A quinolinium-derived turn-off fluorescent anion sensor

Article abstract of DOI:10.1039/b919821k

A quinolinium-derived anion sensor has been synthesised which shows a turn-off fluorescence response in the presence of anions, with selectivity for acetate. The compound exhibits complex anion binding comprising of a host dimer, 2:1 and 1:1 host:guest sp

Full text of DOI:10.1039/b919821k

View Article Online / Journal Homepage / Table of Contents for this issue
PAPER
www.rsc.org/obc | Organic & Biomolecular Chemistry
A quinolinium-derived turn-off fluorescent anion sensor†
Adam N. Swinburne,a Martin J. Paterson,b Andrew Beeby*a and Jonathan W. Steed*a
Received 23rd September 2009, Accepted 18th November 2009
First published as an Advance Article on the web 4th January 2010
DOI: 10.1039/b919821k
A quinolinium-derived anion sensor has been synthesised which shows a turn-off fluorescence response
in the presence of anions, with selectivity for acetate. The compound exhibits complex anion binding
comprising of a host dimer, 2 : 1 and 1 : 1 host : guest species. Fluorescent quenching is due to both
dynamic and static processes with charge transfer being the dominant mechanism.
Introduction
The design of receptors capable of selectively binding and sensing
anions is challenging and is a current area of active research.1–4
We5–9 and others10–13 have used receptors derived from a hexa-
substituted triethyl benzene scaffold as a means of creating
conformationally flexible anion hosts exhibiting some degree of
preorganisation. These hosts have proved effective anion receptors
with the addition of redox active (ferrocenyl)14 or luminescent
(anthracenyl,5 pyrenyl9) reporter groups allowing electrochemical
or ‘turn on’ fluorescent sensing. Several recent examples of turn-
off PET sensors for biologically relevant anions have also been
reported.15–19
The use of quinolinium salts as anion sensors has been
investigated due to their ease of synthesis, their water solubility and
their known fluorescent quenching by anions.20–23 Many studies
have focused on halide detection, particularly chloride. Chloride
is an anion of great importance in biological chemistry, as a
component of extra-cellular fluids, in environmental science as
well as consumer products and industrial processes.20
The quinolinium based sensors reported to date are invariably
sensors based on dynamic quenching processes with no specific
anion recognition built-in.20 To our knowledge there have been
no quinolinium based sensors derived from preorganised anion
receptors capable of binding and discriminating between anions.
We now report a receptor based on quinolinium functionalities
that acts as both binding and ‘turn off’ fluorescent sensing unit
appended to a hexa-substituted benzene core.
Scheme 1 Compounds 1 and 2.
92% for the pyridinium analogue. However, the low yield might
be due in part to the repeated recrystallisation required to remove
the incompletely reacted two arm substituted product. The control
compound 2 has also been prepared in an analogous manner. This
lacks the preorganised, multiply charged binding cavity found
in compound 1 and allows a comparison to be made between
preorganised multivalent24 and non-preorganised, single binding
site quinolinium based anion sensors.
The structure of compound 2 was confirmed by single crystal X-
ray structural determination (Fig. 1). The crystal packing exhibits
p–p stacking of the aryl rings, with a plane to plane distance of
3.616(3) A. The crystal also shows quinolinium to quinolinium
p–p stacking with a plane to plane distance of 3.413(2) A.
˚
˚
Synthesis
The synthesis of compound 1 (Scheme 1) involves the reaction
of quinoline with 1,3,5-tribromomethyl-2,4,6-triethylbenzene in
a manner analogous to related pyridinium-based receptors.5
The reaction time required is considerably longer than for the
pyridinium compounds (ca. 48 h as opposed to 6 h) and reflects
the added steric hindrance found in the quinolinium derivative.
The (unoptimised) isolated yield is also lower (24%) compared to
aDepartment of Chemistry, Durham University, South Road, Durham, UK
DH1 3LE. E-mail: jon.steed@durham.ac.uk; Fax: +44 191 384 4737;
Tel: +44 191 334 2085
bSchool of Engineering and Physical Sciences, Heriot-Watt University,
Edinburgh, UK EH14 4AS
† Electronic supplementary information (ESI) available: Fig. S1–S8.
CCDC reference number 755341. For ESI and crystallographic data in
CIF or other electronic format see DOI: 10.1039/b919821k
Fig. 1 Molecular structure of compound 2.
