Europium complexes of tris(dipicolinato) derivatives coupled to methylumbelliferone: A double sensitization

FULL PAPER

DOI: 10.1002/ejic.201000126

Europium Complexes of Tris(dipicolinato) Derivatives Coupled to

Methylumbelliferone: A Double Sensitization

Julien Andres^[a]and Anne-Sophie Chauvin*^[a]

Dedicated to Professor Jean-Claude Bünzli on the occasion of his 65th birthday

Keywords: N, O ligands / Lanthanides / Luminescence / Europium / Sensitization

A para-poly(oxyethylene) dipicolinic acid derivative coupled

to methylumbelliferone was synthesized and used as a ligand

for coordination with europium ions. Characterization of the

complex by mass spectrometry and spectrophotometry was

performed in aqueous solution and points to the formation of

the 1:3 Eu/L complex. Physicochemical properties are close

to those of already reported dipicolinato complexes, confirm-

ing that the presence of a coumarin chromophore does not

interfere with complex formation. Photophysical measure-

ments result in observed lifetimes of 1.4 ms and intrinsic

quantum yields, Φ_Eu^Eu, ranging from 33 to 50% (depending

on the solvent used). Selective excitation can be performed

either mostly on the dipicolinate (DPA) backbone (at 280 nm)

or exclusively on the coupled coumarin chromophore in the

near visible range (320 nm). Despite the spatial distance be-

tween the antenna and the lanthanide ion, europium emis-

sion was observed upon 320 nm excitation together with the

coumarin emission. The different sensitization pathways

have been explored and rationalized. Finally, the complex

has been used to probe the ratio of a binary mixture of sol-

vent, establishing the potential interest of this new class of

ligands in many fields of research, from photophysics to ap-

plied chemistry.

overall quantum yields.^[20,21]Near-infrared luminescence

from Nd^III, Dy^III, Er^III, Tm^III, and Yb^IIIis also sensitized

by dipicolinate moieties,^[9,22,23]although usually to a lesser

extent than visible emission. For all these reasons, this fam-

ily of complexes was used in several applications, such as

various inorganic–organic hybrid materials,^[24]nanopar-

ticles,^[25]or as analytical probes, particularly for the analysis

of bacterial spores^[26–28]or of nanomolar concentrations of

Introduction

2,6-Pyridinedicarboxylic acid, also known as dipicolinic

acid (DPA) or dipic, is a well-known framework in lantha-

nide coordination chemistry, which forms stable com-

plexes^[1,2]with interesting luminescence properties. The

structures of these complexes have been deduced by using

X-ray diffraction analysis^[3–6]and by analysis of the lantha-

nide-induced shifts (LIS).^[7–10]This small ligand provides a

good sensitization of most of the lanthanide ions, with a

particularly high quantum yield for the europium and ter-

bium complexes,^[11]the sensitization occurring through the

dipic^2–triplet state^[12](with an efficiency of 85% in the solid

state^[13]and 61% in solution^[11]). Apart from the parent di-

pic compound, the pyridine ring can be functionalized, in

particular at its para position, the influence of the grafted

substituent being well documented by several authors.^[14–19]

These couplings have often resulted in a decrease in lumi-

nescence efficiency relative to dipicolinic acid, due to a mis-

match of the triplet states of the complexed ligands with

the excited states of the lanthanides. However, promising

results have also been achieved with substantial increase in

Tb^{III [29]}

Finally, lanthanide dipicolinates display nonlinear

.

optical (NLO) properties^[30,31]and two-photon excitation

luminescence, either through an allowed Tb^IIIf–f transi-

tion^[32]or through the ligand,^[33,34]enabling luminescence

microscopy images to be collected under NIR excitation.

Despite these extensive studies and applications developed

with dipic complexes, the potentialities of these kinds of

ligands still have to be extended, and several routes can be

exploited to improve their capability as luminescent

markers. Above all, dipicolinic acid has a maximum of exci-

tation that is limited to the UVC region (λ_max= 280 nm),

and extending it to the near visible region is of importance.

Recently, coupling polyoxyethylene dipicolinic acid deriva-

tives with biological materials has been successfully under-

taken.^[18,35]The presence of a polyoxyethylene chain at the

fourth position of the pyridine ring has been shown to

slightly affect the coordination ability of dipicolinic acid de-

rivatives relative to their non-polyoxyethylene analogs and

has demonstrated an interesting fine tuning of the lumines-

[a] École Polytechnique Fédérale de Lausanne, ISIC, BCH 1405,

1015 Lausanne, Switzerland

E-mail: anne-sophie.chauvin@epfl.ch

julien.andres@epfl.ch

Supporting information for this article is available on the

WWW under http://dx.doi.org/10.1002/ejic.201000126.

2700

Eur. J. Inorg. Chem. 2010, 2700–2713

Europium Complexes of Tris(dipicolinato) Derivatives

cence properties depending on the coupled molecule at the cal properties of the europium complex both in aqueous

extremity of the polyoxyethylene chain.^[18]This change in solution and in the solid state will be described in an at-

luminescence properties is not straightforward, even if some tempt to rationalize the mode of sensitization by these new

of the results can be rationalized by the introduction of types of ligands.

known quenching functional groups such as free alcohol or

amine groups.

Results and Discussion

Motivated by these interesting results, we decided to at-

Synthesis of the Ligand and Its Europium Complex

tach chromophores or fluorophores at the end of the poly-

oxyethylene chain in order to test whether we could sensit-

ize lanthanide ions through space with an uncoordinated

chromophore. The presence of two different chromophores,

the dipicolinate moiety, and the additional coupled chro-

mophore provides an unusual and novel challenge in under-

standing the photophysics of luminescent lanthanide com-

plexes. Furthermore, the interaction of the two chromo-

phoric units with each other and with the lanthanide ion

might find some useful applications in various domains,

from bioimaging to color reproduction. The choice of the

second chromophore was critical. The first requirement was

that it had to absorb at higher wavelengths than the dipico-

linic moiety, whose absorption occurs below 300 nm. On

the other hand, the absorption had to remain in the UV

range to ensure that the emission would be not too far away

from the europium and terbium emissive excited states. Sim-

ple coumarin derivatives, which have absorptions below

380 nm and emission maxima around 400 nm, have been

found to be good candidates. Moreover, a well-documented

effect of the substitution of these derivatives relative to the

effect on the fluorescence emission is available in the litera-

ture, allowing a fine tuning of the emission and absorption

maxima.^[36–38]Due to its commercial availability, low price,

and matching photophysical properties, the candidate cho-

sen first was 4-methylumbelliferone (4-methyl-7-hy-

droxycoumarin), coupled through its 7-hydroxy group (Fig-

ure 1). In such structures, the maximum distance between

the Eu³⁺ion and the coumarin chromophore is around

17 Å. We report here the synthesis of para-polyoxyethylene

dipicolinic acid derivatives coupled to methylumbelliferone

as ligand for luminescent lanthanide ions. The photophysi-

The strategy used to synthesize ligand H₂DPApC is de-

scribed in Figure 2. Coupling the coumarin chromophore

to a methoxypolyoxyethylene chain is first achieved by re-

acting methylumbelliferone (1) with brominated methoxy-

polyoxyethylene chain 2 obtained from the alcohol deriva-

tive. The methoxy group of the CpOMe chromophore is

then removed by TMSI (trimethylsilyl iodide) to yield free

alcohol 3. Coupling of 3 with diethyl chelidamate (4) is un-

dertaken through a Mitsunobu reaction. The last step con-

sists of a deprotection of the carboxylic acid moieties of 5

to yield ligand H₂DPApC. The global yield of the reaction

is 45%. The sodium salt of the ligand is water-soluble; it is

mixed with europium(III) perchlorate in a 1:3 Eu/L ratio to

form europium complex [Eu(DPApC)₃]^3–.

Figure 2. Synthetic strategy for the synthesis of ligand H₂DPApC

and its europium complex [Eu(DPApC)₃]^3–.

Stoichiometry and Stability of the Ln Complexes

Mass Spectrometry of the Europium Complex

In order to confirm the formation of a major amount of

[Eu(DPApC)₃]^3–complex in 1:3 stoichiometry, a sample of

Figure 1. Ligands derived from dipicolinic acid.

Eur. J. Inorg. Chem. 2010, 2700–2713

www.eurjic.org

2701

J. Andres, A.-S. Chauvin

FULL PAPER

concentration 1ϫ10^–4 in Eu³⁺and 3ϫ10^–4 in ligand formation of a luminescent 1:3 complex under stoichio-

was examined by mass spectrometry with electron spray metric conditions. Additionally, formation of a precipitate

ionization in the negative mode. The experiment shows above 0.66 equiv. was observed. The absorbance of the li-

series of peaks consistent with the 1:3 stoichiometry of the gand was also monitored as a function of the amount of

desired complex [Eu(DPApC)₃]^3–: half the exact mass of the Eu³⁺(Supporting Information, Figure S2), but precipi-

monoprotonated complex ([Eu(DPApC)₃]^3–, m/z = 783.67) tation did occur above 0.66 equiv. Eu³⁺, so that no stability

and half the exact mass of the monoprotonated complex constant data could be extracted from these experiments

with a water molecule ({[Eu(DPApC)₃]^3–+ H₂O}, m/z = and hence no correlation can be established with the dipic

792.68, see Supporting Information, Figure S1).

ligand. However, according to previous results for DPApR

ligands,^[18]it was established that stability constants of the

Eu complexes were close to those of dipic. It seems then

reasonable to consider that the DPApC ligand has the same

lanthanide complexation ability.

