCl3; Me4Si) 10.53 (2 H, s), 8.10 (2 H, d,
J 9.4), 7.22–7.30 (10 H, m), 6.95–7.05 (4 H, m), 6.55 (4 H, dd, J
2.2 and 9.4), 6.24 (2 H, d, J 2.2), 5.55 (4 H, s) and 2.96 (12 H,
s); dC/ppm (75 MHz, CDCl3; Me4Si) 190.0, 161.6, 153.6, 151.7,
142.5, 136.3, 131.2, 130.4, 128.8, 127.7, 127.4, 126.9, 121.4,
Compound 2b. The crude reaction mixture was poured into
H2O (200 cm3) and filtered. The filtrate was washed into a
flask with CH2Cl2 (100 cm3), dried over Na2SO4, and the
solvent was removed in vacuo. The yellow solid was purified
via flash chromatography (THF–hex, 20 : 80 to 50 : 50), and
the product was isolated as a yellow solid (45.6 mg, 57%).
4074 | J. Mater. Chem., 2005, 15, 4073–4077
This journal is ß The Royal Society of Chemistry 2005

View Article Online
111.4, 109.5, 95.7, 46.4 and 40.7; m/z (FAB) 357.2514 (M + Li+.
C44H38LiN4O4S2 requires 357.2495).
Compound 2f. The crude reaction mixture was diluted with
H2O (30 cm3) and extracted with CH2Cl2 (2 6 40 cm3). The
combined organic extracts were dried on Na2SO4, and the
solvent was removed in vacuo. The resulting red–brown solid
was purified via flash chromatography (EtOAc–CH2Cl2, 1 : 3
to 1 : 1), and the product was isolated as a yellow solid (53 mg,
72%). mp 226 uC; nmax(film)/cm21 2924, 1684, 1596, 1497, 1390
and 1170; dH/ppm (500 MHz, CDCl3, Me4Si), 10.53 (2 H, s),
8.29 (2 H, d, J 9.4), 7.27 (12 H, m), 6.63 (2 H, dd, J 2.1 and
9.4), 6.20 (2 H, d, J 2.0), 5.49 (4 H, s),3.10 (4 H, t, J 6.9), 2.98
(12 H, s), 1.54(4 H, m) and 1.46 (2 H, m); dC/ppm (125 MHz
CDCl3; Me4Si) 191.1, 163.2, 155.1, 154.5, 142.8, 137.1, 131.6,
129.6, 128.1, 127.6, 120.9, 113.3, 110.1, 96.5, 47.0, 40.9, 37.7
and 31.0; m/z (FAB) 723.2669 (M + Li+. C41H40N4O4S2Li
requires 723.2651).
Compound 2g. The crude reaction mixture was diluted with
CH2Cl2 (10 cm3), filtered on a plug of cotton, and the solvent
was removed in vacuo to yield a red–yellow solid (55 mg,
100%). mp 122 uC; nmax(film)/cm21 2929, 2858, 1692, 1597,
1497, 1391 and 1170; dH/ppm (300 MHz, CDCl3, Me4Si) 10.59
(2 H, s), 8.34 (2 H, d, J 9.3), 7.26 (12 H, m), 6.65 (2 H, dd, J 2.3
and 9.3), 6.22 (2 H, d, J 2.2), 5.50 (4 H, s), 2.99 (12 H, m), 2.95
(4 H, t, J 7.1), 1.55 (4 H, p, J 6.8) and 1.46(2 H, m); dC/ppm
(75 MHz CDCl3; Me4Si) 190.4, 162.3, 155.3, 153.7, 142.0,
136.4, 130.8, 130.8, 128.8, 127.4, 126.9, 120.3, 112.6, 109.1,
95.8, 46.3, 40.1, 38.2, 29.5 and 27.7; m/z (FAB) 751.2979
(M + Li+. C43H44N4O4S2Li requires 751.2964).