1010 | Org. Biomol. Chem., 2010, 8, 1010–1016
This journal is
The Royal Society of Chemistry 2010
©

View Article Online
Solution state binding properties
The solution state binding properties of compound 1 were
1
investigated using H NMR spectroscopic titration in CD3CN
solvent with a number of anion guests. The binding isotherms are
shown in Fig. 2.
Fig. 4 1H NMR dilution study for 1. The CH2-quinolinium proton is
followed.
dimerisation constant was included into the stoichiometry model
incorporating both 1 : 1 and 2 : 1 host : anion complexes, to give an
improved fit to the complexation data. The binding constants thus
determined for compound 1 for various anions are summarised in
Table 1 as determined by non-linear least squares fitting using
HypNMR 2006.25 DOSY NMR experiments were attempted to
further confirm the presence of a host dimer. However as the dimer
also forms in the presence of anions other than PF6- this technique
was not able to distinguish a change in diffusion coefficient within
the experimental error.
Fig. 2 Binding isotherms for the ortho-quinolinium proton of compound
1 with various anions in acetonitrile-d3. Precipitation occurred after
3 equivalents of acetate.
Significant chemical shift changes in the resonances assigned
to the protons ortho and meta to the quinolinium nitrogen atom
were observed, with very small shifts for the other signals. In all
cases, inflections in the isotherms occur before the addition of
one equivalent of anion, suggesting more a complicated series of
binding equilibria than a 1 : 1 host : guest stoichiometry.
In order to determine the appropriate binding model a Job plot
with TBA-I was determined (Fig. 3). Iodide was chosen due to the
high solubility of the host–iodide complex. The data points appear
skewed further to the right hand side of the plot than might be
expected for a 1 : 1 host : guest stoichiometry. A 2 : 1 host : guest
stoichiometry would give rise to a maximum at a mole fraction
of 0.66, whereas a 1 : 1 host : guest stoichiometry would give a
maximum at 0.5 mole fraction. It is possible that both of these
species coexist resulting in an intermediate value between the two.
Bromide was found to bind the strongest, with chloride having
a similar affinity. Iodide is bound significantly less strongly than
the other anions tested. Interestingly, the compound shows strong
binding to nitrate, with binding constants in the same region as
those to chloride. Binding constants could not be refined for
1
acetate in this solvent due to very strong binding. A H NMR
spectroscopic titration of compound 1 with acetate in DMSO-d6
(a more competitive solvent) was carried out, unfortunately small
chemical shift changes were observed (ca. 0.2 ppm) for which a
binding constant could not be determined.
The 1H NMR spectrum of compound 2 was also recorded with
the addition of 1, 2 and 3 equivalents of TBA-Cl. The major change
is a shift of the peak at 9.17 ppm by up to 0.3 ppm, accompanied by
broadening and is in contrast to compound 1 in which a change
of approximately 1.5 ppm is observed. All other chemical shift
Table 1 Binding constants determined by 1H NMR spectroscopic
titrations
Stoichiometry
(H–G)
Log bHG
1
Anion
Host
Dimer
Cl-
2 : 0
1.57(7)
1 : 1
2 : 1
1 : 1
2 : 1
1 : 1
2 : 1
1 : 1
4.54(5)
8.01(8)
4.75(5)
8.10(7)
3.391(1)
6.120(8)
4.55(2)
8.18(3)
Br-
I-
-
NO3
Fig. 3 Job plot for compound 1 with TBA-I.
2 : 1
-
a
CH3CO2
-
b
b
The presence of a host dimer can be established via a dilution
H2PO4
-
-
HSO4
study on the host as the PF6 salt (Fig. 4). The host shows
modest chemical shift changes as a function of concentration
however a value of log K20 = 1.57(7) fits the data well. This
a Binding constant too large to be determined by this method. b Precipitate.
This journal is
The Royal Society of Chemistry 2010
Org. Biomol. Chem., 2010, 8, 1010–1016 | 1011
©

View Article Online
changes for compound 2 are less than 0.1 ppm. This experiment
implies that there is a markedly weaker interaction between anions
and 2 compared to the preorganised, multidentate compound 1.
DFT calculations were used to further understand the binding
mode of compound 1. The complex geometry was optimised
using a B3LYP/4-31G basis set on the host with the hydrogen
atoms proximal to the anion augmented with additional s and p
diffuse functions and a B3LYP/6-311+G* basis set on chloride
and B3LYP/6-31+G* on acetate.
