Energy transfer in coumarin-sensitised lanthanide luminescence: Investigation of the nature of the sensitiser and its distance to the lanthanide ion

Article abstract of DOI:10.1039/c3cp52279b

A series of lanthanide complexes [Ln(dpxCy)₃]^3- have been synthesised. The ligands are composed of a coordinating dipicolinic acid backbone decorated with a polyoxyethylene arm fitted with a coumarin moiety at its extremity. The nature of the coumarin as well as the length of the linker have been varied. Upon excitation at 320 nm, the coumarin exclusively acts as an antenna while the dipicolinic acid core is not excited. Upon excitation below 300 nm, both parts are excited. With europium as a metal centre, the relaxation of the europium ion (intrinsic quantum yield ΦEuEu and radiative lifetime τ_r) is constant for all the studied ligands. Therefore, the observed differences in overall quantum yield (ΦEuL) in such systems come exclusively from the variation of the terminal coumarin. The overall quantum yields of the studied complexes are low (ΦEuL < 2% in aqueous solution). In order to rationalise the mechanism of the energy transfer and to improve the sensitisation efficiency (η_sens), the distance between the coumarin sensitiser and the lanthanide centre was explored in solution and compared to the solid state. In the solid state, a dramatic effect was confirmed, with an improvement of 80% in the quantum yield ΦEuL for short linkers ((-CH₂CH₂O-)_n with n = 1 compared to n = 3). By monitoring the lifetime decay of the excited state of the lanthanide cation with nanosecond vs. microsecond time-resolved spectroscopy at low temperature, the sensitisation of the lanthanide ions by coumarin derivatives was demonstrated to mainly occur through the singlet excited state of the coumarin and not via the usual triplet pathway. No evidence of a different behaviour at room temperature was found by transient triplet-triplet absorption spectroscopy.

Full text of DOI:10.1039/c3cp52279b

PCCP
View Article Online
View Journal | View Issue
PAPER
Energy transfer in coumarin-sensitised lanthanide
luminescence: investigation of the nature of the
sensitiser and its distance to the lanthanide ion†
Cite this: Phys. Chem. Chem. Phys., 2013,
15, 15981
Julien Andres and Anne-Sophie Chauvin*
A series of lanthanide complexes [Ln(dpxCy)3]3ꢀ have been synthesised. The ligands are composed of a
coordinating dipicolinic acid backbone decorated with a polyoxyethylene arm fitted with a coumarin
moiety at its extremity. The nature of the coumarin as well as the length of the linker have been varied.
Upon excitation at 320 nm, the coumarin exclusively acts as an antenna while the dipicolinic acid core is
not excited. Upon excitation below 300 nm, both parts are excited. With europium as a metal centre,
Eu
Eu
the relaxation of the europium ion (intrinsic quantum yield F and radiative lifetime tr) is constant for
all the studied ligands. Therefore, the observed differences in overall quantum yield (FELu) in such
systems come exclusively from the variation of the terminal coumarin. The overall quantum yields of the
studied complexes are low (FELu o 2% in aqueous solution). In order to rationalise the mechanism of
the energy transfer and to improve the sensitisation efficiency (Zsens), the distance between the
coumarin sensitiser and the lanthanide centre was explored in solution and compared to the solid state.
In the solid state, a dramatic effect was confirmed, with an improvement of 80% in the quantum yield
FELu for short linkers ((–CH2CH2O–)n with n = 1 compared to n = 3). By monitoring the lifetime decay of
the excited state of the lanthanide cation with nanosecond vs. microsecond time-resolved spectroscopy
at low temperature, the sensitisation of the lanthanide ions by coumarin derivatives was demonstrated
to mainly occur through the singlet excited state of the coumarin and not via the usual triplet pathway.
No evidence of a different behaviour at room temperature was found by transient triplet–triplet absorp-
tion spectroscopy.
Received 30th May 2013,
Accepted 25th July 2013
DOI: 10.1039/c3cp52279b
www.rsc.org/pccp
molecules to ensure an extended harvesting of the excitation
light source and can be designed to be incorporated in ligands
Introduction
The luminescence of the lanthanide ions is unique and widespread
in the different domains that utilise luminescent materials.1–3
Their photoluminescence however takes place through sensitisa-
tion of the lanthanide ion, because the lanthanide ions have very
weak extinction coefficients due to the forbidden character of
their sharp 4f–4f transitions. The sensitiser absorbs the excita-
tion light and transfers the energy to the lanthanide ions which
become excited. Upon relaxation, the lanthanide emits light
through luminescence phenomena.4 Organic chromophores
with high extinction coefficients are therefore interesting
which strongly coordinate lanthanide ions.
Three strategies ensure a good luminescence of the lanthanide
complex, based on different ligand designs.
The first strategy is to design ligands devoid of a chromophoric
unit, which strongly coordinate to the lanthanide ion in a 1 : 1
L : Ln ratio (1 ligand per lanthanide ion), and derivatise one of the
coordination sites to introduce a sensitiser. Such systems are often
based on the 1,4,7,10-tetraazacyclododecane-1,4,7,10-tetraacetic
acid backbone (dota) and its cyclen macrocycle, or diethylenetri-
aminepentaacetic acid (dtpa) and its podant architecture. In those
examples, one of the carboxylic acids is typically coupled to a
sensitiser by forming an amide bond with this chromophore.5
In the second strategy, several chromophoric units are
´
´ ´
Ecole Polytechnique Federale de Lausanne, ISIC, BCH 1405, CH-1015 Lausanne,
Switzerland. E-mail: anne-sophie.chauvin@epfl.ch
† Electronic supplementary information (ESI) available: Luminescence variation introduced to enhance the extinction coefficient and the architec-
of the europium emission upon titration of the dp3Cy ligands with europium,
ture allows the presence of several emissive lanthanide centres per
absorption spectra of the free dp3Cy ligands and of their europium complexes,
complex. This is the case for the dendrimeric polyaminoamine
photophysical properties of the [Tb(dp3Cy)3]3ꢀ complexes and of the free dp3Cy
PAMAM.6
ligands including quantum yields FLLn, FLL and lifetimes tobs, fluoresence and
The third strategy is to directly coordinate the sensitiser to
the lanthanide ion. A chromophore with suitable coordination
phosphorescence spectra of the dpxC1 complexes, derivation of the sensitisation
efficiency from the rate of lanthanide population. See DOI: 10.1039/c3cp52279b
c
This journal is the Owner Societies 2013
Phys. Chem. Chem. Phys., 2013, 15, 15981--15994 15981

View Article Online
Paper
PCCP
sites is therefore needed. Examples include b-diketones and coupled to a derivative of dipicolinic acid (dpa) as a coordinating
pyridine containing ligands such as dipicolinates among moiety. We elucidate the sensitisation pathways in such complexes
others. This strategy is often adopted because it ensures that by spectroscopic techniques and show how the ligands can be
the sensitiser is close to the lanthanide ion, which maximises tuned to improve the efficiency of the sensitisation and of the
its energy transfer rate.2
luminescence of the lanthanide ion.
The energy transfer from the sensitiser to the lanthanide ion
¨
is assumed to occur through Dexter’s or Forster’s mechanism.
Design and synthesis of the ligands
The rate of the energy transfer depends, in both cases, on the
overlap between the donor (sensitiser) and the acceptor (lanthanide The design of the ligands was carried out to allow investigating
ion) energy levels, and on the distance between the donor and the the influence of several sensitisers without disrupting the
acceptor.7,8 Dexter’s mechanism relies on short range electron coordination sphere. One of the major problems when comparing
exchanges (exponential decay of the energy transfer rate with the different sensitisers is that the coordination sphere and therefore
¨
separation distance r), whereas Forster’s mechanism occurs the ligand field around the lanthanide ion often depend on the
through longer ranged dipole–dipole couplings (decay proportional sensitiser. It has been indeed proven that the radiative lifetime of
to rꢀ6). The actual sensitisation probably occurs in most cases the lanthanide excited states mainly depends on the refractive
through a complex combination of both mechanisms together index of the medium and on the coordination environment of the
with higher order couplings. Excitation energy transfer is lanthanide ion.20 It is hence difficult to conclude whether an
indeed still discussed and refined nowadays.9,10 The particular improved emission of the luminescent lanthanide ion is due to
case of lanthanide sensitisation, where the excitation of the the coordination, to the structure of the ligand, to its photophysical
acceptor lanthanide ion is forbidden, was also questioned.11–13 properties or to a combination of these contributions.
The dependence of the energy transfer rate on the separation
In order to minimise the effect of the coordination sphere, we
distance between the donor (sensitiser) and the acceptor chose to exclude the sensitiser from the first coordination sphere.
(lanthanide ion) is nonetheless always present, whatever the We opted for a coordination site formed by the dipicolinate (dpa)
mechanism. In this regard, very few examples of a correlation framework, which forms 3 : 1 L : Ln complexes under stoichiometric
between the distance of the sensitiser relative to the lanthanide conditions. The chromophore of the dpa framework is a classic
ion and the quantum yield of the sensitised emission were example of a good sensitiser of several lanthanide ions. Never-
highlighted.14
theless, its absorption is limited to short-wave UV light below
The energy transfer in sensitised luminescence of the 300 nm. A distant sensitiser that would absorb higher than
lanthanide ions is widely accepted to take place mainly from 300 nm would thus be exclusively excited (i.e. no excitation of the
the triplet state of the sensitiser to one of the spectroscopic dpa moiety at lex > 300 nm). We chose here coumarins as
levels of the lanthanide ion. A certain correlation between the sensitisers absorbing above 300 nm, yet below 400 nm to ensure
energy of the triplet state of the sensitiser and the quantum that the excited state is not too low (which would preclude any
yield of the sensitised lanthanide emission was established sensitisation). The sensitiser was grafted at the para position of the
from various ligands.15,16 The triplet pathway seems however dipicolinate moiety, with a linker in between the dpa moiety and
not mandatory. It has been proven that when the intersystem the sensitiser. The tridentate dipicolinate coordination site was
crossing rate is smaller than about 1011 ꢀ1, a singlet transfer preferred to a dota or dtpa one because of its well defined tris
s
mechanism is consistent with the experimental data.13 Recurrent structure forming a nine-coordinated lanthanide complex (whereas
examples of both near infrared (NIR) and visible light emitting dota and dtpa are seven or eight-coordinated and include, in
lanthanide ions sensitised through singlet pathways are evidenced aqueous solution, one or two water molecules in the first coordina-
in several investigations.17–19 Furthermore, it was demonstrated tion sphere), because of its ability to sensitise the lanthanide ion by
that a singlet mechanism is actually interesting to increase the excitation below 300 nm (yet not exploited here), and because of the
absorption range towards the visible spectrum (the singlet state good stability of its lanthanide complexes and rapid complexation
being higher than the corresponding triplet state, and the Stokes in aqueous solution (see the Experimental section).21
shift between the absorption and the fluorescence being smaller
There are a few examples of coumarin-sensitised lanthanide
than between the absorption and the phosphorescence).19 luminescence in the literature. Most of these examples involve
Besides the sensitisation pathway(s) and the associated macrocyclic crown ethers, or the direct coordination of a
mechanism(s) of energy transfer, there are many deactivation coumarin.22–27 Feau et al. were particularly prolific with coumarin-
´
(or relaxation) pathways that are in competition with the sensitised lanthanide complexes, showing nice examples of
sensitisation of the lanthanide ion. For all these reasons, the coumarins as sensitisers in polyaminocarboxylate complexes.28,29
relationship between the photophysical properties of the sensitiser An analyte responsive luminescent probe based on an analyte-
and those of the corresponding luminescent lanthanide complex triggered formation of a coumarin sensitiser was also published
is fairly intricate and any conclusion should always be made with recently by Borbas and co-workers.30 Finally, ligands with
caution.
quinolinone sensitisers were studied by Sevlin and co-workers.31–34
We propose here a versatile strategy that enables investigating The structure of quinolinones is close to the one of coumarins, the
different fundamentals of the sensitisation of the lanthanide endocyclic oxygen atom of coumarins being replaced by an amine in
ions with simple complexes based on coumarins as sensitisers quinolinones.
c
15982 Phys. Chem. Chem. Phys., 2013, 15, 15981--15994
This journal is the Owner Societies 2013