Emission and Absorption As a Function of Europium-to-

Ligand Ratio

The stability of the complex is often critical for good

luminescence properties. In the case of 1:3 complexes such

as dipicolinato derivatives, 1:2 and 1:1 species have poor Emission as a Function of pH

luminescence due to the quenching by water molecules in

Dealing with luminescent materials in aqueous solution

the first coordination sphere. The appearance of these spe-

cies at higher than 1:3 stoichiometries thus decreases the

emission intensity of the luminescent lanthanide ion. Moni-

toring the intensity of the europium emission as a function

of europium-to-ligand ratio hence allows observing the

emergence and vanishing of the species with the strongest

luminescence (the 1:3 species). The excitation wavelength

was chosen at the coumarin absorption maximum, that is,

320 nm. This wavelength would also allow monitoring the

emission intensity of the coumarin moiety as a function of

ligand-to-europium ratio. Complexation with europium ion

might indeed alter the coumarin emission. A titration of a

3ϫ10^–4 solution of the ligand with increasing amounts

of europium perchlorate were undertaken. The pH value of

the solution was set at 7.4 with a Tris buffer, in a manner

similar to other studies with dipicolinato complexes. Fig-

ure 3 shows the emission intensity of the coumarin and sen-

sitized emission from europium upon addition of europium

perchlorate. The emission of europium exhibits a nice bell-

shaped curve with a maximum in the presence of 0.3 equiv.

Eu³⁺. On the other hand, the coumarin emission resembles

a sigmoid with a constant emission starting from 0.33 equiv.

europium ions. These results are then consistent with the

requires finding out the optimum pH values. In order to

establish the best conditions for luminescence, the emission

spectrum of an aqueous solution of [Eu(DPApC)₃]^3–was

monitored as a function of pH. Two excitation wavelengths

were chosen. The first one at 280 nm is centered on the

absorption of the DPA and coumarin moieties. The second

one at 320 nm is at the absorption maximum of the couma-

rin chromophore, where DPA does not absorb any light.

The emission of the coumarin moiety has a maximum at

389 nm. Upon excitation at 280 nm, its intensity slightly in-

creases as pH increases from 2.5 to 10.4. On the other hand,

upon excitation at 320 nm, a significant increase of 175%

in intensity is observed. In parallel, the emission intensity

of the europium ion is higher when excited at 280 nm rather

than at 320 nm. All these results confirm that coumarin is

highly emissive at λ_ex= 320 nm, whereas the luminescence

of europium is more intense than that of coumarin at λ_ex

=

280 nm.

Above pH 6.3, the difference in intensity between 280

and 320 nm excitation is quite constant and close to a ratio

of 1.8. This is also dependent on the pH (Figure 4). Below

pH 6, the ratio is not constant anymore, reflecting that cou-

marin emission is less affected by a fairly acidic medium,

Figure 3. Normalized integrated emission of europium (red) and

coumarin (blue) emission upon 320 nm excitation and upon ad-

dition of europium perchlorate to

a

3ϫ10^–4



solution of

Figure 4. Peak intensity of the [Eu(DPApC)₃]^{3– 5}D₀Ǟ⁷F₂transition

in aqueous solution as a function of pH.

(DPApC)^2–in Tris (0.1 , pH 7.4).

2702

www.eurjic.org

Eur. J. Inorg. Chem. 2010, 2700–2713

Europium Complexes of Tris(dipicolinato) Derivatives

whereas europium emission depends on the stability of the the polyoxyethylene chain removes part of the complexity

complex (with competition between complexation and pro- of the coumarin behavior that comes from this proton. The

tonation of the carboxylic acid functions at low pH values). excitation and absorption spectra in Tris (0.1  pH 7.4)

For comparison with already reported results within this aqueous solution (Figure S5) show good overlap once nor-

series, we fixed the pH at 7.4 (by means of Tris buffer solu- malized, with a maximum absorption at 320 nm and ab-

tion) for all the following experiments.

sorption up to 360 nm. Fitting of the molar extinction coef-

ficients with Gaussian peaks (Figure S6) is achieved with

three Gaussian peaks centered at 291 nm, 320 nm, and

336 nm. As it will be shown later, these three peaks can be

attributed to π–π* and n–π* transitions.

Emission Spectrum of the [Eu(DPApC)₃]^3–Complex

The emission spectrum of the [Eu(DPApC)₃]^3–complex

has the same shape as those of other DPApR complexes

where R = OH, NH₂, OMe, or phthalimide^[18](Figures 7

and S3), with the same relative intensities, so that it can be

established that the coordination spheres around the euro-

pium cation are equivalent for all the dipicolinato ligands.

This also indicates that the presence of the coumarin moiety

does not interfere with the coordination sphere.

On the other hand, the excitation spectrum in the powder

solid state is quite different from that in solution (Figure 5).

The shape is broader with local maxima at 275 nm and

353 nm, with a curvature in between, and a shoulder

around 390 nm. As a consequence, excitation can take place

at higher wavelength in the solid state, up to 400 nm,

whereas in solution, excitation only up to 360 nm is pos-

sible. This additional band with a local maximum at 353 nm

is certainly due to intermolecular interaction in the solid

state. The shape of the spectrum is hence very different in

the solid state, because of all the important changes in the

surroundings of the molecules once the solvent is removed.

The three peaks fitted in Figure S6 are thus hidden in the

broad structure of the solid-state excitation spectrum. The

peak at 291 nm may appear with a slight hypsochromic shift

in the solid state at 275 nm, whereas the other two are less

well-defined. Additionally, the excitation below 300 nm is

much more efficient in the solid state with a maximum at

275 nm relative to 320 nm in solution.

Furthermore, lifetime measurements both in water, τ_obs

-

(H₂O) = 1.4Ϯ0.1 ms, and in deuteriated water, τ_obs(D₂O)

= 2.6Ϯ0.1 ms, lead to the conclusion that no water mole-

cule is in the first coordination sphere (q = 0.02), by using

the Horrocks phenomenological relations [Equations (8)

and (9) in the Experimental Section] with α = 0.31 ms^–1and

A = 1.11 ms, and zero XH oscillators in the first coordina-

tion sphere.^[39]This last result confirms that the coordina-

tion sphere of the europium ion is filled by the three triden-

tate dipicolinate entities, the lifetimes obtained being close

to those previously obtained within the dipicolinato series.

5

High-Resolution Excitation of the D₀Ǟ⁷F₂Transition

A high-resolution excitation spectrum was measured on

the powder solid sample at 12 K by direct excitation on the

⁵D₀ǟ⁷F₀transition. The emission intensity of the ⁵D₀Ǟ⁷F₂

transition at 615 nm was monitored as a function of the

excitation wavelength. The resulting spectrum (Figure S4)

shows a single peak located at 17223 cm^–1(580.61 nm) with

a half-width of 17.9 cm^–1. Hence, only one europium geom-

etry is present in the solid-state sample, which is in all prob-

ability similar to that of Cs₃[Eu(dipic)₃] (D₃symmetry). All

these results are in agreement with the hypothesis of the

similarity between the coordination of the europium ion to

DPApC ligands and that to DPApR or dipic ligands.

Photophysical Properties

Figure 5. Excitation (λ_em= 420 nm) and emission [λ_ex= 275 nm

(s)/320 nm (aq.)] spectra of CpOMe in the solid state (blue curve)

and in Tris (0.1 , pH 7.4) aqueous solution (cyan curve).

Luminescence of the Uncoupled Chromophore CpOMe

To begin with, the luminescence properties of the free

methoxypolyoxyethylene coumarin chromophore CpOMe

(Figure 1) were measured in aqueous solution [saturated

For the emission spectra, both solution and solid-state

aqueous solution diluted ten times in Tris (0.1 , pH 7.4)] samples exhibit a maximum at 384 nm. Nevertheless, this

and in the solid state. The addition of the polyoxyethylene emission goes up to 510 nm in solution and 560 nm in the

chain ensures that the chromophore is in the same confor- solid state. The shape is then slightly different in the solid

mation as it is once coupled to DPA. Indeed, the uncoupled state compared to the smooth emission from solution. The

4-methylumbelliforone can exist in different protonation quantum yield for this emission of CpOMe is around 38%

states depending on the pH (pK_aof the 7-OH group is (Ϯ 4%) upon excitation at 320 nm (maximum absorption)

around 7.4), different tautomeric forms, or different in solution and 31% (Ϯ 3%) in the solid state, so that no

charged structures.^[40]Most of these properties are due to significant difference can be found here due to experimental

the acidic proton at the 7th position. Hence, coupling with error. It will be demonstrated below that these emission and

Eur. J. Inorg. Chem. 2010, 2700–2713

www.eurjic.org

2703

J. Andres, A.-S. Chauvin

FULL PAPER

excitation spectra are retrieved in the DPApC ligand and in significantly reabsorbed by the ligand. As the maximum of

the lanthanide complexes together with additional emission the emission is located near this region, reabsorption shifts

and excitation components arising from the DPA backbone. it towards higher wavelengths.

Luminescence of the DPApC Ligand and Its Europium

Complex

Excitation and emission spectra of [Eu(DPApC)₃]^3–

1ϫ10^–4 in Tris (0.1  pH 7.4) and the solid Cs₃[Eu-

(DPApC)₃] powder were measured at room temperature

and at 77 K (10% glycerol added to the aqueous solution).