The optimised 1 : 1 host : guest complex of 1 with chloride is
shown in Fig. 5. The complex shows C3v symmetry with hydrogen
bonding to the chloride by o- and m-quinolinium hydrogen
atoms. The hydrogen bonding pattern suggested by the calculation
matches the pattern of contact induced chemical shift changes
observed using 1H NMR spectroscopic titrations in solution.
Acetate also binds strongly to the host. The non-spherical shape of
the acetate means the optimised geometry consists of each acetate
oxygen atom hydrogen bonded to two quinolinium arms.
Fig. 6 DFT optimised geometry for the 2 : 1 host : guest structure of 1
with chloride.
Photophysical properties
The absorbance and excitation spectra of 1 as the hexafluorophos-
phate salt are shown in Fig. 7. Both spectra show the same lmax of
317 nm (typical for a compound of this type) and the same overall
band shape.20,26,27 A broad featureless emission band is observed
with a tail reaching to approximately 550 nm and a Stokes shift of
90 nm.
Fig. 5 DFT optimised geometries for (a) 1 with chloride and (b) 1 with
acetate.
Fig. 7 Photophysical properties of compound 1. lex = 317 nm. lem
=
408 nm.
1
The binding equilibrium suggested by H NMR titration data
The fluorescent properties of compound 2, used as a stan-
dard, were also investigated. The band shapes and lmax of the
absorbance, excitation and emission spectra, are identical to that
of compound 1.
The photoluminescent quantum yield (PLQY) of compounds
1 and 2 was calculated via an integration sphere methodology28
to give a PLQY of 0.21 for compound 1 and a markedly lower
value of 0.09 for compound 2. Table 2 collates the important
photophysical data of compound 1 and 2. The effect of the
addition of anions on the fluorescent emission of compound 1 was
investigated using fluorescent spectroscopic titrations. Solutions of
suggests a 2 : 1 host : guest complex. DFT calculations were used
to further test the feasibility of this complex. Fig. 6 shows the
optimised geometry for the 2 : 1 host : chloride complex. The
chloride is equidistant between the two cations and sits higher
in the binding cavity than is observed for the 1 : 1 complex. At
low basis sets the complex falls apart and is only stable using
higher basis sets. This suggests that the interactions in the complex
are weak, as would be expected given the electrostatic repulsion
between the two hosts and the measured association may have a
significant solvophobic component.
1012 | Org. Biomol. Chem., 2010, 8, 1010–1016
This journal is
The Royal Society of Chemistry 2010
©

View Article Online
Table 2 Summary of the photophysical properties of 1 and 2
Compound
1
2
Absorbance
max/nm
Excitation
max/nm
Emission
max/nm
PLQY
Lifetime
ns
317
317
408
317
317
408
l
l
l
0.21
17.7
2.26
-0.79
0.09
5.1
1.84
-0.84
E1/2
V vs. SCE
2.33 ¥ 10-6 mol dm-3 of host were titrated against the TBA-salts
of chloride, bromide, iodide, nitrate and acetate. Fig. 8 shows the
quenching of the fluorescence emission observed with the addition
of acetate.
This quenching is significant and can be observed with the naked
eye, Fig. 9.
Fig. 10 Stern–Volmer plot for compound 1 with various anions.
the data are still useful in a qualitative sense. For ease of
interpretation the number of equivalents of guest rather than
quencher concentration is used for the x-axis. As can be seen
from Fig. 10, acetate is by far the best quencher at virtually
all equivalents of guest, and it can therefore be concluded that
the receptor is selective for acetate and discriminates this anion.
This also corresponds well to the strong binding of acetate by 1
1
shown by H NMR. Interestingly, chloride is also a very good
quencher, showing equal or better quenching than the heavier
atom bromide (especially at low equivalents). Iodide on the other
hand is much worse at quenching the fluorescence. This is the
opposite selectivity observed for related published quinolinium
based systems containing one fluorophore, which are dominated
by the electron transfer properties of the anion in water.29
A
dynamic quenching mechanism has been suggested for these
compounds that does not involve specific complex formation.20
Thus the present preorganised anion receptor is apparently
altering the ability of the quinolinium to discriminate particular
anions by means of its fluorescence emission by specific binding
effects, with strongly bound anions such as chloride quenching
more than iodide.
Fig. 8 Fluorescence quenching of 1 upon addition of acetate.