View Article Online
PCCP
Paper
We chose polyoxyethyene arms (POE, x = 1–3) as linkers the sensitiser is clearly not in direct contact with the lanthanide
between the sensitiser and the coordination site. This choice was ion because of the three dpa moieties that already fill the
motivated by its water-soluble compatibility, by the absence of coordination sphere. Furthermore, in their investigation of
conjugation with the dpa moiety or other chromophoric bridges, trioxyethylenated dpa derivatives, Gassner et al. did not observe
and by the characterisation of lanthanide complexes based on any folding of the POE side chain in the dp3OMe ligand and
similar structures (without the sensitiser at the end of the side in the [Lu(dp3OMe)3]3ꢀ complex in solution by ROESY-NMR
chain).35 This approach is therefore different from other designs spectroscopy.35
involving dpa derivatives, for example as undertaken by Maury
We recently published a first investigation of one such complex
4-methylumbelliferone (4-methyl-7-hydroxycoumarin)
et al., who have developed highly conjugated systems with a with
a
donor–acceptor effect for two photon microscopy.36
fluorophore at the end of a trioxyethylene side chain (x = 3).39
The general structures of the ligands and corresponding This investigation focused on the different sensitisation pathways
complexes are depicted in Scheme 1. In solid states, the POE depending on the excitation wavelength and confirmed that this
side chains usually adopt a trans–gauche–trans conformation at ligand architecture does not preclude the sensitisation of the
each –CH2CH2O– unit, which forms helical structures.37 In lanthanide ion. However, the sensitisation efficiency was found
solution, the movement (diffusion) of polymer pendent arms to be lower when exciting the distant coumarin (at 320 nm)
grafted on a surface generally describes mushroom structures compared to the simultaneous excitation of the dpa and
at room temperature when the density of the polymer pendent coumarin moieties (at 270 nm). Furthermore, the efficiency of
arms is low (brush structure at high density).38 The position of the europium sensitisation was rather low and was attributed
the sensitiser at the end of the side chain relative to the (i) to a poorly located triplet state energy of the donor relative to
lanthanide ion is hence certainly quite flexible; nevertheless, the europium acceptor spectroscopic levels, and (ii) to the
length of the side chain that can extend up to B15 Å. The
double sensitisation ability was yet an interesting way to
modulate the ratio between the coumarin blue fluorescence
and the red luminescence of the europium ion by changing the
excitation wavelength.
In the present article, we only focus on an excitation at 320 nm.
This excitation exclusively populates the distant sensitiser (the
coumarin). Therefore, the dpa moiety is essentially acting only as
a coordination site. However, this does not mean that the dpa
moiety is not involved in the sensitisation process. Its contribution
is yet believed to be identical in all our complexes, because of its
stable and well defined coordination site that should be unaffected
by the distant coumarin, as reported previously.39
To further test the assumptions regarding the poor sensitisation
efficiency of the already reported distant coumarin, we decided
(i) to change the nature of the sensitiser for different
coumarins and (ii) in parallel, to shorten the –(CH2CH2O)x–
linker.
The different ligands are presented in Scheme 1. They will
be referred to as ‘‘dpxCy’’ hereafter where ‘‘px’’ refers to the
length of the side chain (x = 1–3), and ‘‘Cy’’ to the nature of the
sensitiser (coumarins, y = 1–5).
The synthesis of the ligands is performed in four steps. The
first step consists in the coupling of the sensitiser with the POE
side chain. A nucleophilic functional group – an acidic hydroxyl
at the seventh or fourth position on the coumarin core – reacts
with a tosylated oligoethylene glycol monomethyl ether. The
terminal monomethyl ether is then cleaved by trimethylsilyl
iodide (TMSI), thus forming the terminal alcohol. This alcohol
is grafted on a chelidamic diester through a Mitsunobu coupling.
Because the coupled diester is hard to purify by chromatography
without a massive loss of compounds, the Mitsunobu coupling has
to be as clean as possible, in order to ensure that it can be used
directly in the next step. We found that using polymer-supported
Scheme 1 Ligand design forming tris complexes with lanthanide ions in water
(pH 7.4) and allowing the variation of the sensitiser and of the length (x = 1–3) of
the polyoxyethylene (POE) side chain. Structures of the dpxCy ligands. Studied
triphenylphosphine (PS-TPP) instead of free triphenylphos-
phine is a very convenient way to guarantee a minimal amount
variation of the nature of the sensitiser Cy (coumarins) and of the length of the px
linker (POE side chain).
c
This journal is the Owner Societies 2013
Phys. Chem. Chem. Phys., 2013, 15, 15981--15994 15983

View Article Online
Paper
PCCP
of impurities in the ligand diester precursor, since the polymer The tris species is the most luminescent species in solution
can be easily filtered out of the solution. To be certain that no because of the absence of quenching water molecules in the
residual uncoupled chelidamic diester is present in the crude coordination sphere of the lanthanide ion.21,40 The lifetimes from
product of the Mitsunobu reaction, an excess of coumarin– the emissions at a L : Eu ratio of 3 : 1 are all around 1.4 ms ꢁ
oligoethylene glycol is used, and the complete disappearance of 0.1 ms. Such lifetimes are typical for the relaxation of europium
the uncoupled chemidamic diester is checked by TLC and by through f–f transitions in dpa complexes without water molecules
1H-NMR spectroscopy. The final step is the deprotection of the in the first coordination sphere. The parent dpa ligand exhibits for
carboxylic esters to form the carboxylates. The hydrolysis is example a lifetime of 1.6 ms for its tris species and 0.3 ms for
carried out in ethanol upon addition of an aqueous solution of its tris-hydrated bis species.21,40 With the dp3Cy containing
sodium hydroxide. The sodium salt of the ligand, which pre- complexes, the lifetimes at 1.4 ms are well fitted by monoexpo-
cipitates in ethanol, can then be further purified by a series of nential decays; the [LnL3] species are then the major lumines-
precipitations in ethanol. This procedure ensures the complete cent species in solution under those conditions. This result is
removal of impurities resulting from the Mitsunobu reaction in agreement with data obtained on similar structures by differ-
(mainly of the excess of coumarin–oligoethylene glycol) as ent techniques such as NMR spectroscopy, UV-vis absorbance
proven by NMR spectroscopy and microanalysis: the impurities spectroscopy and time-resolved spectroscopy.21,35 The dpa coordi-
being soluble in ethanol, while the carboxylate ligand is not. nation site is then, as expected, not sterically hindered by the
Five different coumarins were coupled to dpa in this way, using distant coumarins. The variation of the terminal coumarin has
triethylene glycol as a side chain. One coumarin (4-methyl- no impact on the formation of the lanthanide complexes.
umbelliferone) was coupled to the dpa moiety using three
different POE linkers (triethylene, diethylene and ethylene Absorption of the different coumarins
glycol) according to the same synthetic strategy. Each ligand
The absorption of the europium complexes are presented in
was characterised by NMR spectroscopy, Electrospray Ionisa-
Fig. 1 together with the absorption of each coumarin (for
tion Mass Spectrometry (ESI-MS) and microanalysis. The
comparison purposes, grafted to the POE side chain but with-
lanthanide complexes were directly prepared in aqueous
out the dpa moiety). The presence of the coumarins extends the
solution by adding the stoichiometric amount of lanthanide
absorption range from below 300 nm (absorption of the dpa
ions (from a titrated solution of a lanthanide salt) to a Tris-
moiety) up to 360 nm. The shapes of the coumarin absorptions
buffered solution at pH 7.4 of the diluted ligand. This is the
above 300 nm are practically unaffected by the coupling to the
normal procedure reported in the literature for the preparation
dpa coordination site and by the complexation of the ligands to
of 3 : 1 L : Ln lanthanide complexes with dpa derivatives.21
the europium ion (Fig. 1 and Fig. S2–S6 of the ESI†). Even
The investigation of the photophysical properties of the
though the extinction coefficients of the maxima (attributed to
ligand-centred and metal-centred emission was carried out in
Tris-buffered aqueous solutions at pH 7.4 (concentration of 0.1 mM
in lanthanide complexes). Such a pH value and concentration
range were demonstrated to be well suited for the luminescence
of the dp3C1 complexes39 and sustain the high stability of the
[LnL3] species with similar complexes (the stability is comparable
whatever the lanthanide ion).21,35
Variation of the sensitiser: coumarin
derivatives
Formation of the complexes in aqueous solution
Dp3C1 and a series of polyoxyethylenated dpa ligands were already
reported elsewhere to form stable 3 : 1 L : Ln complexes.30,39 Their
formation was investigated by NMR spectroscopy, ESI-MS, UV-Vis
spectrophotometry and luminescence. In this study, we do not
report once more the full characterisation of the complexes, but we
illustrate their stability by luminescence comparatively to already
reported results.
The formation of the complexes in solution was verified by
monitoring the emission of the lanthanide ion upon titration
of the dp3Cy ligands with Eu(III) (see Fig. S1, ESI†). A maximum
of the emission is reached at a ligand to lanthanide ratio (L : Ln)
Fig. 1 Molar extinction coefficients of the [Eu(dp3Cy)3]3ꢀ complexes (plain) and
the corresponding coumarin-trioxyethylene monomethyl ether (dashed) scaled
of 3 : 1, thus indicating that the [LnL3] species (tris species)
should form predominantly under stoichiometric conditions.
to show their absorption range in the complexes. Aqueous Tris-buffered solution
(0.1 mM in complex), pH 7.4.
c
15984 Phys. Chem. Chem. Phys., 2013, 15, 15981--15994
This journal is the Owner Societies 2013