Considerable differences in the excitation spectra were ob-

served between the aqueous solution and the solid state

(Figure 6). First of all, the absorption bands of the direct

5

f–f transition are observed. The most intense one is the L₆

5

(excitation at 395 nm); the D₂at 464.5 nm is also rather

5

strong, whereas the D₃and D₁are less intense. Then, the

absorption band of the coumarin chromophore is also dif-

ferent in the solid state relative to that in aqueous solution.

The absorption range is extended to the visible range up to

Figure 7. Emission spectra (λ_ex= 320 nm, room temperature) of

aqueous solution of [Eu(DPApC)₃]^3–(1ϫ10^–4 in Tris 0.1 , pH

7.4, blue curve) and solid-state sample of Cs₃[Eu(DPApC)₃] (red

curve) normalized by the maximum emission intensity from the

Eu³⁺ion (⁵D₀Ǟ⁷F₂transition).

5

the peak of the L₆ǟ⁷F₀and D₃ǟ⁷F₀transitions.

As seen in Figure 7, the emission of the coumarin is also

greatly decreased in the solid state, relative to the emission

of europium. The maximum emission of the coumarin in

5

aqueous solution is up to 50% of the D₀Ǟ⁷F₂europium

transition, while it is less than 10% in the solid state. The

emission peaks from the coumarin and europium can be

integrated (Table S1). Europium emission represents 48%

of the total emission in the solid state, which means that

coumarin and europium emit a similar amount of photons

(the ratio of coumarin emission to europium emission is

1.1). In aqueous solution, europium emission represents

only 12% of the total emission. The coumarin then emits

7.4 times more photons than europium. With respect to the

europium emission, the only structural difference between

the solid sate and the solution is due to the ⁵D₀Ǟ⁷F₁transi-

tion, which is split in two in aqueous solution, consistently

with the emission from tris(dipicolinato)europium.^[13]This

Figure 6. Excitation spectra of [Eu(DPApC)₃]^3–aqueous solution

 in Tris 0.1 , pH 7.4, blue curve) and Cs₃[Eu-

(DPApC)₃] solid state sample (red curve), both at room tempera-

(1ϫ10^–4

ture with λ_em= 615 nm.

Regarding the emission spectrum upon 320 nm exci- difference in relative intensity between europium and cou-

tation (Figure 7), the major difference between the solution marin induces a significant shift in color. Under excitation

and the solid state is provided by the emission of the cou- with a long-wave UV lamp (366 nm), the solid state sample

marin. A significant bathochromic shift of the emission looks magenta, whereas in solution, the color has a stronger

band of the coumarin is observed in the whole excitation blue component and looks more lavender-colored. Upon

range. The maximum is located at 389 nm in aqueous solu- excitation at 254 nm (short-wave UV lamp), the color be-

tion and 411 nm in the solid state. In addition, the ⁵L₆ǟ⁷F₀comes nearly as red as Cs₃[Eu(dipic)₃], most of the blue

transition is also slightly visible in the emission spectrum in component from coumarin being lost.

the solid state. This is due to the reabsorption of the couma-

The excitation spectra with λ_em= 615.5 nm, either in the

rin emission by the europium ion (by a direct f–f transition, solid state or in aqueous solution, are different from the

no energy transfer involved), thus reinforcing the emission ones for the emission at 411 nm (solid) or 389 nm (aqueous

of the sensitized europium. The bathochromic shift of the solution). The ratio of the europium maximum emission

coumarin emission in the solid state may also be due to peak (⁵D₀Ǟ⁷F₂transition, 615.5 nm) and to that of the

a reabsorption-linked phenomenon. Indeed, the excitation coumarin (411 nm) was measured from the excitation spec-

spectra of the solid-state sample have shown that absorp- tra of the solid-state sample (Figure S7), and the results are

tion can go up to 400 nm, with strong absorption still oc- presented in Figure S8. A fairly constant ratio is observed

curring at 380 nm. Consequently, part of the emission of from 310 to 340 nm, where the absorption range of the cou-

the coumarin that would take place below 380 nm can be marin chromophore is mainly located. Below 310 nm, the

2704

www.eurjic.org

Eur. J. Inorg. Chem. 2010, 2700–2713

Europium Complexes of Tris(dipicolinato) Derivatives

absorption of the DPA moiety becomes significant in the 440 up to 630 nm, with a vibrational progression close to

solid samples, whereas at higher wavelengths, the additional 1200 cm^–1, which is typical of Gd complexed with aromatic

absorptions, which probably come from intermolecular in- backbones. The high-energy band is thus attributed to a

teractions as well as f–f transitions, are responsible for the singlet state emission from the ligand, whereas the band at

europium emission. This also rationalizes the red color un- lower energy upon time resolution is attributed to a triplet

der the 254 nm UV lamp, as the ratio of coumarin to euro- state of the ligand. Compared to DPApR series,^[18]the trip-

pium emission in Figure S8 falls down to 0.1 at 250 nm ex- let of the DPApC ligand is redshifted by approximately

citation.

To summarize, there are quite important differences in

90 nm.

photophysical properties between powder solid state sam-

ples and those in solution. Many of these differences arise

from the difference in molecular surroundings and induce

modulations of both excitation and emission characteristics.

The photophysical behavior in aqueous solution will now

be deeply investigated and presented in order to better

understand the mechanism, and thus the efficiency and

shortcomings of such a molecular system for europium sen-

sitization.

Europium-Centered Photophysical Properties

In the presence of an emissive lanthanide, the character-

istic metal-centered emission lines arising from the Eu³⁺

⁵D₀excited states are observed upon excitation at 280 nm

or 320 nm. The broad band at 384 nm is, however, still pres-

ent and accounts for about 88% of the Eu^IIIcomplex signal

(total emission) at room temperature and 57% at 77 K.

When the luminescence decays are sampled with short time

intervals, the function is monoexponential, reflecting life-

times for the Eu^{3+ 5}D₀excited states of 1.4Ϯ0.1 ms at room

temperature and 2.2Ϯ0.1 ms at 77 K. These values are

comparable to those recorded within the DPApR series.

The temperature dependence of the lifetimes reflects the

presence of some vibrational quenching processes of the Eu

excited state by the ligand backbone, as already observed

with DPApR series.^[18]Quantum yields for these emissions

at room temperature were measured in an integration

sphere.^[41]Lifetimes and emission spectra were then used to

calculate the radiative lifetime as well as the intrinsic quan-

tum yield and sensitization efficiency of the europium emis-

sion, according to Equations (5) to (7) in the Experimental

Section. The results are presented in Table 1.

Sensitization Pathway

Ligand-Centered Photophysical Properties

The excitation and emission spectra of the free ligand

DPApC, together with its gadolinium and europium com-

plexes are presented in Figure 8. Excitation at 320 nm of

(DPApC)^2–or [Gd(DPApC)₃]^3–results in a ligand-centered

emission, displaying one broadband with a maximum

around 26000 cm^–1(384 nm), which sustains a hypsoch-

romic shift of about 15 nm upon decreasing the temperature

at 77 K. At 77 K, a faint second band is observed, which

becomes exclusive upon time resolution with a 50 µs delay.

For the nonemissive Gd complex, this band is more intense

and makes the room temperature spectrum look broader.

Upon time resolution, the same spectrum as that of the free

ligand is observed with a maximum at 490 nm [(0-phonon

component at 463 nm, (21600 cm^–1)], and it extends from

Table 1. Quantum yields, Φ_L^L+Ln, of the ligand and lanthanide

emission upon ligand excitation, Φ_L^L, of the ligand residual emis-

sion; intrinsic quantum yields, Φ_Ln^Ln, of the lanthanide ion upon

lanthanide excitation; observed lifetimes, τ_obs^Ln, of the lanthanide

emission and radiative lifetimes, τ_r^Ln, of the lanthanide ion; sensiti-

zation efficiency, η_sens, of the ligand and its gadolinium and euro-

pium complexes in Tris (0.1 , pH 7.4) in aqueous solution except

for entry marked with [a] (solid-state powder). Error in

τ_obs:Ϯ0.1 ms; 10% relative error in the other values; λ_ex= 320 nm.

[Eu(DPApC)₃]^3–

Cs₃[Eu(DPApC)₃]^[a][Gd(DPApC)₃]^3–DPApC

L+Ln

L

Φ_L

0.113

0.097

0.016

0.141

0.093

0.048

1.4

2.8

0.50

0.097

0.115

–

0.316

Ln

0

–

τ_obs^Ln/ms 1.4, 2.2^[b]

τ_r^Ln/ms

4.2

0.33

0.048

Ln

Φ_Ln

η_sens

[a] Solid state powder; refractive index n = 1.517. [b] 77 K.

Interestingly, the emission quantum yield of the uncoor-

dinated ligand is much higher than either its gadolinium or

its europium complex, 31.6% of the absorbed photon being

reemitted in aqueous solution. The gadolinium complex re-

emits only 11.5% of the absorbed photons and the euro-

pium complex 11.3%, whereas 9.7% of the absorbed pho-

tons are emitted by the coumarin moiety and only 1.6% by

Figure 8. Excitation (black curves) and emission spectra (λ_ex

=

320 nm) of the uncoordinated ligand as well as its gadolinium and

europium complexes at room temperature (solid curves) and 77 K

(solid cyan and dotted magenta curves). Emission from the ligand

triplet state (dotted purple curves) recorded 50 µs after 320 nm ex-

citation pulses. 1ϫ10^–4 in Ln³⁺ion, 3ϫ10^–4 in ligand in Tris

(0.1  aqueous solution, pH 7.4) and 10% glycerol for frozen solu-

tions.