The control compound 2 also shows quenching of its fluo-
rescence upon addition of anions. A linear trend with anion
concentration is observed in most cases. Table 3 contains the
calculated Ksv values. Comparison of the Ksv values to the literature
for similar systems shows compound 2 has much less variation in
the Ksv than has previously been observed, although water was
used in previous work instead of acetonitrile.20,26,27,30 This may be
explained by binding processes supported by NMR evidence (vida
supra). This observation along with the use of a less viscous solvent
may, in part, explain the observations.29 Importantly, compound
1 shows significantly higher quenching and greater discrimination
of anions than compound 2.
Table 3 Stern–Volmer data for compound 2
Fig. 9 Compound 1 (1 ¥ 10-4 mol dm-3) in the absence of acetate (left)
and the presence of excess acetate.
Anion
Ksv/M-1
Chloride
Bromide
Iodide
Nitrate
Acetate
570 50
490 40
410 40
600 40
Non-linear
Stern–Volmer plots were created for the various anions titrated
and have been corrected for dilution effects (Fig. 10). Due to
the complex nature of the binding equilibrium it is not possible
to fit the data using standard Stern–Volmer equations although
This journal is
The Royal Society of Chemistry 2010
Org. Biomol. Chem., 2010, 8, 1010–1016 | 1013
©

View Article Online
Typically Stern–Volmer plots are linear for static or dynamic
quenching, showing deviation if mixtures of the two mechanisms
are operating.18,20 The Stern–Volmer plot for acetate quenching
of 1 shows two principal regions—significant quenching up to
2 equivalents of anion followed by a shallower region which
eventually begins to curve. Since the binding affinity is high,
complexation is likely to be essentially complete after the addition
of only a few equivalents of guest and hence the continued
quenching up to a 30 fold excess of anion suggests that a collisional
quenching mechanism is also operating.
UV-vis spectroscopic titrations were used to investigate the
effect of anion on the absorbance bands of compounds 1 and
2 as a means of distinguishing the quenching mechanisms. Fig. 11
shows the UV absorbance spectrum of compound 1 with a lmax of
317 nm in the absence of guest and in the presence of 10 equivalents
of TBA-Cl with negligible dilution effects. It can be seen that there
is an increase in the overall intensity of the band and a small
shift in the lmax (ca. 2 nm), suggesting the formation of a host–
guest complex and static quenching, consistent with the 1H NMR
spectroscopic experiments (vida infra). The same experiment was
also performed on compound 2 there is an overall decrease in
the band intensity and no significant lmax change, consistent with
dilution effects only and dynamic quenching.
between concentration and absorbance. This provides further
evidence for the formation of a host dimer for 1.
A further method of distinguishing static from dynamic quench-
ing is to investigate how the fluorescent lifetime of a species varies
with added anion. The emission from compound 1 and 2 shows a
bi-exponential decay, possibly suggesting complex emissive decay
processes. Similar behaviour has been observed for related systems
and is attributed to nanosecond solvent relaxation processes.31–33
The fluorescent lifetimes of compound 1 were monitored at a
range of anion concentrations for a range of anions. Fig. 13 shows
the logarithmic plot of the decay of the fluorescence at various
equivalents of chloride. The decay becomes steeper, consistent
with shorter lifetimes, and less linear as chloride concentration
increases. In all cases a decrease in the excited state lifetime is
observed and suggests that both static and dynamic processes
contribute to the quenching of the fluorescence of compound 1.
Fig. 13 Logarithmic plot of the fluorescence decay of compound 1 at
several different equivalents of TBA-Cl.
Similar experiments conducted on compound 2 also showed a
general decrease in the excited state lifetime of the compound in
the presence of anion. It seems likely that as with compound 1
there is a mixture of dynamic and collisional quenching occurring
for this system. The 1H NMR data suggests static quenching may
also play a role in the quenching of compound 2.
Fig. 11 Absorbance band of compound 1 in the absence and presence of
10 equivalents of TBA-Cl.
Both compounds exhibit dynamic and collision quenching.
However for compound 1 the 1H NMR, UV-vis and fluorescence
spectroscopic titrations suggest static quenching is the dominant
process. Compound 2 on the other hand appears to be dominated
by dynamic quenching.