View Article Online
PCCP
Paper
Table 1 Location of the singlet (S1) and triplet (T1) excited state of the dp3Cy
p - p* transitions) are changed upon coordination (values
reported in Table S1 of the ESI† for the europium complexes),
when the absorption spectra are scaled at the maximum of the
coumarin absorption, only small differences between the peak
and shoulder of the absorption of the coumarin moieties above
ligands in their gadolinium complex in wavenumber (ꢁ300 cmꢀ1
)
[Gd(dp3Cy)3]3ꢀ
S1(max)/cmꢀ1
T1(0–0)/cmꢀ1
y = 1
y = 2
26 000
27 000
27 000
28 000
28 000
21 500
23 500
23 500
23 500
23 000
300 nm are observed. This behaviour was anticipated because y = 3
y = 4
y = 5
the coumarins are not directly involved in the coordination,
which is performed by the dpa moiety. On the parent dpa, the
coordination indeed induces an increase of the extinction
coefficient, a narrowing of the transitions, and a small batho-
chromic shift (red-shift) as shown in Fig. S2 of the ESI.†
However, those effects are not observed with the dp3Cy ligands
in the absorption range of the dpa chromophore because the
superposed absorptions of the coumarins mask the absorption
structure of the dpa moiety.
Besides those coordination consequences, an essential
result is that an excitation above 300 nm (e.g. 320 nm) only
populates the coumarins, because the dpa moiety absorbs only
below 300 nm. In order to solely investigate the sensitisation
through the coumarins, all photophysical measurements are
performed under excitation at 320 nm. From a comparative
study of the direct excitation of the coumarin versus the double
excitation of the dpa and coumarin moieties, the results are
found to be similar to those already reported with the dp3C1
ligand.39
coupling position on the coumarin, seventh vs. fourth respec-
tively, and are characterised by the same singlet and triplet
values (singlet at 27 000 cmꢀ1 and triplet at 23 500 cmꢀ1).
However, a similar value for the singlet state of two ligands
does not imply that the triplet states are located at the same
energy. For example, dp3C4 and dp3C5 have a first excited
singlet state at the same position (i.e. 28 000 cmꢀ1), but a first
triplet excited state that is slightly shifted by calc. 500 cmꢀ1. It
probably comes from a slightly better stabilisation of the triplet
excited state in the dp3C5 structure (7-chloro) compared to the
dp3C4 structure (unsubstituted seventh position).
Metal-centred photophysical properties
The metal-centred emissions from the europium complexes
upon variation of the terminal coumarin are shown in Fig. 3.
They all exhibit the characteristic europium 5D0
-
7FJ f–f
transitions ( J = 1–4) under excitation at 320 nm, meaning that
the europium is sensitised by each coumarin (so-called antenna
effect). A residual fluorescence of the coumarin is also observed
alongside. These short-lived coumarin emissions are absent on
time-resolved spectra when a delay of a few microseconds is
applied between the pulsed excitation and the measurement.
The efficiencies of the ligand-centred and metal-centred
emissions for the europium complexes are presented in
Table 2 (for terbium, see Table S2 of the ESI†). The europium
complexes are particularly useful because additional photophysical
properties, which are not easy to find for the other lanthanide ions,
Ligand-centred photophysical properties
The ligand-centred emissions from the gadolinium complexes
(non emissive under those conditions) upon variation of the
terminal coumarin are shown in Fig. 2. The fluorescence from
the singlet state (S1) was measured at room temperature,
whereas the phosphorescence from the triplet state (T1) was
recorded at 77 K, 50 ms after a pulsed irradiation.
According to the data in Table 1, a correlation exists between
the structure of the coumarin and the location of its excited
states. For instance, dp3C2 and dp3C3 only differ by the
5
7
can be extracted from the intensity of the D0 - F1 transition
Fig. 2 Fluorescence from S1 and phosphorescence from T1 of the [Gd(dp3Cy)3]3ꢀ
complexes (lex = 320 nm). Aqueous Tris-buffered solution (0.1 mM in complex),
pH 7.4 (S1), and frozen solution at 77 K (10% glycerol added), 50 ms after a pulsed
irradiation (T1).
Fig. 3 Excitation (left hand side, plain for the Eu-centred emission, dashed for
the residual coumarin emission) and emission spectra (right hand side, lex
=
320 nm) of the [Eu(dp3Cy)3]3ꢀ complexes (0.1 mM) in Tris-buffered aqueous
solution, pH 7.4.
c
This journal is the Owner Societies 2013
Phys. Chem. Chem. Phys., 2013, 15, 15981--15994 15985

View Article Online
Paper
PCCP
Table 2 Photophysical properties of the [Eu(dp3Cy)3]3ꢀ and [Gd(dp3Cy)3]3ꢀ
involve the distant sensitiser. Therefore, we can fairly assume
that the difference in quantum yields FELu between the com-
plexes in Table 2 is only due to the difference in sensitisation
efficiency.
Eu
complexes at room temperature (lex = 320 nm). t
= 1.4 ms ꢁ 0.1 ms, trEu
=
obs
4.2 ms ꢁ 0.4 ms FEEuu = 33% ꢁ 5%, for all complexes in this table. Estimated error
of 10% on the quantum yields (sensitised quantum yields of the europium
emission FELu, quantum yields of the ligand-centred emission FLL)
This conclusion is however limited to these [Eu(dp3Cy)3]3ꢀ
complexes. A comparable study was performed with the terbium
complexes (see Table S3, ESI†), and points to a strong deactivation
of the terbium excited state for one of the terminal coumarin. This
was demonstrated with the short observed lifetime for the
[Tb(dp3C1)3]3ꢀ complex (0.6 ms ꢁ 0.1 ms). This is expected
because the dp3C1 ligand has the lowest excited states (either
singlet at 26 000 cmꢀ1 or triplet at 21 500 cmꢀ1). The 5D4 spectro-
scopic level of terbium(III) being located at 20 500 cmꢀ1, a back-
transfer seems highly probable. This also explains why both the
dp3C2 and dp3C3 ligands, with higher excited states, are better
than dp3C1 for the terbium sensitisation (FTLb = 1.4% and 1.7%
versus 0.5% respectively).
Ln = Eu
FELu/%
Ln = Gd
sens/%
[Ln(dp3Cy)3]3ꢀ
FLL/%
Z
FLL/%
y = 1
y = 2
y = 3
y = 4
y = 5
1.7
0.4
0.7
n.d.
0.3
7.7
0.9
1.1
n.d.
n.d.
5.1
1.2
2.1
9.1
1.1
1.6
n.d.
n.d.
o0.6
0.8
N.d. values were too low to be determined with the experimental setup.
5
relative to the total intensity from the D0 spectroscopic level.
5
7
The D0 - F1 transition has indeed a purely magnetic dipole
(MD) nature and no electric dipole (ED) character. This unique
characteristic simplifies the determination of the radiative
lifetime of europium, see eqn (1).41 This equation assumes that
the strength and energy of the MD transition is constant and
independent on the environment around the europium ion and
therefore, that the radiative rate depends only on the ratio of
Eventually, for europium(III) the 5D2 spectroscopic level,
located at 21 500 cmꢀ1, is at the same energy as the triplet of
dp3C1. Since no particular deactivation relative to the other
complexes was observed here, it strongly suggest that the
energy transfer occurs on a lower spectroscopic level such as
5
5
the D1 (at 19 000 cmꢀ1) or the emissive D0 (calc. 17 223 cmꢀ1
5
the MD transition relative to the total emission from the D0
spectroscopic level, and on the refractive index of the medium.
from the high resolution measurement of the 5D0
transition of Cs3[Eu(dp3C1)3]).39
’
7F0
1
Itot
¼ kr ¼ AMD;0 ꢂ n3 ꢂ
(1)
A comparison of the ligand-centred quantum yields of the
non-emissive gadolinium complexes FLL(Ln = Gd) with those of
the emissive europium complexes FLL(Ln = Eu), reveals that the
ligand-centred quantum yields of the gadolinium complexes
are always higher than those of the corresponding europium
complexes. A certain correlation exists between the ligand-
centred quantum yield of the non-emissive complex FLL(Ln =
Gd) with the corresponding europium-centred sensitised quan-
tum yield FELu. When the ligand-centred quantum yield was too
low to be measured, the europium quantum yield was also at
the threshold limit of our setup. The emission from the dp3C1
ligand in the gadolinium complex (9.1%) results in the highest
tr
IMD
From the radiative lifetimes, other parameters such as the
intrinsic quantum yield (FLLnn), which corresponds to the quantum
yield by direct excitation of the lanthanide ion (without the losses
from the sensitisation pathways), and the sensitisation efficiency
(Zsens) can be calculated according to eqn (2) and (3).
tobs
tr
Ln
Ln
F
¼
(2)
FLLn = ZsensꢂF
(3)
Ln
Ln
We observe from the emission spectra on Fig. 3, that all the europium quantum yield of the series (1.7%). The lower emission
europium emission spectra have exactly the same shape. Con- from the dp3C2 and dp3C3 ligands (1.1–1.6%) yields lower
sequently, the radiative lifetimes tr calculated from eqn (1) are quantum yields (0.4–0.7%), whereas the very few emitting
all identical (4.2 ms) whatever the terminal coumarin. This dp3C4 and dp3C5 ligands have a very weak europium emission.
result is consistent with other results found in the literature An expected quenching of the ligand emission by the emissive
(from europium complexes of dpa derivatives as well as of other lanthanide ion is however observed when measurable.
types of ligands), which seems to indicate that the radiative
These first results obtained by variation of the distant
lifetime depends mostly on the coordination sphere and on the sensitiser indicate several important limitations of the sensiti-
neighboring environment.20,35 The observed lifetime tobs sation process. First of all, an appropriate energy difference
(defined as the inverse of the sum of the radiative rate constant between the acceptor spectroscopic level of the lanthanide ion
with all other deactivation rate constants), which is fitted from and the donor excited states of the sensitiser is a very important
the exponential decay of the emission after a pulsed excitation, limitation and has to be optimised to maximise the sensitisa-
is also indistinguishable at 1.4 ms (within experimental error) tion efficiency. This was demonstrated by pointing out the
for all the complexes in Table 2. These lifetimes are similar to strong deactivation of the terbium excited state in the ligand
those of other dpa derivatives. As a consequence, the intrinsic with the lowest excited states. When this limitation is over-
quantum yield, which represents the efficiency of the radiative come, the sensitisation is then limited by the structure of the
deactivation of the lanthanide ion relative to all its deactiva- sensitiser. Small changes on a same backbone can indeed have
tions, has to be the same for each [Eu(dp3Cy)3]3ꢀ complex. This drastic impacts on the photophysics of this molecule. For
suggests that the deactivation of the europium ion does not example, it is well known that the incorporation of a heavy
c
15986 Phys. Chem. Chem. Phys., 2013, 15, 15981--15994
This journal is the Owner Societies 2013