Scheme 3
was prepared in four steps from 3-nitro-dimethylaniline
following close analogy to the coumarin series.8 It was
discovered that compound 6 could be treated with thiolates,
which added in a Michael fashion, to produce thio-ethers
having fluorescent properties which were similar to the original
coumarins. To access diamine sensors, compound 6 was
treated with several dithiols to produce a series of quinolone
dimers 2a–g. The linker connecting the quinolone cores (X)
was varied in order to probe the effects of length and rigidity
on the diamine binding properties.
Compound 7. Compound 6 (200 mg, 0.587 mmol), NaSEt
(74 mg, 0.880 mmol), and DMF (6 cm3) were combined in a
flame-dried, N2-filled flask. The reaction mixture was stirred at
ambient temperature for 28 h. CH2Cl2 was added, and the
solution was washed with H2O (4 6 30 cm3) and brine
(30 cm3). The CH2Cl2 layer was dried on MgSO4, and the
solvent was removed in vacuo. The resulting yellow solid
was purified via flash chromatography (Et2O–CH2Cl2, 5 : 95
to 15 : 85), and the product was isolated as a yellow solid
(86.7 mg, 40%). mp 163u–164u; nmax(film)/cm21 1687, 1681,
1597, 1498, 1483, 1457, 1430 and 1391; dH/ppm (300 MHz,
CDCl3; Me4Si) 10.63 (1 H, s), 8.36 (1 H, d, J 9.3), 7.25–7.30
(5 H, m), 6.62 (1 H, dd, J 2.3 and 9.3), 6.23 (1 H, d, J 2.1), 5.51
(2 H, s), 3.03 (2 H, q, J 7.4), 2.99 (6 H, s) and 1.25 (3 H, t, J
7.4); dC/ppm (75 MHz, CDCl3; Me4Si) 190.4, 162.0, 154.9,
153.5, 141.9, 136.3, 130.7, 128.7, 127.2, 126.8, 120.5, 112.5,
108.8, 95.6, 46.9, 40.0, 32.3 and 14.9; m/z (FAB) 373.1575 (M +
Li+. C21H22LiN2O2S requires 373.1562).
Spectroscopic analysis
The interactions of sensors 2a–g were probed spectroscopically
by UV–vis and fluorescence titration experiments. The series
of sensors displayed poor water solubility properties compared
to monomeric quinolones such as 7; thus, spectroscopic
analysis was carried out in a 1 : 1 methanol–buffer system.
The absorption spectra showed trends similar to those
observed with the coumarin analogs in which a large red shift
in absorption maximum was observed upon addition of
diamines to the sensors. As shown in Fig. 1(a), when sensor
2g was titrated with diaminopropane, a 28 nm shift in
Results and discussion
Synthesis
absorbance was observed, consistent with
a shift from
aldehyde to iminium ion forms (i.e., Scheme 2). The red shift
in absorption has been attributed to the hydrogen bond
between the formed iminium ion and the carbonyl group of the
chromophore. In fluorescence mode, by exciting the chromo-
phore at 495 nm, a large increase in fluorescence was observed
The synthesis of the series of diamine sensors is shown in
Scheme 3. One of the drawbacks to the original coumarin
sensors (i.e., 1, Scheme 1) was the limited functionality (R) that
could be appended to the coumarin core. To expand the range
of functionality available, the quinolone core (6, Scheme 3)
This journal is ß The Royal Society of Chemistry 2005
J. Mater. Chem., 2005, 15, 4073–4077 | 4075

View Article Online
Scheme 4
Results of titrations with six different guests for each of the
eight sensors are tabulated in Table 1.
NMR characterization
The mode of binding and stoichiometry were confirmed by
NMR experiments. First, compound 7 was titrated with
butylamine in MeOH-d4 and clean conversion to the imine
(8 in Scheme 4) was observed. The use of pure methanol
solvent for the NMR experiments was necessitated by the poor
solubility of the sensor. In contrast to butylamine, ethylene
diamine appeared to produce an aminal with the aldehyde
(compound 9). The dimeric sensors, however, behaved as
expected (i.e., Scheme 2). For example, addition of one
equivalent of diaminopentane to a solution of sensor 2a
(1.4 mM in CDCl3) gave complete conversion to the bis-imine
with no sign of aminal formation. Similarly ethylene diamine
also produced the bis-imine product by NMR. These experi-
ments served to demonstrate the mode of binding in this
potentially complicated system and furthermore confirmed the
1 : 1 stoichiometry of the interaction since one equivalent of
diamine completely converted the sensor aldehydes to imines.9
Fig. 1 (a) UV–vis and (b) fluorescence titrations of sensor 2g with
diaminopropane (10 mM sensor in 50% MeOH–buffered water [50 mM
HEPES, 240 mM NaCl, pH 5 7.4]). Curves in (a) correspond to the
addition of 0.05, 0.1, 0.015, 0.26, 0.36, 0.46, 0.61, 0.76, 1.1 mM guest.