A dilution study was also conducted on the host compound
in the absence of guest and is shown in Fig. 12. The data shows
a distinct non-linear trend and deviates from the Beer–Lambert
law. The control compound 2 however shows a linear relationship
There are several possible quenching mechanisms for quinolin-
ium based sensors including electron transfer from the anion,34-36
charge-transfer complex or exciplex formation,37,38 or heavy atom
quenching.39 Jayaraman et al.29 have shown that the log kq is
linearly proportional to the indicated excited state reduction
potential of the quinolinium species and the free energy change
for charge transfer. In addition the quenching rate decreases with
increasing anion oxidation potential for dynamically quenched
systems in water. They conclude therefore that the quenching
mechanism is via a dynamic excited state charge transfer process.
Whilst a full scale mechanistic study for compounds 1 and 2 has
not been undertaken, insights into the mechanism can be gained
via simple experiments. An estimate of the excited state reduction
potential (known as the indicated excited state reduction potential)
Fig. 12 UV-vis dilution study of compound 1.
1014 | Org. Biomol. Chem., 2010, 8, 1010–1016
This journal is
The Royal Society of Chemistry 2010
©

View Article Online
General procedure for 1H NMR spectroscopic titrations
for compound 1 and 2 can be calculated by Ered + E0,0 where Ered
is the reduction potential and E0,0 is the singlet excitation energy.
The reduction potential can be obtained by cyclic voltammetry
(see Table 2). Both compounds 1 and 2 have the same singlet
excitation energy of 3.90 eV and therefore give an indicated excited
state reduction potential of 3.11 V and 3.06 V for compound
1 and 2, respectively. This is exactly in line with that observed
for methyl quinolinium.29 Given the very similar excited state
reduction potentials for compounds 1 and 2 to that of methyl
quinolinium, which is known to be quenched by charge transfer
processes, it is logical to assume that this mechanism also occurs
for compounds 1 and 2. However further experiments would be
required to confirm this.
All chemical shifts are reported in ppm relative to residual solvent.
A solution of the host species of known concentration typically
0.5–1.5 mM, was made up in an NMR tube using the appropriate
deuterated solvent (0.5 ml). Solutions of the anions, as TBA
salts (1 ml) were made ten times the concentration of the host
solution. The guest solution was typically added in 10 ml aliquots,
representing 0.1 equivalents of the guest with respect to the host.
Larger aliquots were used in some cases where no inflection of
the trace was evident. Spectra were recorded after each addition
and the trace was followed simultaneously. Results were analysed
using the curve-fitting program HypNMR 2006.25
Compound 1 is also water soluble and has the potential to
act as an anion sensor in aqueous media. Unfortunately 1H
NMR titrations in water showed very small changes in all proton
chemical shifts, suggesting the solvent is too competitive for 1 to
bind anions effectively. In addition, there is negligible quenching
by anions over that of dilution effects. Polar aprotic solvents such
as DMSO also proved too competitive for effective anion binding.
General procedure for UV-vis spectroscopic titrations
UV-vis titration were carried out using a Perkin-Elmer Lambda
35. A solution of typical concentration 1 ¥ 10-4 mol dm-3 of
host was made in a volumetric flask. A 3 ml sample of host
solution of concentration of 2.33 ¥ 10-6 mol dm-3 for 1 and 1 ¥
10-5 mol dm-3 for 2 were prepared by dilution of the stock solution.
Guest solutions were prepared such that 300 ml of guest solution
corresponds to 10 equivalents of host. Solutions were prepared
using acetonitrile as solvent.
Conclusions
It has been shown that a tripodal anion receptor with quinolinium
arms providing charge assisted hydrogen bonding functionality,
binds anions strongly with complicated speciation consisting of
an unbound host, a host dimer, and both 1 : 1 and 2 : 1 host : guest
complexes. The fluorescence of the quinolinium groups can be
used as an anion sensing reporter, with quenching occurring in the
presence of anions. Crucially the preorganised, tripodal receptor,
1 exhibits a different quenching selectivity to simple quinolinium
derivatives. A mixture of static and dynamic quenching was ob-
served, with selectivity for acetate. In the case of the preorganised
compound 1 static quenching dominates, however for compound
2 dynamic quenching dominates. The mechanism of quenching
has provisionally been assigned as via the formation of a charge
transfer complex.
Fluorescence spectroscopy
Emission and excitation spectra were obtained using a Jobin-Yvon
Horiba Fluorolog 3-22 Tau-3 spectrofluorimeter with a right angle
illumination method and were corrected for the spectral response
of the instrument.