View Article Online
PCCP
Paper
Table 3 Photophysical properties of the [Eu(dpxC1)3]3ꢀ and [Gd(dpxC1)3]3ꢀ
atom such as a bromine or an iodine atom on a fluorophore
increases the intersystem crossing rate by spin–orbit coupling.42
In our case, we observed that several complexes with similar
excited states but different structures exhibit completely different
quantum yields. The limitation is then probably the competing
deactivation processes, and particularly the quenching of each
sensitiser by its environment.
The sensitisation efficiency, which is defined as the ratio of
the number of sensitised lanthanide ions per number of excited
sensitisers, can be divided into several contributions according
to the sensitisation pathways. For a sensitisation through the
triplet state of the sensitiser, it is defined as the product of the
intersystem crossing efficiency, Zisc (number of triplet states
populated per number of excited singlet states), by the energy
complexes at room temperature in aqueous solution (aq) and in solid state (s).
l
ex = 320 nm. tEoubs = 1.4 ms ꢁ 0.1 ms, tEru = 4.2 ms ꢁ 0.4 ms (aq) and 2.8 ms ꢁ 0.4 ms
(s), FEEuu = 33% ꢁ 5% (aq) and 50% ꢁ 5% (s), for all complexes in this table. Estimated
error of 10% on the quantum yields (sensitised quantum yields of the europium
emission FELu, quantum yields of the ligand-centred emission FLL)
Ln = Eu
Ln = Gd
FELu/%
[Ln(dpxC1)3]3ꢀ aq
FLL/%
Z
sens/%
FLL/%
s
aq
s
aq
s
aq
s
x = 3
x = 2
x = 1
1.7
1.2
4.5
6.8
7.7 5.6 5.1
9.0
6.3 5.7 3.6 13.6
9.1
7.1
6.1
6.3
1.8 25.2 14.7 7.8 5.4 50.4 15.5 10.7
transfer efficiency, Zet (number of sensitised lanthanide ions per which remains between 1.2% and 1.8% (see Table 3). The ligand
number of triplet state sensitisers). To better understand the with the intermediate length (dp2C1) seems even slightly less
sensitisation process, the rate constants of each photophysical efficient than the other two ligands with a quantum yield FLEu
=
phenomenon must be considered, because the sensitisation is in 1.2%. This suggests that any increase of the energy transfer rate
competition with many different deactivation pathways. To that would be due to the shorter side chain (if any), would be
maximise the sensitisation efficiency, the energy transfer rate compensated by a decrease of another component of the sensi-
constant has to be as high as possible compared to the other tisation efficiency, or of the europium intrinsic quantum yield.
deactivations, ideally much higher, so that the other processes In terms of rate constants, it then corresponds to a competitive
can be neglected, which would yield a sensitisation efficiency process that deactivates the sensitiser or the europium ion.
of 100%.
In order to identify why the shortening of the side chain has
The rate constant of the energy transfer is assumed to no real impact on the quantum yield in aqueous solution, the
depend on the overlap between the donor state of the sensitiser following strategies were adopted. First, quenching phenomena
and the acceptor state of the lanthanide ion, as well as on the were decreased by studying solid state samples and by freezing
distance between the donor and the acceptor. We have already the aqueous solutions at 77 K. This strategy suppresses most
illustrated the importance of the energy of the excited states in of the quenching due to the presence of water molecules,
this first part, but we also pointed out the importance of the and lowers the rate constants of diffusion limited quenching
competing processes that can easily overcome the energy phenomena, respectively. The mechanisms of the sensitisation
transfer rate. In an attempt to rationalise the mechanism of were then investigated through time-resolved spectroscopy. The
the energy transfer in a sensitised luminescent lanthanide chemical environment of the complexes changes quite a lot
complex, we undertook a shortening of the POE side chain in once in the solid state (compared to the situation in aqueous
our original ligand that exhibits the best photophysical properties solution). The refractive index is higher in the solid state
in the series of Table 2, i.e. dp3C1. We then synthesised ligands (1.517 from ref. 43 versus 1.333 in water); there is a high density
dp2C1 and dp1C1 (Scheme 1) and investigated the photophysical of complexes probably in close contact, and a more rigid
properties of their respective europium complexes.
environment than in solution.
The ligand-centred and metal-centred photophysical properties
of the solid state samples were measured from the solid state
samples and not extrapolated from the aqueous solution. As seen
in Table 3, nearly all photophysical properties are altered relative to
those in aqueous solution.
Variation of the length of the linker
Photophysical properties in aqueous solution and in the
solid state
In aqueous solution, all the [Eu(dpxC1)3]3ꢀ (x = 1–3) complexes
The trend in the solid state demonstrates the expected behaviour
exhibit the same characteristic red emission from the europium when the side chain is shortened, i.e. an increase of the quantum
centre as well as the residual coumarin emission. The ligand- yield. The quantum yield FELu increases up to 25% when the
centred fluorescence and phosphorescence spectra are unchanged sensitiser is only separated from the coordination site by one
upon variation of the length of the side chain (same as Fig. 2, with –CH2CH2O– unit. The shape of each characteristic europium emis-
[Gd(dp3C1)3]3ꢀ at 26 000 and 21 500 cmꢀ1, see Fig. S7 of the ESI†), sion in the solid state is very similar throughout the series, so that
meaning that any difference in metal-centred photophysical the radiative lifetime can be estimated to be fairly identical for the
properties ought to be due to the shorter side chain and not to a three complexes. Since the observed lifetime is also not altered, the
different location of the donor excited state. The quantum yields intrinsic quantum yield stays at 50%. The difference in sensitised
together with other photophysical properties of the corresponding quantum yield then ought to come from an increased sensitisation
[Ln(dpxC1)3]3ꢀ (x = 1–3) complexes are presented in Table 3.
efficiency, and presumably from a higher energy transfer rate.
Noteworthily, the emission from the coumarin seems also to
Unexpectedly, the shortening of the side chain has practically
no influence on the quantum yield FELu in aqueous solution, be affected by the length of the side chain. This is even true in
c
This journal is the Owner Societies 2013
Phys. Chem. Chem. Phys., 2013, 15, 15981--15994 15987

View Article Online
Paper
PCCP
aqueous solution where the quantum yield of the coumarin both occur from the corresponding singlet or triplet excited
emission is twice that of the ligands with the longer linkers. states of the sensitiser. The singlet pathway is most of the time
This last point stresses one of the major problems of any neglected. However, efficient energy transfers were already
system that tries to investigate the relationship between the observed from the singlet excited state.17–19 Both contributions
distance of a donor–acceptor and the rate of the energy trans- should hence be considered first. The major one can then be
fer: at close distances, the environment of the donor and of the determined by observing whether a quenching of the triplet
acceptor is altered by the presence of its partner. In our case, state is observed in a luminescent lanthanide complex relative
the presence of the coordination site seems somehow to to a corresponding non-luminescent lanthanide complex. The
increase the efficiency of the coumarin emission, which means sensitisation efficiency should therefore contain the contribu-
either that the radiative lifetime of the coumarin is increased, tion from the singlet state ZSet, as well as the contribution from
or that the quenching and/or non-radiative relaxations are the triplet state ZTet. The sensitisation efficiency should thus be
decreased. By measuring the lifetimes of the europium 5D0 approximated as in eqn (4) (for the complete derivation from
spectroscopic level at 77 K, we also noticed that it jumped from the rate of lanthanide population, see ESI†).
2.2 ms with the longer linkers up to 4.7 ms for the shortest side
Zsens = ZSet + ZTetꢂZisc
(4)
chain, while the room temperature lifetimes in aqueous
solution are similar within experimental error (at 1.4 ms ꢁ
0.1 ms). This result strongly suggests that the efficiency in
aqueous solution is limited by the quenching from diffusing
solvent molecules, and that when the coumarin is close to the
coordination site and frozen it may also participate in the
second coordination sphere and help prevent the non-radiative
deactivations of the europium ion. The lifetime of 4.7 ms is
indeed what may be expected from a purely radiative relaxation
since the radiative lifetime was calculated in aqueous solution
at 4.2 ms (the radiative lifetime is identical with the three
ligands).
Each of those parameters can be expressed in terms of the
rate constant as the ratio of the considered process relative to
all the deactivations of the associated excited state as shown in
eqn (5).
kisc
kSobs
kSet
kSobs
kTet
kTobs
Zisc
¼
;
ZSet
¼
;
ZTet
¼
(5)
The sum of all the deactivations of the associated excited
state kobs can be approximated as in eqn (6), where kf is the
fluorescence radiative lifetime, kp the phosphorescence radia-
tive lifetime and knr the sum of all the non-radiative deactiva-
tions of the excited state.
The fact that the quantum yield FELu is unchanged upon
shortening of the linker may also be understood as a proof that
the sensitiser is standing at a similar average distance from the
lanthanide ion in aqueous solution and at room temperature,
which should be at the upper limit the length of the shortest
side chain (B5 Å). It would mean that the POE side chain may
be fairly folded rather than extended. On the other hand, it may
also come from a diffusion limited energy transfer. The excited
sensitiser at the end of the side chain moves around the
complex until the energy transfer rate is sufficiently high, and
therefore until the sensitiser is close enough to the lanthanide
ion, to allow an excitation transfer onto the lanthanide ion.
kSobs = kf + kisc + keSt + knSr
kTobs = kp + kTet + knTr
(6)
The observed relaxation rate constants kobs are usually
expressed as their inverse, which gives observed lifetimes tobs
(eqn (7)).
1
tobs
¼
(7)
kobs
In order to determine whether a preferential sensitisation
Those phenomena would be removed in the solid state since pathway is followed in our case, the sensitisation mechanisms
the structure is in that case more rigid and hence better were studied by time-resolved spectroscopy both for the solid
defined. The increased quantum yield of the coumarin emis- state samples and for the frozen aqueous solution at 77 K.
sion FLL for the complex with the shortest side chain (i.e. with
On the nanosecond time-resolved emission spectra of
the dp1C1 ligand) in aqueous solution could then be under- the solid state and frozen solution, the fluorescence from the
stood as a decreased self-quenching of the coumarin by the coumarin is observed together with the transitions from the
other coumarin moieties on the two remaining ligands of the higher 5D1 spectroscopic level of europium. The transitions
complex. Self-quenching was indeed observed in a previous from the typical 5D0 state are however absent. It then means
study to be
a
significant deactivation process in the that the energy transfer most probably occurs mainly on the 5D1
5
[Eu(dp3C1)3]3ꢀ complex.39 This decreased self-quenching in level, which lies B1700 cmꢀ1 higher than the D0 level. In the
the dp1C1 complex could be due to a limited diffusion range microsecond time scale (see Fig. 4), the 5D1 emission bands
when the linker is smaller, thus restricting the contact of the decrease while the D0 peaks rise.
5
5
coumarin moieties within the complex.
The decay rates of the D1 level are identical (within experi-
mental error) for all the ligands in the dpxC1 series, with a
lifetime of 1.3 ms ꢁ 0.2 ms. This decay is also observed on the
parent dipicolinate complex, so that it seems to be defined
Time-resolved luminescence: a tool for deciphering the
sensitisation pathway
There are actually two possible pathways that yield an energy mostly by the coordination sphere. The rise time of the transi-
transfer, the singlet S pathway, and the triplet T pathway, which tions from the 5D0 level is in the same range as the decay time
c
15988 Phys. Chem. Chem. Phys., 2013, 15, 15981--15994
This journal is the Owner Societies 2013