Eur. J. Inorg. Chem. 2010, 2700–2713

www.eurjic.org

2705

J. Andres, A.-S. Chauvin

FULL PAPER

the europium ion. These decreases in quantum yield upon resulting transfer to the europium ion is then much more

lanthanide coordination can be accounted either by struc- temperature-dependent than the excitation at 280 nm,

tural changes of the ligand upon complexation, affecting its where most of the excitation is on the DPA moiety.

photophysical properties, or by the influence of the coordi-

nating ion on the photophysical properties, such as on the

relaxation of the forbidden triplet state formation (intersys-

tem crossing, ISC). Indeed, the highly paramagnetic gado-

linium ions induce a higher ISC rate and hence increase the

population of the triplet state.^[42,43]Nevertheless, as triplet

states are by essence long-lived, quenching of the triplet

states can be a quite effective way of nonradiative deactiva-

tion. Most of the energy stored in the triplet state would

then be lost by nonradiative deactivation, and hence, the

overall quantum yield for the mixed emission from singlet

(fluorescence) and triplet (phosphorescence) states would

be decreased as a result of the decrease in fluorescence

quantum yield that is related to the increase in the ISC rate.

For [Eu(DPApC)₃]^3–, part of the triplet state energy is

transferred onto the europium ion and gives rise to the ob-

served characteristic europium emission. The percentage of

each component responsible for the total emission was esti-

Figure 9. Smoothed (FFT filter, 5% cutoff) normalized excitation

spectra for the emission of the ⁵D₀Ǟ⁷F₂europium transition at

mated from the integration ratio of each part of the emis-

sion spectrum and amounts to 58.8% for the singlet emis-

sion, 27.0% for the triplet emission, and 14.2% for the eu-

ropium emission, at room temperature. The phosphores-

room temperature (r.t.) and at low temperature (77 K).

To better understand the sensitization pathway, low-tem-

cence was extracted by subtracting the normalized emission perature, time-resolved measurements were carried out on

spectrum of the uncoordinated ligand, which shows very the uncoordinated ligand and its gadolinium complex, with

little or no phosphorescence at room temperature, from the excitations at 320 and 270 nm (Figure 10). At 77 K, a sec-

normalized emission of the europium complex. The nor- ond phosphorescence band centered at 410 nm appears

malization was done with respect to the maximum of the upon excitation at 270 nm, which is not observed upon exci-

coumarin emission at 389 nm. Most of the emission arises tation at 320 nm. As a comparison, the scaled phosphores-

from the singlet state with more than half of the intensity cence spectrum of the dipicolinate gadolinium complex

emitted as fluorescence. Nevertheless, a significant contri- [Gd(dipic)₃]^3–is also shown in Figure 10 (green line): its

bution to the emission from the triplet state is observed triplet emission is located at the exact position of the

(27%), certainly accounting for the important thermally as- higher-energy triplet emission. This could indicate that this

sisted back-transfer from the excited europium ion to the new band comes from the DPA moiety of the DPApC li-

ligand triplet state. Finally, a fair 14.2% comes from the gand, whereas the other triplet emission that is visible upon

europium ion. These results then suggest that the low quan- both excitations arises from the coumarin moiety. However,

tum yield of the europium emission comes from a low net the same phenomenon was also observed when measuring

energy transfer rate due to significant back-transfer.

the polyoxyethylene coumarin chromophore CpOMe (Fig-

Compared to the [Eu(dipic)₃]^3–complex, the dipic com- ure 11), meaning that the high energy triplet emission can,

Eu

plex has a higher quantum yield (Φ_dipic

= 24% vs. in fact, come from either the coumarin moiety or the coor-

Φ_DPApC^Eu= 1.6%). Nevertheless, two important factors are dinating DPA.

to be taken into account. First, the excitation of the dipic

The most rational explanation would be that two triplet

complex can only take place below 300 nm, and second, its excited states can be populated on the coumarin chromo-

extinction coefficient is also smaller with an ε₂₇₀([Eu- phore, depending on the excitation wavelength. The one

(dipic)₃]^3–) of 16470 lmol^–1cm^–1at 270 nm vs. ε₃₂₀([Eu- higher in energy can be reached by an excitation at 270 nm

(DPApC)₃]^3–) = 25286 lmol^–1cm^–1at 320 nm. As a conse- and is located at the same energy as the DPA triplet state.

Eu

quence, Φ_dipic^Eu/Φ_DPApC= 15, while the εΦ_Lratio is The second one is reached at 320 nm, where no excitation

only a factor of 10, which makes DPApC an interesting of the DPA moiety can take place. Its energy is lower than

candidate, considering the gain of 50 nm in excitation wave- the one excited at 270 nm, but can also be populated upon

length.

excitation at 270 nm. This is perfectly coherent, because of

As observed from Figure 9, the difference in europium the internal conversion from higher excited state to lower

intensity between room temperature and low temperature excited state.

depends on the excitation wavelength. The difference in

To confirm this hypothesis, calculations of the molecular

emission intensity is maximal upon excitation around orbitals with INDO/S parameters after optimizing the ge-

320 nm, whereas it is much weaker upon excitation at ometry with ZINDO INDO/1 calculations were performed

280 nm. The excitation of the coumarin at 320 nm and the with the Scigress Explorer software.^[44]Two absorption

2706

www.eurjic.org

Eur. J. Inorg. Chem. 2010, 2700–2713

Europium Complexes of Tris(dipicolinato) Derivatives

284 nm. Similar calculations performed on the DPApOMe

ligand demonstrated that one major π–π* transition (pre-

dicted at 272.5 nm) is responsible for the absorption around

270 nm (Supporting Information, Figure S10).

The wavelength-dependent difference between low-tem-

perature and room-temperature excitation spectra can then

be rationalized according to Figure 12.

Figure 10. Triplet emission of the DPApC ligand and its gadolin-

ium complex 50 µs after pulsed excitation. [DPApC^2–] = 3ϫ10^–4,

[Gd³⁺] = 1ϫ10^–4 in a frozen Tris 0.1  aqueous solution (10%

glycerol), pH 7.4, T = 77 K. Emission of [Gd(dipic)₃]^3–scaled to

show location of the triplet emission.

Figure 12. Relative energy diagram of the coumarin (Coum) and

dipicolinate (DPA) moieties with their possible interactions (ISC_xy

:

Intersystem crossing from the xth to the yth excited state,

ET_LϽ–ϾC: Energy transfer between ligand L and chromophore C,

IC_Txy: internal conversion from the xth triplet excited state to the

yth triplet excited state). Wavelet arrows point at nonradiative deac-

tivation, whereas straight arrows point at radiative transitions (ab-

sorptions: solid purple, emissions: solid black). Thick black lines

indicate the lower energy states of the singlet ground (S₀) or nth

excited states (S_n), thick colored lines indicate the lower energy

states of the triplet nth excited states (T_n) and dashed black lines

are higher vibrational modes of the parent lower energy states.

Single pathway T₁^CoumǞEu: Upon 320 nm excitation, the

absorption by the coumarin moiety promotes an electron

from the π HOMO to the π_α* LUMO of the chromophore,

followed by an ISC from this singlet ¹ππ_α* excited state (1st

excited state of the coumarin, S₁^Coum) to its triplet state

³ππ_α* (T₁^Coum). Then, an energy transfer (ET) to the euro-

pium ion can occur, as this triplet state (³ππ_α*) is located

Figure 11. Triplet emission of the polyoxyethylene coumarin chro-

mophore (CpOMe) 50 µs after pulsed excitation. Frozen saturated

solution diluted 10ϫ in Tris (0.1  aqueous solution containing

10% glycerol, pH 7.4) at T = 77 K.

bands at 286.5 nm and at 309.9 nm were predicted and at-

tributed to π–π* transitions (Supporting Information, Fig-

ure S9). The band located at 309.9 nm is more intense than

the one at 286.5 nm. It involves the same π HOMO as the

band at 286.5 nm, but lower π_α* LUMO, relative to the

higher energy π_β* molecular orbital. Finally, a third weak

transition is predicted at 348.6 nm and is attributed to an

5

above the emissive D₀spectroscopic level of the europium

ion. Nevertheless, due to the relative proximity of the ³ππ_α*

excited state of the ligand with the accessible europium

spectroscopic levels, a non-negligible back-transfer from the

3

europium excited level to the ππ_α* state of the ligand also

occurs.

Coum

n–π* transition from an n HOMO to the π_α* LUMO. The Mixed pathway {T₂

p T₁^DPA}ǞEu: When the exci-

n–π* transition is very weak in accord with its forbidden tation wavelength is at higher energy, such as 270 nm, the

character and is barely seen in the experimental UV/Vis ab- absorption on the coumarin promotes an electron to an-

sorption or excitation spectrum (Figure S5) as a slight other molecular orbital, the π_β*, higher in energy than the

1

shoulder at 340 nm. The most intense π–π_α* transition is π_α* LUMO. After ISC from the singlet ππ_β* (2nd excited

experimentally observed as the absorption maximum at state of the coumarin, S₂^Coum) to the triplet ³ππ_β* (T₂^Coum),

320 nm, whereas the more energetic π–π_β* transition is ob- the triplet can undergo an internal conversion to the lower

served in the excitation spectrum as an apparent shoulder at ³ππ_α*, in which case the same problem is encountered as

Eur. J. Inorg. Chem. 2010, 2700–2713

www.eurjic.org

2707

J. Andres, A.-S. Chauvin

FULL PAPER

upon 320 nm excitation. Alternatively, it can directly trans- Table 2. Quantum yields, lifetimes, and sensitization efficiencies of

a 1ϫ10^–4 solution of [Eu(DPApC)₃]^3–in different solvents. Error

fer its energy to the europium ion, with little issues of back-

in τ :Ϯ0.05 ms; 10% relative error in the other values; λ_ex

=

transfer now, or transfer its energy to the DPA moiety

(which has a triplet excited state located in the same region

of the spectrum). The transfer to the DPA ³ππ* state would

320 nm.