Inset in (b) is the theoretical fit to the binding isotherm.
Selectivities
The data shown in Table 1 provide a number of trends in
relative affinity. In all cases, the diamines bound better than
butylamine with the obvious exception of the monoaldehyde 7.
The extent of the difference varied from 2.5 fold for sensor 2a
to 160 fold for 2g. This effect validated our assumptions
about the utility of aldehyde dimers as diamine sensors. The
upon titration with the diamine (Fig. 1(b)). The fluorescence
increase was fitted to a one-site binding isotherm which gave a
binding constant of 6700 M21 (inset in Fig. 1(b)) with a maxi-
mum fluorescence increase at saturation (Isat/I0) of 6.6-fold.
Table 1 Binding constants and maximum fluorescence enhancements for various amine guests binding to sensors 2a–g and 7
Butylamine
Diaminopropane
Diaminobutane
Diaminopentane
Ornithine
Lysine
b
Sensor
Ka/M21a
Isat/I0
Ka/M21
Isat/I0
Ka/M21
Isat/I0
Ka/M21
Isat/I0
Ka/M21
Isat/I0
Ka (M21
)
Isat/I0
2a
2b
2c
2d
2e
2f
2g
7
160
93
130
72
30
31
43
76
12
20
13
8.8
10
26
38
32
590
820
3300
3400
350
2100
6700
—
8.3
14
6.5
16
8.5
15
6.6
—
640
600
520
1600
160
690
1500
—
16
21
10
15
10
24
27
—
470
620
480
1200
350
270
2200
—
17
19
10
15
4.3
24
18
—
580
1000
840
3300
680
550
2400
—
14
22
12
13
7.3
22
20
—
410
540
780
2100
800
1500
2800
129
15
29
14
4.7
10
24
27
41
a
b
Error in Ka is ¡20% based on duplicate titrations. Isat/I0 is the maximum change in fluorescence derived from the theoretical fit to the
binding isotherm.
4076 | J. Mater. Chem., 2005, 15, 4073–4077
This journal is ß The Royal Society of Chemistry 2005

View Article Online
interesting feature of the series of sensors was their relative
selectivity for the various guests. The a priori expectation
for the dialdehyde sensors was that all would bind to
most diamines, however, shorter linkers would favor short
chain diamines over long chain diamines and vice versa.
Furthermore, guests which do not fit properly to a particular
sensor would produce lower fluorescence response. This
expectation is predicated on the observation that only a
properly aligned hydrogen bond in the iminium ion state
would produce an adduct with strongly shifted absorption
characteristics compared to the unbound sensor.4 Thus, a
mismatch between the sensor and guest might result in poor
hydrogen bonding and consequently poor fluorescence
responses.
enhancements were observed in partly aqueous methanol.
Furthermore, useful selectivities for different length guests
were achieved by variation of the linker portion of the sensor.
Acknowledgements
This work was supported by the National Institutes of Health
(R01 GM059245).
References
1 P. Jiang and Z. Guo, Coord. Chem. Rev., 2004, 248, 205; B. Valeur
and I. Leray, Coord. Chem. Rev., 2000, 202, 40; A. P. de Silva,
H. Q. N. Gunaratne, T. Gunnlaugsson, A. J. M. Huxley,
C. P. McCoy, J. T. Radermacher and T. E. Rice, Chem. Rev.,
1997, 97, 1515.