View Article Online
PCCP
Paper
needed to quantify them precisely and thereby to determine
whether an expected dependence of the rate constants on the
length of the side chain is actually observed. Those data are still
in agreement with a sensitisation through the singlet state of
the coumarin (S1) since the energy transfer occurs in the same
time scale as the deactivation of the singlet state. Triplet–triplet
transient absorption spectroscopy could provide further informa-
tion that should confirm that the triplet excited state is not involved
in the sensitisation of the lanthanide ion in aqueous solution and
at room temperature. This hypothesis is investigated in the next
section.
Transient absorption spectroscopy: time resolved T–T absorption
At room temperature, the triplet excited state of the coumarin
moiety could not be observed by luminescence. In order to
completely exclude the triplet excited state of the coumarin as a
sensitisation pathway of the europium ion, triplet–triplet (T–T)
transient absorption experiments were carried out at room
temperature. 4-Methylumbelliferone has a T–T absorption
between 400 nm and 550 nm.44 Aqueous solution of the
uncoupled mp3C1 chromophore (0.3 mM in Tris 0.1 M,
pH 7.4) and both the gadolinium and europium complexes of
dp3C1 (0.1 mM in Tris 0.1 M, pH 7.4) were excited by 355 nm
nanosecond pulses (20 Hz, 6 ns, 1.5 mJ) and the transient
absorption measured at 450 nm with a continuous Xenon light
source perpendicular to the excitation laser.
Fig. 4 Time-resolved emission spectrum of Na3[Eu(dp3C1)3] in the solid state
(lex = 320 nm) and the extracted decay of the 5D1 and rise of the 5D0 spectro-
scopic levels.
Transient absorptions with decays in the microsecond range
were measured for each solution containing the coumarin
chromophore. Solutions of the parent dpa complexes yielded
no transient signal. The difference between the decays of the
transient absorption is weak as shown in Fig. 5 by the normalised
absorption changes. An average lifetime of 2.9 ms was found.
Degassing the solution was not conclusive in showing any
significant quenching that would be due to the presence of
dioxygen in aerated solutions. However, the removal of oxygen
sometimes has very small effects on the relaxation of the triplet
5
of the D1 level. It is therefore obvious that the energy is first
transferred from the sensitiser onto the 5D1 level and from
there onto the D0 level.
5
In aqueous solution and at room temperature, the emission
from the 5D1 level is very difficult to observe (very weak
intensity). It probably comes from the rapid quenching of this
level by water molecules, which relax the europium ion down to
the 5D0 spectroscopic level. Water has indeed a vibrational
bending transition at 1645 cmꢀ1, which would be a good
acceptor for the energy gap between the 5D1 (19 000 cmꢀ1
)
and the D0 states (17 223 cmꢀ1, from ref. 39).
5
By increasing the time scale up to milliseconds, the decay of
the 5D0 spectroscopic level of the europium ion down to the 7FJ
spectroscopic levels is clearly seen by the luminescence decays
of the characteristic transitions. This decay is observed at low
temperature in frozen solution, in the solid state at room
temperature and in aqueous solution at room temperature.
In order to observe the phosphorescence of the ligand, low
temperature measurements were required. At 77 K in frozen
solution, the decay of the triplet excited state was found to be in
the second time scale, resulting in an emission that is clearly
visible up to 5 seconds after a laser excitation.
Because the europium emission is not present during this
long-lived triplet emission, the sensitisation through the triplet
state (T1) of the coumarin does certainly not happen. Further-
more, both the rise time of the D1 - F1 transition and the
decay of the singlet state are clearly in the nanosecond time
Fig. 5 Transient absorption at 450 nm after nanosecond pulsed excitation at
355 nm of aerated Tris buffered (pH 7.4) aqueous solutions of mp3C1 (0.3 mM,
cyan line), [Eu(dp3C1)3]3ꢀ (0.1 mM, red line) and [Gd(dp3C1)3]3ꢀ (0.1 mM,
blue line) at room temperature. The green line represents the fitted average
5
7
scale (o100 ns). However, a higher time resolution would be exponential decay.
c
This journal is the Owner Societies 2013
Phys. Chem. Chem. Phys., 2013, 15, 15981--15994 15989

View Article Online
Paper
PCCP
excited state, and particularly with water as solvent.45,46 The phosphorescent emission in the europium complex (B3%) com-
attribution of this transient absorption to T–T absorptions is based pared to the gadolinium analogue indicates that the intersystem
on the location of the T–T absorption in 4-methylumbelliferone.44 crossing rate is slower in the europium complex. This behaviour is
Because no substantial quenching is measured between the consistent with a singlet mediated energy transfer. Since no
gadolinium and europium complex, the observed excited state is quenching of the triplet excited state by energy transfer to the
certainly not involved in the sensitisation of europium. In compar- europium ion was observed by transient absorption at room
ison, known triplet sensitisation showed quenching by a factor of temperature, and because the europium emission is not observed
more than 2.5.47 Moreover, a strong correlation between the alongside the phosphorescent emission upon time resolution at
relaxation rate constant of the triplet excited state and the growing 77 K, the triplet excited state is undeniably not mediating any
rate constant of the europium luminescence is usually observed energy transfer. The lower population at 77 K of the triplet excited
when the triplet excited state mediates energy transfer onto the state in the europium complex can then be accounted to a slower
lanthanide ion. In our case, no rise of the europium population was intersystem crossing rate due to a faster relaxation of the singlet
observed during the deactivation of the triplet excited state. There- excited state by energy transfer onto the europium ion.
fore, a sensitisation of europium via the triplet excited state of the
coumarin moiety is unlikely to occur under such conditions.
Conclusions
Effect of the sensitisation on the intersystem crossing rate
In conclusion, we have been able to demonstrate with cou-
marin-based sensitisers that despite similar structures and
energy levels, the sensitisation efficiency cannot rely only on
the energetics of the excited states, but has to take into account
the possible deactivations and photophysical processes that
depend on the structure of each sensitiser. We have then
revealed for the first time with minimal perturbations of the
ligand design that a correlation between the length of the linker
(POE side chain) separating the sensitiser from the lanthanide
ion and the quantum yield is indeed observed in the solid state
as expected by energy transfer theories, but may not show up in
aqueous solution perhaps due to the flexible structure of the
polyoxyethylene linker in aqueous solution. We have also shown
that the sensitisation of the lanthanide ions by coumarin derivatives
at low temperature is occurring mainly through the singlet excited
state of the coumarin and not via the usual triplet pathway. No
evidence for a triplet-mediated sensitisation at room temperature
was found. It is therefore, the first occurrence of a relationship
between the quantum yield of a sensitised lanthanide luminescence
and the ‘‘distance’’ of the sensitiser relative to the lanthanide ion via
a singlet pathway sensitisation. We have proved that such a ligand
design is a versatile system for the investigation of sensitisers for
lanthanide ion. The synthesis (see the Experimental section) of
those ligands is simple: the only requirement is that the sensitiser to
investigate has a nucleophilic functional group that can be grafted
on a polyoxyethylene side chain. Other sensitisers may be investi-
gated that way with minimal adaptations of the synthesis. Chromo-
phores with known triplet pathways would be particularly
interesting to test comparatively to the singlet pathway sensitisers.
Besides, this simple structural design may then be a nice way to
screen a library of chromophores for potential lanthanide sensiti-
sers, and select promising candidates for an incorporation in more
complex architectures with a higher density of chromophores and
luminescent metal centres such as dendrimers.
In order to further confirm that the singlet excited state is
involved in the sensitisation pathway, the effect of the sensiti-
sation on the intersystem crossing rate was investigated. The
population of the triplet excited state was studied at 77 K in
frozen solutions. For this purpose, the emission spectra of the
gadolinium dp3C1 complex [Gd(dp3C1)3]3ꢀ, of the europium
dp3C1 complex [Eu(dp3C1)3]3ꢀ and of the free dp3C1 ligand
dp3C12ꢀ were measured at 77 K under continuous excitation at
320 nm (see Fig. 6). The emission spectrum of the gadolinium
complex shows a phosphorescence component that is exclusively
observed upon time resolution. On the other hand, the europium
complex has a weak phosphorescent component at this location.
Under time resolution, the phosphorescence is still observed and
lasts for seconds. By fitting the contribution of the time resolved
phosphorescence spectra in the total emission spectra, the triplet
state emission is estimated to account for 13% of the total
emission of the gadolinium complex at 77 K. In the free ligand,
the contribution is lower (B9%) as expected by the heavy atom
effect of the lanthanide ion in the complex. The much weaker
Experimental section
General procedures
Fig. 6 Emission spectra upon continuous excitation at 320 nm of the gadoli-
nium and europium complexes together with the free ligand in frozen solution
(77 K) and compared with the time resolved phosphorescence spectrum of the
coumarin.
The solvents were purified by a non-hazardous procedure by
passing them onto activated alumina columns (Innovative
c
15990 Phys. Chem. Chem. Phys., 2013, 15, 15981--15994
This journal is the Owner Societies 2013