H₂O Tris 0.1 ^[a]EtOH^[b]

DMSO^[c]

L+Eu

then offer another pathway to efficiently sensitize the euro- Φ_LL

0.120

0.106

0.014

1.45

4.6

0.32

0.142

0.127

0.015

1.85

4.6

0.40

0.051

0.049

0.002

2.11

4.0

0.53

Φ_LEu

pium ion with very little or no back-transfer to the original

ligand triplet state. Finally, absorption of the 270 nm exci-

tation light source can also directly promote an electron

from a π MO of DPA moiety to a π* (¹ππ* state or S₁^DPA),

Φ_L

Eu

τ_obs/ms

τ_r^Eu/ms

Eu

Φ_Eu

which, after ISC, becomes a ππ* system (T₁^DPA) with en-

3

η_sens

0.044

0.037

0.004

3

ergy similar to the ππ_β* state and offers the same alterna-

[a] n = 1.333. [b] n = 1.361. [c] n = 1.4785.

3

tives to the ππ* relative to the ππ_β*.

Nevertheless, the ³ππ_α* state remains accessible wherever

the sensitization comes from. Thus, back-transfer should be

independent of the sensitization pathway. On the other

hand, what ought not to be the same is the energy transfer

rate from either a ³ππ_β* or a ³ππ_α*. It then becomes obvious

that the higher triplet state (³ππ_β*) has a higher energy

transfer rate to the europium ion than the LUMO so that

the global transfer rate taking into account the unavoidable

Mixtures of two solvents were also investigated. Fig-

ure 13 presents the ratio of coumarin to europium emission,

and Figure S11 (Supporting Information) shows the quan-

tum yields for the europium and coumarin emission. When

increasing the ethanol percentage in a water solution, the

quantum yield is maximal at 20% of ethanol, then falls

down and is again increased in pure ethanol. The coumarin

emission is maximal with 40% of ethanol and then keeps

gently decreasing up to 100% ethanol. When it comes to

the ratio of coumarin to europium emission, we observe a

maximum with 60% ethanol. In a mixture of water/DMSO,

the quantum yield is maximal for 20% of DMSO and de-

creases in pure DMSO, whereas the coumarin emission is

maximal at 40% DMSO. Interestingly, the ratio keeps in-

creasing with increasing percentage of DMSO, so that it

might be used to determine the ratio of DMSO in Tris-

buffered water. This can be of interest for analytical applica-

tions, to determine the amount of DMSO present in water

solutions.

3

back-transfer to the ππ_β* is increased upon 270 nm exci-

tation.

Luminescence of the Europium Complex in Different

Solvents

Emission of coumarins is known to be dependent on the

solvent in which they are dissolved. Yet, it has been shown

that 7-hydroxy-based coumarins are less apt to strong solva-

tochromism than 7-amino-based coumarins. This behavior

is rationalized by taking into account the different deactiva-

tion pathways of the excited state that these two types of

fluorophores experience.^[38]In order to check the influence

of the coumarin in our system, its photophysical properties

were measured in different solvents. EtOH, DMSO, and

Tris-buffered H₂O were tested as pure solvents or as mix-

tures.

Table 2 shows the quantum yields, lifetimes, and sensiti-

zation efficiencies for the emission of [Eu(DPApC)₃]^3–in

Tris-buffered water (pH 7.4), ethanol, or DMSO. Similar

quantum yields for the europium emission upon excitation

at 320 nm were measured in water and ethanol (ca. 1.5%),

while the value decreases drastically in DMSO (ca. 0.2%).

The emission of the coumarin part is more affected by the

L

solvent, with Φ_L≈ 11% in water, ca. 13% in ethanol, and

only ca. 5% in DMSO. However, lifetimes of the Eu^{3+ 5}D₀

excited state follow the opposite trend, ranging from τ =

1.45 ms in water to 2.11 ms in DMSO. Sensitization effi-

ciencies of ca. 4% were calculated in water and ethanol but

only 0.4% in DMSO. This low efficiency explains the lower

overall quantum yield in DMSO despite a higher intrinsic

quantum yield (53%). Furthermore, as the coumarin emis-

sion is also lower in DMSO than in the other two solvents,

Figure 13. Ratio of the coumarin to europium emission as a func-

tion of solvent mixture (1ϫ10^–4 in complex, λ_ex= 320 nm).

Europium Complex with Mixed Ligands

According to the literature, the coordination ability of

it seems that DMSO is a good quencher of the ligand ex- the dipicolinate moiety (DPA) in DPApR ligands is slightly

cited states, but not of the europium.

affected by the nature of the R group.^[18]Hence, we can

2708

www.eurjic.org

Eur. J. Inorg. Chem. 2010, 2700–2713

Europium Complexes of Tris(dipicolinato) Derivatives

obtain mixed Eu/dipic/DPApC complexes by mixing both

To further prove this assumption, lifetime measurements

ligands in different ratios in the range 0.33:x:(3 – x) with x were conducted at room temperature. The higher lifetime

= 0–3. In the following discussion, the [Eu(dipic)_x- was found for the [Eu(dipic)₃]^3–complex [1.55(5) ms], and

(DPApC)_3–x] complexes will be labeled as [x:(3 – x)] for pur- the lifetime slightly decreased in [2:1] [1.49(5) ms], [1:2]

poses of clarity.

[1.44(5) ms], and [Eu(DPApC)₃]^3–[1.42(5) ms]. Even if the

The quantum yield upon excitation at 280 nm is maxi- observed lifetimes are not significantly different, a general

mum with the [Eu(dipic)₃]^3–complex [3:0] and is decreased lowering trend can be reasonably deduced despite the exper-

when a lower ratio of dipic ligand is in solution (Figure 14). imental error, at least between 1.55(5) and 1.42(5) ms. As

The [Eu(DPApC)₃]^3–complex [0:3] has the lowest quantum the observed lifetime is a good indicator of the quenching

yield of this series, whereas the mixed-ligand complexes of the europium excited state, the affirmation that more

have intermediate values. On the other hand, absorption in- back-transfer is introduced as the ratio of DPApC is in-

creases from the [3:0] to [0:3] complex; the DPApC ligand creased relative to the dipic ligand is consistent both with

is a better absorber than the dipic one.

the experimental data and the sensitization hypothesis.

Quantum yields upon excitation at 320 nm were then

measured for the europium and coumarin emissions. As

shown in Figure 15, both europium and coumarin quantum

yields increase when changing the stoichiometry from the

pure [Eu(DPApC)₃]^3–to increasing amounts of dipic ligand.

Nevertheless, absorption at 320 nm is of course reduced

when the complex concentration is kept constant until no

more photons are absorbed at 320 nm by the pure [Eu-

(dipic)₃]^3–complex. The inverse of these quantum yields

were then plotted as a function of DPApC concentration.

The fit showed a good linearity, thus confirming a Stern–

Volmer deactivation kinetics, according to Equations (1) to

(4) in the Experimental Section.

Figure 14. Quantum yields of the pure and mixed-ligand complexes

upon excitation at 280 nm, Tris (0.1  aqueous solution, pH 7.4),

1ϫ10^–4 in complex.

When it comes to the excitation at 320 nm (Figure S12),

the trend is the opposite. The [3:0] complex cannot be ex-

cited at this wavelength in solution. However, as soon as a

DPApC ligand is introduced, excitation can take place and

gives rise to a europium emission. The gain in europium

intensity between [0:3] and the mixed-ligand [1:2] and [2:1]

complex is here less important than the decrease in intensity

with excitation at 280 nm. More than 75% of the emission

intensity of the pure complex (the most emissive of the

Figure 15. Quantum yield for the europium (red) and coumarin

(blue) emission of the mixed-ligand complexes upon excitation at

series) is conserved with only one equivalent of sensitizing 320 nm.

ligand, only 40% remains upon excitation at 280 nm with

one equivalent of the dipic ligand. There are several expla-

Stern–Volmer plots for the emission of europium and

nations to this different loss in intensity. First of all, when coumarin as a function of [DPApC] (Supporting Infor-

the number of dipic ligands is decreased, the probability mation, Figure S13) point to a self-quenching of the cou-

that they absorb the incoming light is reduced. Then, fewer marin excited states by the coumarin moiety itself. From the

photons will be absorbed by the dipic ligand and transfer- linear fits of both emissions, the emission quantum yields

red to the europium center with the high sensitization effi- without quenching can be extracted as the inverse of the y-

ciency of this ligand. Second, and according to the location intercepts. Thus, Φ_em⁰(Eu³⁺) = 4.3%, whereas Φ_em⁰(Coum)

of the triplet state of each kind of ligand, a nonradiative = 17.7%. This means that the emission of the coumarin is

transfer from the excited dipic ligand to an unexcited quenched by 40% by the coumarin moiety itself in the pure

DPApC ligand might occur and hence further decrease the [Eu(DPApC)₃]^3–complex and that nearly 70% of the euro-

amount of net sensitization of the europium. The third pium emission is also quenched by the coumarin, probably

point is that, the europium, once sensitized, can back-trans- by energy back-transfer to its low-lying triplet excited state.

fer its excitation to the DPApC ligand, as observed in the

present work. As soon as a DPApC ligand is coordinated, Table 1 and approximation of the radiative lifetime of the

back-transfer should become more relevant. coumarin emission in the nanosecond range, we found that

By use of the radiative lifetime of europium found in

Eur. J. Inorg. Chem. 2010, 2700–2713

www.eurjic.org

2709

J. Andres, A.-S. Chauvin

FULL PAPER

the quenching rate for the europium emission is near a dif- transfer to the lowest triplet state ought to be independent

fusion-limited quenching rate according to the Stokes–Ein- of the sensitization pathway. This complex has also shown

stein relation. On the other hand, the coumarin emission is potential in analytical applications due to the sensitivity of

quickly quenched, much quicker than the diffusion rate, the coumarin emission to the solvent. Binary mixtures of

since quenching should happen in the picosecond range.