Fluorescence spectroscopic titrations were carried out using
the equipment described above. A solution of concentration 1 ¥
10-4 mol dm-3 of host was made in a volumetric flask. A 3 ml sample
of host solution of typical concentration 2.33 ¥ 10-6 mol dm-3 for
1 and 1 ¥ 10-5 mol dm-3 for 2 were prepared by dilution of the
stock solution. Guest solutions were prepared such that 300 ml of
guest solution corresponds to 10 equivalents of host. Solutions
were prepared using acetonitrile as solvent.
Experimental
Photoluminescent quantum yields were calculated using either
a Jobin-Yvon Horiba Fluorolog 3-22 Tau-3 spectrofluorimeter
with a right angle illumination method and an integration sphere,
following the integration sphere method of Beeby et al.28
Lifetimes were obtained via the time-correlated single-photon
counting technique. The method described by Beeby et al.41
was used. Samples were excited using the third harmonic of
mode-locked cavity dumped Ti-sapphire laser. The emission was
collected at 90◦ to the source of excitation and the emission
wavelength selected by a monochromator (Jobin Yvon Triax
190). Fluorescence detection was obtained using a single photon
avalanche diode (ID-Quantique ID-100) that was linked to a time-
to-amplitude converter (Ortec 567) and a pulse height analyser,
PHA, (E.G. & G. Trump Card and Maestro for Windows v
5.10). Fluorescence decays were recorded to a minimum of 10 000
counts in the peak channel of the PHA with a record length
of 1000 channels. The decay and IRF data were transferred to
PC for analysis in Microsoft Excel via the method of iterative
reconvolution and nonlinear least squares fitting.
Instrumentation
All NMR spectra were performed on a Varian Mercury-400
(400 MHz for 1H), Varian Inova-500 machine (500 MHz for 1H,
126 Hz for 13C) or a Varian DD-700 (700 MHz for 1H, 176 MHz
for 13C) and were referenced to residual solvent. Electrospray
(ES) mass spectroscopy was recorded on a Thermo-Finnigan LTQ
instrument, whilst Matrix Assisted Laser Desorption Ionisation
(MALDI) experiments were recorded on an ABI Voyager-DC
STR. Fourier transform infrared spectra were recorded with a
Perkin Elmer Spectrum 100 ATR instrument (Perkin-Elmer, Nor-
walk, Ct., USA). For each spectrum, 64 scans were conducted over
a spectral range of 4000 to 600 cm-1 with a resolution of 4 cm-1. The
analysis was carried out with the Spectrum Express 1.01 software.
Elemental analysis was performed using an Exeter Analytical inc.
CE-400 Elemental Analyser. Commercial reagents were used as
supplied, without further purification. 1,3,5-Tri(bromomethyl)-
2,4,6-triethylbenzene was prepared as previously reported.40
This journal is
The Royal Society of Chemistry 2010
Org. Biomol. Chem., 2010, 8, 1010–1016 | 1015
©

View Article Online
General procedure for cyclic voltammetry experiments
References
1 J. L. Sessler, P. A. Gale and W.-S. Cho, Anion Receptor Chemistry,
Royal Society of Chemistry, Cambridge, 2006.
2 C. Caltagirone and P. A. Gale, Chem. Soc. Rev., 2009, 38, 520.
3 P. A. Gale, Coord. Chem. Rev., 2006, 250, 2917, preface to the special
issue and subsequent reviews.
4 T. Gunnlaugsson, M. Glynn, G. M. Tocci, P. E. Kruger and F. M.
Pfeffer, Coord. Chem. Rev., 2006, 250, 3094.
5 K. J. Wallace, W. J. Belcher, D. R. Turner, K. F. Syed and J. W. Steed,
J. Am. Chem. Soc., 2003, 125, 9699.
6 D. R. Turner, M. J. Paterson and J. W. Steed, J. Org. Chem., 2006, 71,
1598.
7 S. J. Dickson, E. V. B. Wallace, A. N. Swinburne, M. J. Paterson, Gareth.
O. Lloyd, A. Beeby and J. W. Steed, New J. Chem., 2008, 32, 786.
8 A. Barnard, S. J. Dickson, M. J. Paterson, A. M. Todd and J. W. Steed,
Org. Biomol. Chem., 2008, 7, 1554.