View Article Online
PCCP
Paper
Technology Inc. system). The chemicals were ordered from buffered 0.3 mM ligand aqueous solution (pH 7.4). The excita-
Fluka and Aldrich and used without further purification. The tion and emission spectra, as well as the titration experiments
´
ESI-MS spectra were obtained by Dr Laure Menin and the were measured in 1 cm or 0.2 cm path length quartz cells. The
´
elemental analyses were performed by Dr E. Solari, at the Ecole low temperature measurements were performed at 77 K in
Polytechnique Federale de Lausanne. H, 13C and HSQC NMR quartz Suprasils capillaries with 10% glycerol added to the
1
´ ´
spectroscopy was performed on a Bruker Avance DRX 400 Tris-buffered aqueous solutions.
spectrometer at 25 1C, using deuterated solvents as internal
The quantum yields were determined in an integration
standards. The chemical shifts are given in ppm relative sphere by measuring the ratio of the emitted corrected intensity
to TMS.
over the absorbed corrected intensity. Empty capillaries have
been used as a blank. A 75 W Xenon light source with a
monochromator was used as a light source. The emissions
Photophysical measurements
The analytical grade solvents and chemicals were used without from the integrating sphere were collected with an optical fibre
further purification. The aqueous solutions were prepared from and analysed on a Hamamatsu photonic multichannel analyser
doubly distilled water. The lanthanides solutions were prepared C8808 detector. The correction function of the setup was
from the corresponding perchlorate salt and titrated by calculated with a calibrated standards Deuterium and Halogen
complexometry using a standardised Na2H2EDTA solution in light sources as reference irradiances. All concentrations were
a urotropine-buffered medium and with xylenol orange as an set at 0.1 mM.
indicator.
The time-resolved emission spectra were measured by
The aqueous solutions containing lanthanide complexes exciting the samples with an Ekspla NT 342/3/UV pulsed laser.
were directly prepared by adding the stoichiometric amount The excitation wavelength was set at 320 nm and gave a 6 ns
of lanthanide ions (from a freshly titrated solution of a lanthanide pulse of 0.53 ꢁ 0.04 cm2 with a typical energy of 1–5 mJ at a
salt) to a Tris-buffered aqueous solution at pH 7.4 of the diluted frequency of 20 Hz. The resulting emissions were collected
ligand. Aliquots of the solutions containing lanthanide complexes with an optical fibre and analysed on a Hamamatsu photonic
were titrated with a standardised Na2H2EDTA solution in a multichannel analyser C8808 detector.
urotropine-buffered medium and with xylenol orange as an
indicator to ensure that no free (i.e. uncomplexed) lanthanide cells. Aqueous solutions were photoexcited by short light pulses
ion were present in the solutions. (5 ns FWHM) generated by a frequency-tripled Q-switched
The transient absorptions were measured in 1 cm quartz
In order to better reproduce the properties of the prepared Nd:YAG laser (Ekspla NT-340). The pump fluence was of
solid state samples, we found that drying Tris-buffered solu- the order of 20 mJ cmꢀ2. The probe light from a Xenon arc
tions instead of unbuffered solutions helped producing repea- lamp was passed through filters, various optical elements, the
table results. The solid state samples are then solid state sample, and a grating monochromator, before being detected
complexes in a solid Tris matrix. The solid state samples were by a fast photomultiplier tube and recorded by a digital
prepared by mixing 50 mg of the ligand to the stoichiometric oscilloscope. The kinetic traces were typically averaged over
amount (3 ligands per lanthanide) of lanthanide chloride 3000–4000 consecutive laser shots.
(previously titrated to determine the hydration number) in
18 ml of hot water (90 1C), adding 2 ml of a Tris 1 M aqueous
Synthesis of the ligands
solution at pH 7.4 and slowly evaporating at room temperature The dp3C1 ligand was synthesised from commercial 4-methyl-
the 0.1 M Tris-buffered aqueous solutions until a solid residue umbelliferone according to our previous procedure.39 The
remained. The solid residue was then collected and further other ligands were synthesised according to the following
dried at 65 1C for 2 hours under slightly reduced pressure new synthesis. The tosylated oligoethylene glycol monomethyl
(B600 mbar). The warm samples were finally stored in a ethers were synthesised according to the procedure in the
desiccator at room temperature.
supplementary materials of Kohmoto and et al.48 Diethylcheli-
The UV-Vis absorption spectra were measured on a Perkin- damate (Et2chelida) was prepared according to a previously
Elmer Cary 1E spectrophotometer using 0.2 cm path length reported procedure.35
quartz cells.
General procedure for the synthesis of 4-hydroxycoumarins
The room temperature excitation and emission spectra of
the europium and gadolinium dp3Cy complexes were recorded 4-Hydroxycoumarins were synthesised according to Jung et al.49
on a Fluorolog 3-22 spectrofluorimeter from Jobin-Yvon. The The corresponding 4-substituted 2-hydroxyacetophenone
room temperature excitation and emission spectra of the (50 mmol) was dissolved with diethyl carbonate (75 mmol) in
terbium dp3Cy complexes and of the europium and gadolinium toluene (25 ml). The solution was added dropwise under an
complexes of dpxC1 were measured on a Perkin Elmer LS50B inert atmosphere and at room temperature to a vigorously
spectrofluorimeter in phosphorescence mode with a zero delay. stirred suspension of sodium hydride (60% w/w moistened
The titration experiments were also performed on the Perkin with oil, 250 mmol) in toluene (125 ml). After complete addition,
Elmer LS50B in phosphorescence mode by integrating the the reaction was heated up to 105 1C for 3 hours. The solvent was
intensity at 615 nm and monitoring the variation upon addition then evaporated. Water (100 ml) was added to the residual solid
of aliquots containing 1/15 europium equivalent to a 0.1 M Tris to quench the excess of sodium hydride. The solution was
c
This journal is the Owner Societies 2013
Phys. Chem. Chem. Phys., 2013, 15, 15981--15994 15991