DMSO with Tris-buffered water was found to be particu-

In order to further check this behavior, solutions with larly well suited for this complex, since emissions of the

0.33 and 0.66 equiv. DPApC ligands and hence 2.66 and europium and coumarin are differently affected by the in-

2.33 equiv. dipic ligands, respectively, were prepared. The crease in DMSO content. This gives access to a method of

resulting emission of the coumarin was found to follow the determining the content of DMSO in a water solution. Fi-

Stern–Volmer plot, whereas the emission of the europium nally, complexes with two ligands, DPApC and dipic, were

was too weak to be properly integrated (Figure S13). The prepared at different ratios of ligands. This study revealed

advantage of mixing dipic and DPApC ligands is that the that the quantum yield for either the coumarin emission or

DPApC can be diluted without diluting the Eu³⁺, thus re- the europium emission is self-quenched by the DPApC li-

ally changing the ratio of DPApC/Eu³⁺. Furthermore, since gand upon excitation of the coumarin moiety at 320 nm.

the dipic ligand does not absorb at 320 nm, no photophysi- This technique offers a new tool in the rationalization of the

cal interactions should occur. The constant Eu/Ligand^totra- europium sensitization by considering the self-quenching of

tio being set at 1:3, a constant concentration of the complex the ligand as a possible factor of deactivation.

is kept, so that only the effect of the dilution of the couma-

rin sensitizing group can be studied in this way.

To conclude, the advantage of this class of ligand over

many other ligands is that: first, it allows to finely tune the

sensitization efficiency and extinction coefficient by a care-

ful choice of the coupled chromophore, without signifi-

cantly interfering with the coordinating ability of the dipic-

olinate moiety as long as the chromophore is noncoordina-

ting; second, it offers a huge scope of derivatization with

Conclusions

A photophysical study of para-polyoxyethylene dipico- potential abilities in biological media; and finally, it pro-

linic acid derivatives coupled to methylumbelliferone as li- vides a unique framework for the investigation and better

gand for europium complex is reported. This new class of understanding of the photophysical properties of lumines-

ligands is easily synthesized and its coordination to a euro- cent lanthanide complexes.

pium ion gives the desired 1:3 complex under stoichiometric

conditions, with properties quite similar to those already

obtained with tris(dipicolinato) complex in terms of sta-

bility and coordination environment. The complex can be

excited at 280 nm, corresponding to the DPA backbone of

Experimental Section

the ligand, but, additionally, due to the presence of the cou-

marin chromophore, the excitation range is expanded up to

360 nm in aqueous solutions and up to 400 nm in the solid

state. This opens perspectives in view of exciting the dipicol-

inato complexes at a higher wavelength than in the short-

wave UV. Reaching excitation wavelengths in the near-vis-

ible range can give access to applications in several do-

mains, such as biology, color reproduction, or document

security. The coumarin chromophore can be introduced in

the ligand, but other dipicolinato complexes can also be

doped with a few percent of the DPApC ligand, as stated

above. As a drawback, a significant thermally assisted back-

transfer from the excited europium ion to the lowest triplet

state of the ligand was highlighted. A careful investigation

of the triplet excited sates of the ligand together with the

ones of the uncoupled moieties by time resolved emission

spectrometry showed that the coumarin moiety possesses

two excited states coming from π–π* transitions to two dif-

ferent π* molecular orbitals. A difference in energy transfer

from these triplet states may explain the difference in euro-

pium quantum yields upon excitation at low energy

General Procedures: Solvents were purified by a nonhazardous pro-

cedure by passing them through activated alumina columns (Innov-

ative Technology Inc. system).^[45]Chemicals were ordered from

Fluka and Aldrich and used without further purification. ESI-MS

spectra were obtained with a Finningan SSQ 710C spectrometer by

using 10^–5–10^–4 solutions in acetonitrile/H₂O/acetic acid (50:50:1)

or MeOH, a capillary temperature of 200 °C, and an acceleration

potential equal to 4.5 keV. The instrument was calibrated with

horse myoglobin standard and analyses were conducted in the posi-

tive mode with a 4.6 keV ion spray voltage. Mass spectrometry ex-

periments were performed by Dr Laure Ménin and elemental

analyses by Dr. E. Solari, at the École Polytechnique Fédérale de

Lausanne. ¹H NMR spectroscopy was performed with a Bruker

Avance DRX 400 spectrometer and ¹³C NMR spectroscopy with a

Bruker AV 600 MHz instrument at 25 °C, with deuteriated solvents

as internal standards. Chemical shifts are given in ppm relative to

TMS.

Physicochemical Measurements

Spectrophotometric Measurements

Analytical grade solvents and chemicals were used without further

purification. Aqueous solutions were prepared from doubly dis-

tilled water. Lanthanide solutions were prepared from the corre-

sponding perchlorate salt and titrated by complexometry with a

standardized Na₂H₂EDTA solution in urotropine-buffered medium

Eu

(320 nm, Φ_L= 1.4%) or higher energy (270 nm, Φ_L

=

2.7%). An increase in energy transfer from the higher triplet

state is then needed in order to rationalize the higher emis-

sion quantum yield upon 270 nm excitation, since back- and with xylenol orange as indicator.^[46]

2710

www.eurjic.org

Eur. J. Inorg. Chem. 2010, 2700–2713

Europium Complexes of Tris(dipicolinato) Derivatives

UV/Vis absorption spectra were measured with a Perkin–Elmer

Lambda 900 spectrophotometer by using 1 cm path length quartz

cells. Factor analysis and mathematical treatment of the spectro-

photometric data were performed with the SPECFIT program.^[47]

A_MD,0is the spontaneous emission probability for the purely mag-

netic dipole (MD) transition, whose intensity is practically not in-

fluenced by the chemical environment of the lanthanide ion

(⁵D₀Ǟ⁷F₁in Eu³⁺, A_MD,0= 14.65 s^–1), n is the refractive index of

the solvent, and I_tot/I_MDis the ratio of total emission intensity to

the emission intensity of the MD transition.^[48]

Luminescence Measurements

Luminescence measurements were recorded with a Fluorolog 3–

22 spectrofluorometer from Jobin–Yvon. Emission and excitation

spectra were measured in 1 cm path length quartz cells or quartz

Suprasil^®capillaries at room temperature and in quartz Suprasil^®

capillaries at 77 K with 10% glycerol added to the solutions. Room

temperature measurements in quartz Suprasil^®capillaries were per-

formed in an integration sphere.^[41]All emission spectra were

measured in photon counts and corrected for the instrumental

function. Luminescence evolution of the europium complex was

performed by variation of the pH in a fashion similar to that in

the titration of the ligands but on a 1:3 europium/ligand solution.

Water molecules in the first coordination sphere were estimated by

the Supkowski and Horrocks formulas [Equations (8) and (9)].

q = A[τ_obs^–1(H₂O) – τ_obs^–1(D₂O) – k_XH

]

(8)

(9)

k_XH= α + βn_OH+ γn_NH+ δn_O=CNH

Here τ_obs(H₂O) is the lifetime in water, τ_obs(D₂O) is the lifetime in

deuteriated water, and n_XHis the number of XH oscillators in the

first coordination sphere.^[39]

Synthesis of the Ligand: The synthesis of Br(CH₂CH₂O)₃OMe (2)

has already been reported in the literature, starting from 2-[2-(2-

methoxyethoxy)ethoxy]ethanol and PBr₃, with yields of 18–

25%,^[49]whereas diethyl ester 4 was prepared from chelidamic

acid.^[18]

Quantum yields were determined in the integration sphere by meas-

uring the ratio of the emitted corrected intensity to the absorbed

corrected intensity.^[41]Empty capillaries were used as blank. The

obtained values were checked by comparison to the tris(dipicolin-

ato)europium standard, which exhibits 24% quantum yield in Tris-

buffered aqueous solutions at pH 7.4.^[11]In solution, all concentra-

tions were set at 1ϫ10^–4, including the concentration of the ref-

erence tris(dipicolinato) complex. For solid-state samples, the com-

plex or the uncoupled chromophore (CpOMe) was used as a pow-

der.