Cyclic voltammetry was carried out on a Chi Instruments Model
420 Electrochemical Analyser using two Pt wire counter electrodes
and a 2 mm Pt working electrode or a 3 mm glassy carbon working
electrode. The ferrocene/ferrocinium redox couple was used as an
internal reference. E1/2 values are reported vs. SCE (Fc = 0.4 V
with TBAPF6 in MeCN).42 A 0.1 mol dm-3 solution of TBAPF6
in dry MeCN was used as electrolyte. Solutions were degassed by
bubbling through N2 and the cell was kept under positive pressure
of N2 at all times.
Synthesis of 1
1,3,5-Tri(bromomethyl)-2,4,6-triethylbenzene (0.50 g, 1.13 mmol)
and quinoline (0.51 g, 3.97 mmol) were dissolved in dry DCM
and heated at reflux for 24 h. The solvent was evaporated and
50 ml of methanol was added to the residue. Ten equivalents of
9 M. H. Filby, S. J. Dickson, N. Zaccheroni, L. Prodi, S. Bonacchi, M.
Montalti, C. Chiorboli, M. J. Paterson, T. D. Humphries and J. W.
Steed, J. Am. Chem. Soc., 2009, 130, 4105.
10 V. Amendola, M. Boiocchi, L. Fabbrizzi and A. Palchetti, Chem.–
Eur. J., 2005, 11, 5648.
11 H. Ihm, S. Yun, H. G. Kim, J. K. Kim and K. S. Kim, Org. Lett., 2002,
4, 2897.
12 Y. Bai, B. G. Zhang, J. Xu, C. Y. Duan, D. B. Dang, D. J. Liu and Q. J.
Meng, New J. Chem., 2005, 29, 777.
13 Y. Bai, B.-G. Zhang, C. Y. Duan, D.-B. Dang and Q.-J. Meng,
New J. Chem., 2006, 30, 266.
14 J. Zhang, A. M. Bond, J. Belcher, K. J. Wallace and J. W. Steed, J. Phys.
Chem. B, 2003, 107, 5777.
15 W. X. Liu and Y. B. Jiang, Org. Biomol. Chem., 2007, 5, 1771.
16 C. M. G. Dos Santos, T. McCabe and T. Gunnlaugsson, Tetrahedron
Lett., 2007, 48, 3135.
17 R. M. Duke and T. Gunnlaugsson, Tetrahedron Lett., 2007, 48, 8043.
18 C. M. G. Dos Santos, M. Glynn, T. McCabe, J. S. S. De Melo, H. D.
Burrows and T. Gunnlaugsson, Supramol. Chem., 2008, 20, 407.
19 E. B. Veale, G. M. Tocci, F. M. Pfeffer, P. E. Kruger and T.
Gunnlaugsson, Org. Biomol. Chem., 2009, 7, 3447.
20 C. D. Geddes, Meas. Sci. Technol., 2001, 12, R53.
21 C. D. Geddes, J. Photochem. Photobiol., A, 2000, 137, 145.
22 M. S. Mehata and H. B. Tripathi, J. Lumin., 2002, 99, 47.
23 O. van den Berg, W. F. Jager and S. J. Picken, J. Org. Chem., 2006, 71,
2666.
-
NH4PF6 were added and the product was precipitated as the PF6
salt. Yield = 0.28 g, 0.27 mmol, 24%. m.p. 210–222 ◦C; 1H NMR
(CD3CN, 700 MHz) d = 9.18 (3H, d, J = 8.7 Hz, ArH), 8.77 (3H,
d, J = 7.8 Hz, ArH), 8.74 (3H, d, J = 6.0 Hz, ArH), 8.47 (3H, d,
J = 7.8 Hz, ArH), 8.40 (3H, t, J = 7.8 Hz, ArH), 8.13 (3H, t, J =
7.8 Hz, ArH), 8.1 (3H, dd, J = 8.7 and 6.0, ArH), 6.13 (6H, s, CH2-
1
N+), 2.43 (6H, br q, CH2CH3), 1.16 (9H, br t, CH2CH3); 13C-{ H}
NMR (CD3CN, 176 MHz) d = 151.9, 148.5, 145.8, 139.3, 136.6,
131.3, 130.9, 130.6, 127.9, 122.5, 118.7, 54.0, 24.1, 14.6; n/cm-1
=
=
3096 (C–H), 2972 (C–H), 1625 (Ar C C), 1591 (Ar C C), 1524
=
(Ar C C), 1377, 1229, 987, 817. Found: C, 45.39; H, 4.44; N, 3.74.
Calc. for C42H42F18N3P3·5H2O: C, 45.29; H, 4.71; N, 3.77%. Mass
spectrometry could not be obtained for this compound.