View Article Online
Paper
PCCP
+
concentrated under vacuum to remove the remaining toluene (m, 4H), 3.42 (m, 2H), 3.23 (s, 3H). ESI-MS C16H21O6 , calcd
and acidified with concentrated hydrochloric acid down to pH 1 309.134 m/z, found 309.137 m/z. Mp3C5: yield 42%. 1H NMR d:
to form a precipitate that can be filtered out, washed with water 7.81 (d, J = 8.5 Hz, 1H), 7.62 (d, J = 2.0 Hz, 1H), 7.46 (dd, J1 =
and dried in vacuo. Recrystallisation from ethanol can be per- 8.5 Hz, J2 = 2.0 Hz, 1H), 5.98 (s, 1H), 4.37 (m, 2H), 3.87 (m, 2H),
formed to obtain higher purity 4-hydroxycoumarins if needed.
C4: 1H NMR d: 7.84 (dd, J1 = 7.8 Hz, J2 = 1.6 Hz, 1H), 7.66 (td, ESI-MS C15H20ClO6 , calcd 343.095 m/z, found 343.093 m/z.
J1 = 7.9 Hz, J2 = 1.7 Hz, 1H), 7.37 (m, 2H), 5.62 (s, 1H). The resulting yields are in agreement with the ability of the
3.65 (m, 2H), 3.56 (m, 2H), 3.53 (m, 2H), 3.42 (m, 2H), 3.23 (s, 3H).
+
Non commercial 2-hydroxyacetophenones like 4-chloro-2- substituent to lower the nucleophilicity of the coupling hydroxyl
hydroxyacetophenone can be synthesised through a Fries rear- group on the coumarin. Hence, an electron withdrawing group
rangement of the 3-substituted acetoxybenzene, for the at position 7, like a chlorine (C5), diminishes the yield compared
4-chloro-2-hydroxyacetophenone,
obtained from 3-chlorophenol, was reacted with AlCl3.50
C5: 1H NMR d: 7.83 (d, J = 8.5 Hz, 1H), 7.58 (d, J = 1.7 Hz, 1H yield up to a quantitative coupling.
3-chloroacetoxybenzene, to the unsubstituted 4-hydroxycoumarin (C4). On the other
hand, an electron donating group like a methoxy increases the
d1 = 20 sec.), 7.42 (m, 1H), 5.65 (s, 1H).
Coupling of POEated coumarins with chelidamic acid. POE-
Synthesis of 7-hydroxycoumarins. 7-Hydroxy-4-methoxycoumarin MEated coumarins were demethylated at the end of the POE
(C2) was synthesised from 4-hydroxy-7-methoxycoumarin (C3) (1 g, side chain by reacting them with trimethylsilyl iodide (TMSI,
5.20 mmol) by demethylation of the 7-methoxy group with hydro- 1.2 eq.) in acetonitrile (0.15 M in POEMEated coumarin). TMSI
iodic acid (75 ml) in a 1/1 (v/v) acetic acid/acetic anhydride solution was added dropwise in the warm solution (at 75 1C) and further
(total 60 ml) under reflux (120 1C) for 3 h 30 min. The obtained stirred 2 hours at 75 1C. The solvent was then evaporated, the
4,7-dihydroxycoumarin (471 mg, 51%) was then remethylated selec- residue dissolved in dichloromethane (DCM) and the solution
tively at position 4 by heating in methanol (0.1 M solution, 16 ml) washed with Na2S2O3 (0.09 M, 20 mmol Na2S2O3 per mmol
with concentrated sulfuric acid (1.25 ml acid per mmol dihydroxy- TMSI). The aqueous phase was then further extracted with
coumarin) for 1 hour and a half. 7-Hydroxy-4-methoxycoumarin (C2) DCM (2 times) and the collected organic layers dried over
is obtained as a slightly pinkish white powder (163 mg, 54% from Na2SO4, filtered and evaporated under reduced pressure to
4,7-dihydroxycoumarin).
yield the crude product used without purification in the
C2: 1H NMR d: 10.57 (s, 1H), 7.63 (d, J = 8.7 Hz, 1H), 6.80 (dd, next step.
J1 = 8.6 Hz, J2 = 2.2 Hz, 1H), 6.71 (d, J = 2.2 Hz, 1H), 5.69 (s, 1H),
Mitsunobu reaction with resin supported TPP. Diethylcheli-
3.98 (s, 3H). C3: 1H NMR d: 7.73 (d, J = 8.6 Hz, 1H), 6.94 (m, 2H), damate (Et2chelida) was dissolved in DCM (0.06 M in Et2ch-
5.43 (s, 1H), 3.86 (s, 3H).
elida) with the POEated coumarin (1.5 eq.) and resin supported
Synthesis of the POEMEated coumarins. Coumarins were triphenylphosphine (PS-TPP, 1.5 eq.). The dispersion was
coupled to a polyoxyethylene chain by attacking tosylated cooled down to 0 1C and diisopropylazodicarboxylate (DIAD,
triethylene glycol monomethyl ether (1.2 eq.) in DMF (0.3 M 1.5 eq.) was then added dropwise to the cold stirred solution.
in coumarin) with an excess of K2CO3 (22 eq.) at 75 1C for The reaction was then allowed to warm at room temperature
48 hours. The solvent was then evaporated, the residual slurry and stirred for 5 hours. The solution was then filtered to remove
dissolved in dichloromethane, washed three times with half the TPP resin. The resin was washed with DCM (50 ml) and
saturated NH4Cl aqueous solution, dried with Na2SO4, and the the filtrate was washed with aqueous HCl 1M (50 ml) once and
solvent evaporated in vacuo. Purification by silica gel chroma- three times with H2O (3 ꢃ 50 ml). The organic layer was dried
tography with ethyl acetate 100% as eluent gave the pure with Na2SO4, filtered and evaporated under reduced pressure to
desired product as a solid after evaporation of the solvent. In yield on oil composed of the coupled desired product together
the case of mp3C2 and mp3C3, the oily residue obtained after with residual uncoupled POEated coumarin. The excess of POE
washing with aqueous NH4Cl was solidified by adding a small containing coumarin is removed in the next step. The crude
amount of diethyl ether and sonicating at 40 1C. The white solid product was then used without further purification in the
was filtrated, rinsed with diethyl ether and further dried to yield final step.
pure mp3Cy.
Deprotection of the carboxylic acid by hydrolysis of the
Mp3C2: yield 100%. 1H NMR d: 7.70 (d, J = 8.8 Hz, 1H), 7.02 diethyl ester moiety. The diesters groups of the POE coupled
(d, J = 2.3 Hz, 1H), 6.97 (dd, J1 = 8.8 Hz, J2 = 2.4 Hz, 1H), 5.77 coumarin are removed by hydrolysis with aqueous NaOH in
(s, 1H), 4.22 (m, 2H), 4.00 (s, 3H), 3.78 (m, 2H), 3.61 (m, 2H), EtOH. The diethyl ester is first dissolved in ethanol (0.08 M)
3.55 (m, 2H), 3.52 (m, 2H), 3.43 (m, 2H), 3.24 (s, 3H). ESI-MS and aqueous NaOH (2.2 eq., 0.5 M in NaOH) added dropwise to
+
C
17H23O7 , calcd 339.144 m/z, found 339.141 m/z. Mp3C3: yield the stirred solution at room temperature. A precipitate gently
100%. 1H NMR d: 7.71 (d, J = 8.8 Hz, 1H), 7.01 (d, J = 2.3 Hz, 1H), appears after the NaOH addition. The reaction mixture is
6.97 (dd, J1 = 8.8 Hz, J2 = 2.6 Hz, 1H), 5.78 (s, 1H), 4.34 (m, 2H), further stirred for 30 minutes. The product is then isolated in
3.86 (m, 5H), 3.65 (m, 2H), 3.55 (m, 4H), 3.43 (m, 2H), 3.24 the form of the sodium dicarboxylate salt.
+
(s, 3H). ESI-MS C17H23O7 , calcd 339.144 m/z, found 339.143
The pure sodium dicarboxylate salt is obtained by recrystal-
1
m/z. Mp3C4: yield 50%. H NMR d: 7.82 (dd, J1 = 7.9 Hz, J2 = lisation of the precipitate in ethanol with a minimum amount
1.6 Hz, 1H), 7.68 (td, J1 = 7.6 Hz, J2 = 1.7 Hz, 1H), 7.40 (m, 2H), of water followed by filtration. The filtered solid is washed with
5.95 (s, 1H), 4.37 (m, 2H), 3.88 (m, 2H), 3.65 (m, 2H), 3.55 cold ethanol and collected as a wet powder. The powder is then
c
15992 Phys. Chem. Chem. Phys., 2013, 15, 15981--15994
This journal is the Owner Societies 2013

View Article Online
PCCP
Paper
dried in vacuo to obtain the desired ligand. It is characterised by i.e. tosylated diethylene glycol monomethyl ether for dp2C1 and
1H-NMR and 13C-NMR in D2O and by elemental analysis.
tosylated ethylene glycol monomethyl ether for dp1C1, accord-
Na2dp3C2: 1H NMR (400 MHz, D2O) d 7.48 (d, 1H, J = 8.4 Hz), ing to the same synthetic pathway.
7.20 (s, 2H), 6.80 (m, 2H), 5.43 (s, 1H), 4.18 (m, 2H), 4.00 (m,
Na2dp2C1: 1H NMR (400 MHz, D2O) d 7.45 (d, 1H, J = 8.8 Hz),
2H), 3.85 (m, 5H), 3.80 (m, 2H), 3.74 (m, 2H), 3.67 (m, 2H). 13
C
7.30 (s, 2H), 6.88 (d, 1H, J = 2.3 Hz), 6.82 (dd, 1H, J = 8.8, 2.4 Hz),
NMR (101 MHz, D2O) d 171.9 (Ccarboxylate), 168.0 (4CoumCar–O), 6.10 (s, 1H), 4.33 (m, 2H), 4.22 (m, 2H), 3.99 (m, 4H), 2.32
166.6 (4pyrCar–O), 166.0 (7CoumCar–O), 161.7 (2CoumCarQO), 154.2 (s, 3H). 13C NMR (HSQC) d 126.1 (5CoumCar–H), 113.9 (6CoumCar–
(2+6)pyr
(
(
C ), 153.6 ((8ꢀ1)-bridgeCoumCar), 124.0 (5CoumCar–H), 112.8 H), 111.2 ((3+5)pyrCar–H), 110.2 (8CoumCar–H), 101.7 (3CoumCar–H),
ar
6Coum
C –H), 111.1 ((3+5)pyrCar–H), 108.3 ((4ꢀ5)-bridgeCoumCar), 68.8 (OCH2–C–H2), 68.6 (OCH2–C–H2), 67.6 (OCH2–C–H2), 67.4
ar
100.7 (8CoumCar–H), 86.3 (3CoumCar–H), 69.9 (OCH2–C–H2), 69.8 (OCH2–C–H2), 17.9 (4CoumCar–C–H3). Na2C21H17NO9ꢂ2.0NaOH
(OCH2–C–H2), 68.7 (OCH2–C–H2), 68.6 (OCH2–C–H2), 67.4 (553.34): calcd C 45.58, H 3.46, N 2.53; found C 45.56, H 3.20,
1
(OCH2–C–H2),
67.3
(OCH2–C–H2),
56.6
(O–C–H3). N 2.77. Na2dp1C1: H NMR (400 MHz, D2O) d 7.57 (d, 1H, J =
Na2C23H21NO11ꢂ2NaOH (613.40): calcd C 45.04, H 3.78, N 8.9 Hz), 7.47 (s, 2H), 6.94 (d, 1H, J = 8.8 Hz), 6.90 (s, 1H), 6.11
2.28; found C 45.05, H 3.87, N 2.15. ESI-MS: C23H24NO11+ calcd (s, 1H), 4.53 (m, 2H), 4.48 (m, 2H), 2.34 (s, 3H). 13C NMR (HSQC)
490.1349 m/z, found 490.1348 m/z. Na2dp3C3: 1H NMR d 125.9 (5CoumCar–H), 113.1 (6CoumCar–H), 111.4 ((3+5)pyrCar–H), 110.3
8Coum
(400 MHz, D2O) d 7.27 (d, J = 8.9 Hz, 1H), 7.05 (s, 2H), 6.49
(
C –H), 101.8 (3CoumCar–H), 67.0 (OCH2–C–H2), 66.8 (OCH2–
ar
(dd, J = 2.5, 8.9 Hz, 1H), 6.33 (d, J = 2.4 Hz, 1H), 5.32 (s, 1H), 4.12 C–H2), 18.0 (4CoumCar–C–H3). Na2C19H13NO8ꢂ2.0H2O (465.33): calcd
(m, 2H), 3.99 (m, 2H), 3.87 (m, 2H), 3.79 (m, 2H), 3.74 (m, 2H), C 49.04, H 3.68, N 3.01; found C 49.02, H 3.69, N 2.94.
3.69 (m, 2H), 3.62 (s, 3H). 13C NMR (101 MHz, D2O) d 171.8
(Ccarboxylate), 167.1 (4CoumCar–O), 166.5 (4pyrCar–O), 165.9 (7CoumCar–
O), 162.7 (2CoumCarQO), 154.1 ((2+6)pyrCar), 153.8 ((8ꢀ1)-bridgeCoumCar),
Acknowledgements
124.0 (6CoumCar–H), 112.5 (5CoumCar–H), 110.9 ((3+5)pyrCar–H), 108.0
The authors would like to thank the group of photochemical
dynamics headed by Prof. Jacques Moser at the EPFL for
hosting and granting us access to lasers and time resolved
equipment, as well as Jelissa De Jonghe for her help with the
transient absorption setup.
(4ꢀ5)-bridgeCoum
8Coum
3Coum
(
C ), 99.9
ar
(
C –H), 86.9
ar
(
C –H),
ar
70.2 (OCH2–C–H2), 69.9 (OCH2–C–H2), 68.7 (OCH2–C–H2),
68.5 (OCH2–C–H2), 68.2 (OCH2–C–H2), 67.3 (OCH2–C–H2), 55.7
(O–C–H3). Na2C23H21NO11ꢂ2.3NaOH (625.40): calcd C 44.17, H
+
3.76, N 2.24; found C 44.19, H 3.62, N 2.52. ESI-MS: C23H24NO11
calcd 490.1349 m/z, found 490.1352 m/z. Na2dp3C4: 1H NMR (400
MHz, D2O) d 7.91 (m, 1H), 7.62 (m, 1H), 7.40 (m, 2H), 7.33 (m, 2H),
5.91 (m, 1H), 4.49 (m, 2H), 4.26 (m, 2H), 4.12 (m, 2H), 4.01
(m, 2H), 3.94 (m, 2H), 3.89 (m, 2H). 13C NMR (101 MHz, D2O) d
Notes and references
1 S. V. Eliseeva and J. C. G. Bunzli, New J. Chem., 2011, 35,
1165–1176.
2 S. V. Eliseeva and J. C. G. Bunzli, Chem. Soc. Rev., 2010, 39,
189–227.
3 K. Binnemans, Chem. Rev., 2009, 109, 4283–4374.
4 J. Andres and A.-S. Chauvin, Lanthanides: Luminescence, to be
published in: The Rare Earths: Fundamentals and Applications,
Wiley, Chichester, 2012.
5 D. Parker, R. S. Dickins, H. Puschmann, C. Crossland and
J. A. K. Howard, Chem. Rev., 2002, 102, 1977–2010.
6 J. P. Cross, M. Lauz, P. D. Badger and S. Petoud, J. Am.
Chem. Soc., 2004, 126, 16278–16279.
4Coum
4pyr
172.0 (Ccarboxylate), 166.8
(
C –O), 166.1
(
C –O),
ar
ar
166.0 (2CoumCarQO), 154.3 ((2+6)pyrCar), 152.0 ((8ꢀ1)-bridgeCoumCar),
133.0 (7CoumCar–H), 124.6 (6CoumCar–H), 122.8 (5CoumCar–H), 116.4
8Coum
(
C –H), 114.6 ((4ꢀ5)-bridgeCoumCar), 111.0 ((3+5)pyrCar–H), 89.4
ar
3Coum
(
C –H), 70.1 (OCH2–C–H2), 69.8 (OCH2–C–H2), 68.6 (2ꢃ
ar
OCH2–C–H2),
68.2
(OCH2–C–H2),
67.3
(OCH2–C–H2).
Na2C22H19NO10ꢂ0.85NaOHꢂ0.25CH3CH2OH (548.89): calcd
C
49.24, H 3.92, N 2.55; found C 49.24, H 3.97, N 2.48. ESI-MS:
+
C22H22NO10 calcd 460.1244 m/z, found 460.1243 m/z. Na2dp3C5:
1H NMR (400 MHz, D2O) d 7.45 (m, 1H), 7.08 (m, 2H), 7.01 (m, 2H),
5.56 (s, 1H), 4.22 (m, 2H), 4.04 (m, 2H), 3.92 (m, 2H), 3.82 (m, 2H),
3.76 (m, 2H), 3.72 (m, 2H). 13C NMR (101 MHz, D2O) d 171.7
7 D. L. Dexter, J. Chem. Phys., 1953, 21, 836–850.
8 T. Forster, Discuss. Faraday Soc., 1959, 7–17.
9 A. Olaya-Castro and G. D. Scholes, Int. Rev. Phys. Chem.,
2011, 30, 49–77.
4Coum
4pyr
(Ccarboxylate), 166.0
(
C –O), 165.9
(
C –O), 165.3
ar
ar
2Coum
(
C QO), 154.1 ((2+6)pyrCar), 152.0 ((8ꢀ1)-bridgeCoumCar), 138.1
ar
7Coum
5Coum
6Coum
(
(
C –Cl), 124.9
ar
(
C –H), 124.1
ar
(
C –H), 116.4
ar
C –H), 113.2 ((4ꢀ5)-bridgeCoumCar), 110.9 ((3+5)pyrCar–H), 89.4 10 V. May, Dalton Trans., 2009, 10086–10105.
8Coum
ar
3Coum
(
C –H), 70.2 (OCH2–C–H2), 70.0 (OCH2–C–H2), 68.9 (OCH2– 11 M. D. Ward, Coord. Chem. Rev., 2010, 254, 2634–2642.
ar
C–H2), 68.7 (OCH2–C–H2), 68.2 (OCH2–C–H2), 67.4 (OCH2–C–H2). 12 O. L. Malta, J. Lumin., 1997, 71, 229–236.
Na2C22H18ClNO10ꢂ0.4CH3CH2OH (556.25): calcd C 49.23, H 3.70, N 13 M. Kleinerman, J. Chem. Phys., 1969, 51, 2370–2381.
+
2.52; found C 49.30, H 3.68, N 2.30. ESI-MS: C22H21ClNO10 calcd 14 I. M. Clarkson, A. Beeby, J. I. Bruce, L. J. Govenlock,
494.0854 m/z, found 494.0855 m/z.
M. P. Lowe, C. E. Mathieu, D. Parker and K. Senanayake,
Synthesis of the dpxC1 ligands. The variation of the length
New J. Chem., 2000, 24, 377–386.
of the POE side chain was performed by replacing the tosylated 15 M. Latva, H. Takalo, V. M. Mukkala, C. Matachescu,
triethylene glycol monomethyl ether from the synthesis of the
dp3Cy ligands by the desired oligoethylene glycol derivative,
J. C. RodriguezUbis and J. Kankare, J. Lumin., 1997, 75,
149–169.
c
This journal is the Owner Societies 2013
Phys. Chem. Chem. Phys., 2013, 15, 15981--15994 15993