As seen in Table 1, the range of binding constants for the
different length guests varied over approximately an order of
magnitude from 160–6700 M21. Between the xylene linked
sensors 2a–c, very little selectivity was observed except for
the preference for the smaller ortho-linked sensor to bind
the smaller diamine. The meta-linked sensor (2b) demonstrated
a small preference for ornithine which was coupled with a very
large fluorescence increase (22 fold) for that guest.
2 T. W. Bell and N. M. Hext, Chem. Soc. Rev., 2004, 33, 589.
3 J. Raker and T. E. Glass, J. Org. Chem., 2002, 667, 6113.
4 E. K. Feuster and T. E. Glass, J. Am. Chem. Soc., 2003, 125, 16174.
5 K. E. Secor and T. E. Glass, Org. Lett., 2004, 6, 3727.
6 G. J. Mohr, C. Demuth and U. E. Spichiger-Keller, Anal. Chem.,
1998, 70, 3868; E. Mertz and S. C. Zimmerman, J. Am. Chem. Soc.,
2003, 125, 3424; E. Mertz, J. B. Beil and S. C. Zimmerman, Org.
Lett., 2003, 5, 3127; E. Mertz, S. L. Elmer, Am. M. Balija and
S. C. Zimmerman, Tetrahedron, 2004, 60, 11191–11204.
7 K. Fuji, K. Tsubaki, T. Kiyoshi, N. Hayashi, T. Otsubo and
T. Kinoshita, J. Am. Chem. Soc., 1999, 121, 3807; T. Mizutani,
K. Wada and S. Kitagawa, J. Am. Chem. Soc., 1999, 121, 11425;
A. P. de Silva and K. R. Sandanayake, Angew. Chem., Int. Ed. Engl.,
1990, 29, 1173; R. Ballardini, V. Balzani, A. Credi, M. T. Gandolfi,
F. Kotzyba-Hilbert, J.-M. Lehn and L. Prodi, J. Am. Chem. Soc.,
1994, 116, 5741.
Looking at the phenyl linked sensors 2d and 2e, it is
interesting to note that the meta-phenyl linked sensor had
overall higher affinity for all guests compared to the ortho-
phenyl linked sensor. In fact, 2d was selective between the
amino acid guests with a preference for ornithine which was
not observed for 2e. Indeed, the maximum fluorescence change
for 2d was much larger for ornithine than for lysine which may
indicate a better binding geometry for ornithine. Surprisingly,
the most selective sensors of the group were the flexible
sensors 2f and 2g. Excluding diaminopropane, the preferred
guest in both cases was lysine. Although the selectivity was
higher for 2f, the overall binding constant was higher for 2g.
Indeed, in both cases the fluorescence increase was much
smaller for diaminopropane than the other guests. Thus, the
sensors could distinguish guests based not only on the binding
constant, but also on the relative fluorescence change which
was induced.
8 M. A. Kirpichenok, V. M. Baukulev, L. A. Karandashova and
I. I. Grandberg, Chem. Heterocycl. Compd., 1991, 27, 1193;
A. Knierzinger and O. S. Wolfbeis, J. Heterocyclic Chem., 1980,
17, 225.
9 The NMR experiments revealed a secondary reaction in which
amine guests can substitute the thiol group at C4 (e.g., compound
10). The existance of this species could be verified specroscopically.
The amine adduct 10 (lmax 5 378 nm) was blue-shifted compared to
the parent thioethyl quinolone 7 (lmax 5 441 nm) and grew in quite
rapidly in the pure methanol experiment. In the methanol–buffer
experiments, this blue-shifted band did not appear except at very
high concentration of amine guest over long times. Presumably, in
the neutral buffered conditions of the latter experiments, the amines
were largely protonated and not nucleophilic. Thus, in the course of
the fluorescence titrations reported in Table 1, the change in
fluorescence was, in fact, only due to iminium ion formation.
Conclusions
Sensors for diamines were readily prepared by dimerization of
a quinolone aldehyde core. The aldehyde sensors bound the
diamine guests by the formation of 1 : 1 bis-iminium ion
adducts. Good binding constants and excellent fluorescence
This journal is ß The Royal Society of Chemistry 2005
J. Mater. Chem., 2005, 15, 4073–4077 | 4077