Synthesis of 2
Quinoline (0.50 g, 3.9 mmol) was dissolved in dichloromethane
(60 ml). Benzyl bromide was added (3.31 g, 19.1 mmol, 5 eq)
and the solution refluxed for 24 h. The solvent was removed in
vacuo resulting in a white powder. The powder was washed with
diethyl ether then dissolved in methanol. Excess (1.50 g) NH4PF6
was added and a white powder precipitated out of solution slowly.
The powder was isolated by filtration, washed with◦diethyl ether
and dried under ambient conditions. m.p. 140–146 C; 1H NMR
(CD3CN, 700 MHz) d = 9.17 (2H, m, ArH), 8.39 (1H, d, J =
8.5 Hz, ArH), 8.34 (1H, d, J = 9.6 Hz ArH), 8.16 (1H, m, ArH),
8.08 (1H, dd, J = 9.0, 6.4 Hz, ArH), 7.99 (1H, d, J = 7.3 Hz, ArH),
7.43 (3H, m, ArH), 7.31 (2H, m, ArH), 6.17 (2H, s, CH2-N+); 13C-
24 L. Baldini, A. Casnati, F. Sansone and R. Ungaro, Chem. Soc. Rev.,
2007, 36, 254.
25 P. Gans, HypNMR 2006, University of Leeds, Leeds.
26 C. D. Geddes and P. Douglas, J. Appl. Polym. Sci., 2000, 76, 603.
27 R. Krapf, N. P. Illsley, H. C. Tseng and A. S. Verkman, Biophys. J.,
1988, 53, A196.
28 L. Porres, A. Holland, L. O. Palsson, A. P. Monkman, C. Kemp and
A. Beeby, J. Fluoresc., 2006, 16, 267.
29 S. Jayaraman and A. S. Verkman, Biophys. Chem., 2000, 85, 49.
30 C. D. Geddes, K. Apperson, J. Karolin and D. J. S. Birch, Dyes Pigm.,
2001, 48, 227.
31 C. D. Geddes, K. Apperson and D. J. S. Birch, Dyes Pigm., 2000, 44,
69.
32 S. G. Schulman, R. M. Threatte, A. C. Capomaccia and W. L. Paul,
J. Pharm. Sci., 1974, 63, 876.
33 D. Pant, H. B. Tripathi and D. D. Pant, J. Lumin., 1992, 51, 223.
34 J. Najbar and M. Mac, J. Chem. Soc., Faraday Trans., 1991, 87,
1523.
1
{ H} NMR (CD3CN, 176 MHz) d = 150.4, 149.6, 139.4, 137.2,
133.7, 132.0, 131.5, 131.4, 130.4, 130.4, 128.7, 123.2, 119.9, 62.0;
m/z (ES+) 220.1 [M - PF6]+; n/cm-1 3097 (C–H), 1627 (Ar C C),
1590 (Ar C C), 1528 (Ar C C), 1456, 1379, 1361, 1045, 817.
Found: C, 51.65; H, 3.85; N, 3.91. Calc for C16H14F6NP·0.3H2O:
C, 51.85; H, 3.97; N, 3.78%.
=
=
=
35 H. Shizuka, M. Nakamura and T. Morita, J. Phys. Chem., 1980, 84,
989.
36 H. Shizuka, T. Saito and T. Morita, Chem. Phys. Lett., 1978, 56, 519.
37 K. A. Abdullah and T. J. Kemp, J. Photochem., 1985, 28, 61.
38 A. R. Watkins, J. Phys. Chem., 1973, 77, 1207.
39 J. Bendig, S. Helm and D. Kreysig, Chem. Phys. Lett., 1978, 54, 466.
40 K. J. Wallace, R. Hanes, E. Anslyn, J. Morey, K. V. Kilway and J. Siegel,
Synthesis, 2005, 2080.
41 A. Beeby, S. FitzGerald and C. F. Stanley, Photochem. Photobiol., 2001,
74, 566.
42 N. G. Connelly and W. E. Geiger, Chem. Rev., 1996, 96, 877.
Acknowledgements
We are grateful to the EPSRC and Durham University for funding,
the Heriot-Watt University Information and Computing Services
for use of the Heriot-Watt high performance cluster service.
1016 | Org. Biomol. Chem., 2010, 8, 1010–1016
This journal is
The Royal Society of Chemistry 2010
©