7-(Methylpolyoxyethylenyl)methylumbelliferone (CpOMe): To

a

solution of 7-hydroxy-4-methylcoumarin (1) (4.2 mmol, 0.74 g) in

DMF (15 mL) was added K₂CO₃(91.8 mmol, 12.7 g). The solution

was stirred for 5 min and turned yellow. Compound 2 (4.2 mmol,

0.95 g) was then added dropwise, and the solution was stirred for

4 h at room temperature. Afterwards, the solvent was evaporated

under reduced pressure, and the crude product was redissolved in

CH₂Cl₂(150 mL). The solution was washed three times with a satu-

rated NH₄Cl solution (3ϫ80 mL), dried with Na₂SO₄, filtered, and

concentrated under reduced pressure to yield CpOMe (0.9 g, 66%).

¹H NMR (CDCl₃, 400 MHz): δ = 7.48 (d, J = 8.8 Hz, 1 H, H_ar),

6.88 (dd, J₁= 8.8, J₂= 2.5 Hz, 1 H, H_ar), 6.82 (d, J = 2.5 Hz, 1

H, H_ar), 6.13 (d, J = 1.1 Hz, 1 H, H_ar), 4.18 (m, 2 H, -CH₂-), 3.89

(m, 2 H, -CH₂-), 3.74 (m, 2 H, -CH₂-), 3.67 (m, 4 H, -CH₂-), 3.55

(m, 2 H, -CH₂-), 3.37 (s, 3 H, -CH₃), 2.39 (d, J = 1.1 Hz, 3 H,

-CH₃) ppm.

Stern–Volmer plots were fitted by using Equations (1), (2), (3), and

0

(4), where Φ_emis the emission quantum yield, Φ_emis the emission

quantum yield without quenching, k_qis the quenching rate, τ_obs⁰is

the observed lifetime without quenching, [Q] is the concentration

0

of quencher Q, k_ris the radiative emission rate (k_r= 1/τ_r), and k_obs

is the sum of all deactivation rates for the excited state of the emis-

0

sive molecule (k_obs= 1/τ_obs⁰).

0

Φ_em

= 1 + k_qτ_obs⁰[Q]

(1)

(2)

(3)

(4)

Φ_em

7-(Hydroxypolyoxyethylenyl)methylumbelliferone

(3):

CpOMe

0

1

τ_obs

(2.34 mmol, 0.75 g) was dissolved in CH₃CN (15 mL) under nitro-

gen. The solution was heated up to 60 °C and TMSI (3.51 mmol,

0.5 mL) was then added dropwise. The reaction mixture was stirred

for 80 min at 60 °C before the solvent was evaporated under re-

duced pressure. CH₂Cl₂(50 mL) was added to the remaining me-

dium, and the solution was washed with an aqueous solution of

Na₂S₂O₃(0.09 , 50 mL). The brown aqueous solution was then

extracted three times with CH₂Cl₂(3ϫ50 mL), and the collected

organic solutions were dried with Na₂SO₄, filtered, and concen-

trated under reduced pressure to yield 3 (0.60 g, 80%). ¹H NMR

(CDCl₃, 400 MHz): δ = 7.49 (d, J = 8.8 Hz, 1 H, H_ar), 6.89 (m, 1

H, H_ar), 6.84 (dd, J₁= 2.4, J₂= 1.2 Hz, 1 H, H_ar), 6.13 (d, J =

1.0 Hz, 1 H, H_ar), 4.19 (m, 2 H, -CH₂-), 3.90 (m, 2 H, -CH₂-), 3.73

(m, 6 H, -CH₂-), 3.62 (m, 2 H, -CH₂-), 2.39 (d, J = 1.0 Hz, 3 H,

-CH₃) ppm.

=

+ k_q

₀[Q]

0

Φ_emΦ_em

Φ_em

0

k_r

τ_obs

0

Φ_em

=

0

k_obs

τ_r

1

=

+ k_qτ_r[Q]

0

Φ_emΦ_em

Radiative lifetimes, τ_obs, intrinsic quantum yields, Φ_Ln^Ln, and sensi-

tization efficiency, η_sens, were calculated according to Equations (5),

(6), and (7).

1

= k_r= A_MD,0n³(I_tot/I_MD

)

(5)

τ_r

Diethyl 4-[(4-Methylcoumarin-7-yl)oxypolyoxyethylenyl]dipicolinate

(5): Compounds 3 (1.86 mmol, 0.60 g), 4 (0.93 mmol, 0.22 g), and

PPh₃(1.86 mmol, 0.49 g) were added into THF (15 mL). After

dropwise addition of diisopropyl azodicarboxylate (DIAD) (1.86,

0.37 mL) to the solution, it was heated under reflux for three hours

(controlled by TLC, CH₂Cl₂/MeOH, 97:3). The solvent was then

evaporated under reduced pressure, and the remaining mixture was

dissolved in CH₂Cl₂(100 mL). The solution was washed three

times with water (3ϫ100 mL), dried with Na₂SO₄, filtered, and

τ_obs

Ln

Φ_Ln

=

(6)

(7)

τ_r

Ln

Φ_L= η_sensΦ_Ln

In these equations, k_ris the radiative deexcitation rate, τ_obsis the

observed lifetime upon radiative and nonradiative deactivation,

Eur. J. Inorg. Chem. 2010, 2700–2713

www.eurjic.org

2711

J. Andres, A.-S. Chauvin

FULL PAPER

concentrated under reduced pressure to yield 1.73 g of crude prod-

uct, which was dissolved in a minimum amount of AcOEt and kept

in a freezer overnight to yield a precipitate, which was filtered off.

The precipitation was repeated until no solid formed. The clear

filtrate solution was concentrated under reduced pressure to yield

pure 5 mixed with triphenylphosphane oxide (ca. 15%). The prod-

uct was used as such, the triphenylphosphane oxide being elimin-

ated in the next step. Alternatively, 5 can be purified by chromatog-

raphy (SiO₂, CH₂Cl₂/MeOH 100 to, 95:5), but then the yield is low

(20%), as a result of partial hydrolysis of the ester moieties, the

carboxylic acid product being retained on the column. ¹H NMR

(CDCl₃, 400 MHz): δ = 7.79 (s, 2 H, H_ar), 7.42 (m, 1 H, H_ar), 6.86

(m, 1 H, H_ar), 6.82 (m, 1 H, H_ar), 6.14 (m, 1 H, H_ar), 4.96 (m, 2

H, -CH₂-), 4.45 (m, 2 H, -CH₂), 4.30 (m, 2 H, -CH₂-), 4.18 (m, 4

H, -CH₂-), 3.89 (m, 4 H, -CH₂-), 3.75 (s, 2 H, -CH₃), 2.38 (s, 3 H,

-CH₃), 1.44 (t, J = 7.1 Hz, 6 H, -CH₃) ppm.

Acknowledgments

We thank Jean-Claude G. Bünzli for hosting us in his laboratory,

Frédéric Gumy and Svetlana Eliseeva for high-resolution excitation

measurements, and Roger D. Hersch for providing financial sup-

port. This research is supported through grants from the Swiss

National Science foundation (FNS 126757).

[1] I. Grenthe, J. Am. Chem. Soc. 1961, 83, 360–364.

[2] I. Grenthe, I. Tobiasson, Acta Chem. Scand. 1963, 17, 2101–

2101.

[3] A. Mondry, P. Starynowicz, J. Alloys Compd. 1995, 225, 367–

371.

[4] P. A. Brayshaw, J.-C. G. Bünzli, P. Froidevaux, J. M. Har-

rowfield, Y. Kim, A. N. Sobolev, Inorg. Chem. 1995, 34, 2068–

2076.

[5] P. A. Brayshaw, J. M. Harrowfield, A. N. Sobolev, Acta Crys-

tallogr., Sect. C: Cryst. Struct. Commun. 1995, 51, 1799–1802.

[6] J. M. Harrowfield, Y. Kim, B. W. Skelton, A. H. White, Aust.

J. Chem. 1995, 48, 807–823.

4-[(4-Methylcoumarin-7-yl)oxypolyoxyethylenyl]dipicolinic

Acid

(H₂DPApC): To a solution of crude product 5 (0.81 mmol, 0.43 g)

in EtOH (20 mL) was added dropwise an aqueous solution of

NaOH (0.5 , 1.77 mmol, 3.54 mL), and the solution was stirred

at room temperature for 20 min. The solvent was then evaporated

under reduced pressure, and the aqueous solution was washed three

times with CH₂Cl₂(3ϫ50 mL) to remove triphenylphosphane ox-

ide. It was acidified with HCl (1 ) down to pH 1.5. The acidic

solution was then extracted three times with CH₂Cl₂(3ϫ150 mL),

the organic phase was dried with Na₂SO₄, filtered, and concen-

trated under reduced pressure to yield H₂DPApC (315 mg, 85%

[7] H. Donato, R. B. Martin, J. Am. Chem. Soc. 1972, 94, 4129–

4131.

[8] C. N. Reilley, B. W. Good, J. F. Desreux, Anal. Chem. 1975, 47,

2110–2116.

[9] C. Platas, C. Piguet, N. André, J.-C. G. Bünzli, J. Chem. Soc.,

Dalton Trans. 2001, 3084–3091.

[10] N. Ouali, B. Bocquet, S. Rigault, P.-Y. Morgantini, J. Weber,

C. Piguet, Inorg. Chem. 2002, 41, 1436–1445.

[11] A.-S. Chauvin, F. Gumy, D. Imbert, J.-C. G. Bünzli, Spectrosc.