View Article Online
Paper
PCCP
16 N. M. Shavaleev, S. V. Eliseeva, R. Scopelliti and J. C. G. 33 P. R. Selvin, J. Jancarik, M. Li and L. W. Hung, Inorg. Chem.,
Bunzli, Chem.–Eur. J., 2009, 15, 10790–10802. 1996, 35, 700–705.
17 G. A. Hebbink, S. I. Klink, L. Grave, P. G. B. O. Alink and 34 G. Vereb, E. Jares-Erijman, P. R. Selvin and T. M. Jovin,
F. C. J. M. van Veggel, ChemPhysChem, 2002, 3, 1014–1018. Biophys. J., 1998, 74, 2210–2222.
18 J. R. G. Thorne, J. M. Rey, R. G. Denning, S. E. Watkins, 35 A. L. Gassner, C. Duhot, J. C. G. Bunzli and A. S. Chauvin,
M. Etchells, M. Green and V. Christou, J. Phys. Chem. A,
2002, 106, 4014–4021.
19 C. Yang, L. M. Fu, Y. Wang, J. P. Zhang, W. T. Wong, X. C. Ai,
Y. F. Qiao, B. S. Zou and L. L. Gui, Angew. Chem., Int. Ed.,
2004, 43, 5010–5013.
Inorg. Chem., 2008, 47, 7802–7812.
36 A. D’Aleo, A. Picot, A. Beeby, J. A. G. Williams, B. Le
Guennic, C. Andraud and O. Maury, Inorg. Chem., 2008,
47, 10258–10268.
37 H. Tadokoro, Y. Chatani, T. Yoshihara, S. Tahara and
S. Murahashi, Makromol. Chem., 1964, 73, 109–127.
38 P. G. de Gennes, Adv. Colloid Interface Sci., 1987, 27, 189–209.
20 J. C. G. Bunzli, A. S. Chauvin, H. K. Kim, E. Deiters and
S. V. Eliseeva, Coord. Chem. Rev., 2010, 254, 2623–2633.
21 A. S. Chauvin, F. Gumy, D. Imbert and J. C. G. Bunzli, 39 J. Andres and A. S. Chauvin, Eur. J. Inorg. Chem., 2010,
Spectrosc. Lett., 2004, 37, 517–532. 2700–2713.
22 M. T. Alonso, E. Brunet, C. Hernandez and J. C. Rodriguezubis, 40 R. M. Supkowski and W. D. Horrocks, Inorg. Chim. Acta,
Tetrahedron Lett., 1993, 34, 7465–7468. 2002, 340, 44–48.
23 M. T. Alonso, E. Brunet, O. Juanes and J. C. Rodriguez-Ubis, 41 M. H. V. Werts, R. T. F. Jukes and J. W. Verhoeven, Phys.
J. Photochem. Photobiol., A, 2002, 147, 113–125. Chem. Chem. Phys., 2002, 4, 1542–1548.
24 C. Feau, E. Klein, P. Kerth and L. Lebeau, Synth. Met., 2009, 42 D. S. Mcclure, J. Chem. Phys., 1949, 17, 905–913.
´
159, 528–536.
43 A. Aebischer, F. Gumy and J. C. G. Bunzli, Phys. Chem. Chem.
Phys., 2009, 11, 1346–1353.
44 B. R. Henry and E. A. Lawler, J. Mol. Spectrosc., 1973, 48,
117–123.
45 C. Schweitzer and R. Schmidt, Chem. Rev., 2003, 103,
1685–1757.
46 C. B. Nielsen, M. Johnsen, J. Arnbjerg, M. Pittelkow,
S. P. McIlroy, P. R. Ogilby and M. Jorgensen, J. Org. Chem.,
2005, 70, 7065–7079.
25 J. C. Rodriguez-Ubis, M. T. Alonso, O. Juanes and E. Brunet,
Luminescence, 2000, 15, 331–340.
26 S. G. Roh, N. S. Baek, K. S. Hong and H. K. Kim, Bull. Korean
Chem. Soc., 2004, 25, 343–344.
27 P. G. Tarassoff and N. Filipescu, J. Chem. Soc., Chem.
Commun., 1975, 208–209.
´
28 C. Feau, E. Klein, C. Dosche, P. Kerth and L. Lebeau, Org.
Biomol. Chem., 2009, 7, 5259–5270.
´
29 C. Feau, E. Klein, P. Kerth and L. Lebeau, Bioorg. Med. Chem. 47 A. Beeby, S. Faulkner, D. Parker and J. A. G. Williams,
Lett., 2007, 17, 1499–1503. J. Chem. Soc., Perkin Trans. 2, 2001, 1268–1273.
30 E. Pershagen, J. Nordholm and K. E. Borbas, J. Am. Chem. 48 S. Kohmoto, E. Mori and K. Kishikawa, J. Am. Chem. Soc.,
Soc., 2012, 134, 9832–9835. 2007, 129, 13364.
31 J. Y. Chen and P. R. Selvin, J. Photochem. Photobiol., A, 2000, 49 J. C. Jung, Y. J. Jung and O. S. Park, Synth. Commun., 2001,
135, 27–32. 31, 1195–1200.
32 M. Li and P. R. Selvin, J. Am. Chem. Soc., 1995, 117, 50 J. I. Luengo and J. D. Elliott, 2002. U.S. Pat., 6353116, filed
8132–8138.
Dec. 14, 1999.
c
This journal is the Owner Societies 2013
15994 Phys. Chem. Chem. Phys., 2013, 15, 15981--15994