Lett. 2004, 37, 517; erratum: A.-S. Chauvin, F. Gumy, D. Im-

bert, J.-C. G. Bünzli, Spectrosc. Lett. 2006, 40, 193.

[12] F. G. Prendergast, J. Lu, P. J. Callahan, J. Biol. Chem. 1983,

258, 4075–4078.

[13] A. Aebischer, F. Gumy, J.-C. G. Bünzli, Phys. Chem. Chem.

Phys. 2009, 11, 1346–1353.

[14] J. B. Lamture, Z. H. Zhou, A. S. Kumar, T. G. Wensel, Inorg.

Chem. 1995, 34, 864–869.

1

over two steps) as a white powder. H NMR (D₂O, 400 MHz): δ =

7.43 (s, 2 H, H_ar), 7.36 (m, 1 H, H_ar), 6.78 (m, 1 H, H_ar), 6.70 (m,

1 H, H_ar), 6.02 (m, 1 H, H_ar), 4.27 (d, J = 0.8 Hz, 2 H, -CH₂-),

4.16 (m, 2 H, -CH₂-), 3.97 (m, 2 H, -CH₂-), 3.93 (m, 2 H, -CH₂-),

3.84 (m, 4 H, -CH₂-), 2.20 (s, 3 H, CH₃) ppm. C₂₃H₂₃NO₁₀·1.6H₂O

(502.26): calcd. C 55.00, H 5.26, N 2.79; found C 55.07, H 5.23, N

3.04.

Tris(4-[(4-methylcoumarin-7-yl)oxypolyoxyethylenyl]dipicolinato)-

europium {Cs₃[Eu(DPApC)₃]^3–}: Solid H₂DPApC (0.53 g, 1.12

mmol) was mixed with solid Eu₂O₃(0.17, 0.48 mmol) and stirred

in refluxed H₂O (5 mL) for 30 min. Cs₂CO₃(0.53 g, 1.62 mmol)

was then added to the hot solution. When necessary, additional

cesium carbonate was added to obtain a clear solution. The solu-

tion was then filtered and kept at room temperature overnight. No

crystallization was observed. The solvent was evaporated under re-

duced pressure until the formation of an oily aggregate. This aggre-

gate then formed a vitrified solid (small beads) when left to dry in

air. Further drying was needed under reduced pressure to yield

[15] X.-H. Yin, M. Y. Tan, Zhongguo Xitu Xuebao (J. Chin. Rare

Earth Soc.) 2001, 19, 555–560.

[16] X.-H. Yin, M. Y. Yang, H. H. Shi, G. Yang, X. B. Chen, L.-Q.

Gu, Zhongguo Xitu Xuebao (J. Chin. Rare Earth Soc.) 2000,

18, 211–214.

[17] M. W. George, C. A. Golden, M. C. Grossel, R. J. Curry, Inorg.

Chem. 2006, 45, 1739–1744.

[18] A.-L. Gassner, C. Duhot, J.-C. G. Bünzli, A.-S. Chauvin, Inorg.

Chem. 2008, 47, 7802–7812.

[19] M. Latva, H. Takalo, V. M. Mukkala, C. Matachescu, J.-C.

Rodriguez-Ubis, J. Kankare, J. Lumin. 1997, 75, 149–169.

[20] K. Binnemans, Chem. Rev. 2009, 109, 4283–4374.

[21] N. M. Shavaleev, S. V. Eliseeva, R. Scopelliti, J.-C. G. Bünzli,

Chem. Eur. J. 2009, 15, 10790–10802.

[22] W. Yang, J. Z. Gao, M. Chen, J. W. Kang, Spectrosc. Lett.

1997, 30, 367–378.

Cs₃[Eu(DPApC)₃]

as

a

yellow

to

orange

solid.

Cs₃C₆₉H₆₃EuN₃O₃₀·4H₂O (2037.01): calcd. C 40.69, H 3.51, N

2.06; found C 40.70, H 3.24, N 2.15.

Supporting Information (see footnote on the first page of this arti-

cle): ESI-TOF-MS analysis of the complex, UV/Vis absorption

upon Eu³⁺addition, europium emission compared to similar com-

plexes, CpOMe absorption and excitation spectra, peak fitting of

the coumarin extinction coefficient as a function of wavenumber,

ratio of coumarin and europium emission in solution and in the

solid state, excitation spectra for the emission of europium and cou-

marin in the solid state and ratio of coumarin to europium emis-

sion relative to excitation at 320 nm, calculated molecular orbitals

of CpOMe and DPApOMe, quantum yields for the europium and

coumarin emission in different solvent mixtures, corrected exci-

tation spectra of the pure and mixed-ligand complexes, Stern–

Volmer plot for the emission of europium and coumarin with in-

creasing ratio of DPApC ligand.

[23] C. Reinhard, H. U. Güdel, Inorg. Chem. 2002, 41, 1048–1055.

[24] D. C. Lai, B. Dunn, J. I. Zink, Inorg. Chem. 1996, 35, 2152–

2154.

[25] P. C. R. Soares-Santos, H. I. S. Nogueira, V. Felix, M. G. B.

Drew, R. A. S. Ferreira, L. D. Carlos, T. Trindade, Chem. Ma-

ter. 2003, 15, 100–108.

[26] P. M. Pellegrino, N. F. Fell, D. L. Rosen, J. B. Gillespie, Anal.

Chem. 1998, 70, 1755–1760.

[27] M. L. Cable, J. P. Kirby, K. Sorasaenee, H. B. Gray, A. Ponce,

J. Am. Chem. Soc. 2007, 129, 1474–1475.

[28] J. P. Kirby, K. J. Venkateswaran, A. Ponce, 2007, USA.

[29] T. D. Barela, A. D. Sherry, Anal. Biochem. 1976, 71, 351–357.

[30] N. Tancrez, C. Feuvrie, I. Ledoux, J. Zyss, L. Toupet, H.

Le Bozec, O. Maury, J. Am. Chem. Soc. 2005, 127, 13474–

13475.

2712

www.eurjic.org

Eur. J. Inorg. Chem. 2010, 2700–2713

Europium Complexes of Tris(dipicolinato) Derivatives

[31] H. Hou, Y. Wei, Y. Song, Y. Fan, Y. Zhu, Inorg. Chem. 2004,

43, 1323–1327.

[32] A. DЈAléo, G. Pompidor, B. Elena, J. Vicat, P. L. Baldeck, L.

Toupet, R. Kahn, C. Andraud, O. Maury, ChemPhysChem

2007, 8, 2125–2132.

[33] A. Picot, F. Malvolti, B. LeGuennic, P. L. Baldeck, J. A. G.

Williams, C. Andraud, O. Maury, Inorg. Chem. 2007, 46, 2659–

2665.

[40]

[41]

[42]

T. Moriya, Bull. Chem. Soc. Jpn. 1983, 56, 6–14.

F. Gumy, PCT/IB2007/054187, 2009.

S. Tobita, M. Arakawa, I. Tanaka, J. Phys. Chem. 1984, 88,

2697–2702.

S. Tobita, M. Arakawa, I. Tanaka, J. Phys. Chem. 1985, 89,

5649–5654.

Scigress Explorer Ultra 7.7.0.47, Fujitsu Limited, 2007.

[43]

[44]

[34] A. Picot, A. DЈAleo, P. L. Baldeck, A. Grichine, A. Duperray,

C. Andraud, O. Maury, J. Am. Chem. Soc. 2008, 130, 1532–

1533.

[35] A.-S. Chauvin, S. Comby, B. Song, C. D. B. Vandevyver, J.-

C. G. Bünzli, Chem. Eur. J. 2008, 14, 1726–1739.

[36] P. Hammond, A. Fletcher, R. Henry, R. Atkins, Applied Phys-

ics A: Mater. Science Process. 1975, 8, 311–314.

[37] D. P. Specht, P. A. Martic, S. Farid, Tetrahedron 1982, 38,

1203–1211.

[45] A. B. Pangborn, M. A. Giardello, R. H. Grubbs, R. K. Rosen,

F. J. Timmers, Organometallics 1996, 15, 1518–1520.

[46] G. Schwarzenbach, Complexometric Titrations, Chapman &

Hall, London, 1957.

[47] H. Gampp, M. Maeder, C. J. Meyer, A. D. Zuberbühler, Tal-

anta 1985, 23, 1133–1139.

[48] M. H. V. Werts, R. T. F. Jukes, J. W. Verhoeven, Phys. Chem.

Chem. Phys. 2002, 4, 1542–1548.

[49] A. Cappelli, S. Galeazzi, M. Anzini, S. Vomero, PCT/EP2007/

059698, 2008.

[38] B. D. Wagner, Molecules 2009, 14, 210–237.

[39] R. M. Supkowski, W. d. Horrocks Jr., Inorg. Chim. Acta 2002,

340, 44–48.

Received: February 3, 2010

Published Online: ᭿

Eur. J. Inorg. Chem. 2010, 2700–2713

www.eurjic.org

2713

Article Doi

Europium complexes of tris(dipicolinato) derivatives coupled to methylumbelliferone: A double sensitization

DOI: 10.1002/ejic.201000126

Source and publish data:

Authors:

Article abstract of DOI:10.1002/ejic.201000126

Full text of DOI:10.1002/ejic.201000126

Products guided by the article

R&D Labs maybe for 1243174-02-8

Relevant to this article

Hot Product