Lanthanide(III) Complexes of Cyclen Triacetates and Triamides Bearing Tertiary Amide-Linked Antennae

Abstract

The coordination compounds of the trivalent lanthanide ions (Ln(III)) have unique photophysical properties. Ln(III) excitation is usually performed through a light-harvesting antenna. To enable Ln(III)-based emitters to reach their full potential, an understanding of how complex structure affects sensitization and quenching processes is necessary. Here, the role of the linker between the antenna and the metal binding fragment was studied. Four macrocyclic ligands carrying coumarin 2 or 4-methoxymethylcarbostyril sensitizing antennae linked to an octadentate macrocyclic ligand binding site were synthesized. Complexation with Ln(III) (Ln = La, Sm, Eu, Gd, Tb, Yb and Lu) yielded species with overall −1, 0, or +2 and +3-charge. Paramagnetic ¹H NMR spectroscopy indicated subtle differences between the coumarin- and carbostyril-carrying Eu(III) and Yb(III) complexes. Cyclic voltammetry showed that the effect of the linker on the Eu(III)/Eu(II) apparent reduction potential was dependent on the electronic properties of the N-substituent. The Eu(III), Tb(III) and Sm(III) complexes were all luminescent. Coumarin-sensitized complexes were poorly emissive; photoinduced electron transfer was not a major quenching pathway in these species. These results show that seemingly similar emitters can undergo very different photophysical processes, and highlight the crucial role the linker can play.

1. Introduction

The luminescence of the coordination compounds of lanthanide ions (Ln(III)) has found application in a variety of fields, such as in biological sensing and imaging [1], in fluorescent lamps and lasers [2], and in anti-counterfeiting [3,4,5]. The direct excitation of Ln(III) is inefficient due to the low absorption coefficients of the Laporte-forbidden 4f-4f transitions [6]. A common method for overcoming the challenge of low Ln(III) absorptions is to excite the metal ion through a light-harvesting chromophore, a so-called antenna [6,7]. The antenna can transfer the excitation energy to the Ln(III), thus combining the attractive Ln(III) luminescence properties (sharp, spiked emission peaks, long excited-state lifetimes) with the strong absorption of common organic chromophores. The presence of the antenna offers additional benefits. It can carry reactive groups for attachment to biomolecules [8], or labels (e.g., F-19) [9] for creating multimodal imaging agents. The antenna can also provide a way to render the Ln(III) complex analyte-responsive [10,11,12,13,14].

Given the importance of the energy transfer for Ln(III) sensitization, it is not surprising to see the amount of research dedicated to the optimization of the antenna photophysical properties. Helpful guidelines are available for the matching of Ln(III) with the antenna, which discuss the antenna single-triplet gap, and the energy gap between the antenna triplet and the Ln(III) receiving levels [15]. Both the antenna and the lanthanide excited states are susceptible to quenching, which needs to be minimized if highly emissive complexes are to be obtained. Ln(III) have 8–9 coordination sites, most of which need to be occupied by a multidentate ligand for the complex to be kinetically and thermodynamically stable [16]. Furthermore, X-H (X = O, N, C) oscillators quench Ln(III) excited states [17,18], and should thus be eliminated if possible. The complexes shown in Figure 1 have only one inner-sphere water molecule.

Recently, we reported that Ln(III) complexes sensitized by carbostyril antennae that are attached to the metal binding site via tertiary amide linkers are substantially more luminescent than the analogous secondary amide-linked ones (Figure 1a) [19]. While the amide linker certainly influences the antenna excited states, the effect is rather subtle. Furthermore, an X-ray crystal structure of a precursor of the tertiary amide-linked complexes indicated that the orientation of the antenna in the two types of emitters may be different, which could have a substantial effect on the energy transfer efficiency.

Possible structural effects that explain the differences observed between the photophysical properties of secondary and tertiary amide-linked complexes are investigated in this work. We studied the effect of charge on the amide linker substituent, and whether the effect is carried over to complexes carrying a coumarin antenna or a ligand binding site wherein the three carboxylate donors were replaced by three amides (Figure 1b).

2. Results

2.1. Synthesis

The syntheses of LnL1a^Car and LnL1d^Cou have been reported previously, except their La, Sm, Yb and Lu analogues [19,20,21]. Triamides LnL2a^Car, LnL2c^Car and LnL2d^Cou were synthesized following analogous procedures, these are summarized in Scheme 1. Detailed synthetic protocols and complete characterization of all new compounds are given in the Materials and methods. Syntheses started from known monoalkylated cyclens 1 [22] or 6 [20]. Trialkylation with bromoacetamide was carried out in the presence of diisopropyl ethylamine (DIPEA) or K₂CO₃ to yield L2c^Car or L2d^Cou, respectively. Heating the latter at 55 °C in an aqueous-ethanolic mixture with a slight excess of LnCl₃ afforded LnL2d^Cou (Ln = La, Sm, Eu, Gd, Tb, Yb and Lu) in high yield. Stirring L2c^Car in a 1:1 mixture of CH₂Cl₂ and trifluoroacetic acid (TFA) overnight cleaved the tert-butyl ester in the tertiary amide side chain to give L2a^Car, which was then used to form the complexes LnL2a^Car analogously to LnL2d^Cou. The reaction mixtures of LnL2a^Car had to be purified via semipreparative HPLC to afford analytically pure samples, as the complexes partially decomposed during synthesis. We attempted to synthesize the tert-butyl-protected LnL2c^Car the same way as the other LnL2 triamide species. The ^tBu ester in the Ln(III) complexes was labile, and hydrolyzed quantitatively to LnL2a^Car. This was confirmed by LC-MS analysis of their reaction mixtures stirred overnight which showed the presence of both LnL2a^Car and LnL2c^Car species in a 1:2 ratio (Figures S1–S3). The ¹H spectra of LaL2c^Car and LuL2c^Car confirmed the presence of both hydrolyzed and ^tBu ester-protected species in 1:0.45 and 1:0.25 ratios, respectively (Figures S4 and S5).

2.2. Paramagnetic ¹H NMR Spectroscopy

Cyclen-based DO3A-type Ln complexes interconvert between several structures in solution [23]. This dynamic process was studied by paramagnetic ¹H NMR spectroscopy of the Eu(III) and Yb(III) chelates (Figure 2 and Figure 3). The spectra of the diamagnetic La(III) and Lu(III) complexes were similar to those of the ligands (Figures S6–S13). Eu(III) and Yb(III) paramagnetic centers caused broadening and relocation of the signals due to the exchange between square antiprism (SAP) and twisted SAP (TSAP) isomers (Figures S14–S25). The resonances belonging to SAP and TSAP isomers can be distinguished from their different signals of axial cyclen ring protons [24]. Well-defined signals could be obtained upon cooling down the solutions of the Eu(III) complexes [25]. Tricarboxylate complexes EuL1a^Car and EuL1d^Cou were present as mixtures of TSAP (12.6–12.8 ppm and 11.9–13.5 ppm) and SAP (32.2–36.3 ppm and 32.5–36.3 ppm) conformers in D₂O at 10 °C (Figure 2). The ratios of the signals attributed to the SAP and TSAP species were 1:0.22 and 1:0.34 for EuL1a^Car and EuL1d^Cou, respectively (Figures S14 and S18). The spectra of triamide EuL2a^Car, EuL2c^Car and EuL2d^Cou complexes in CD₃OD at 0 °C contained only signals from the SAP isomer (30.6–33.1, 29.9–32.8 and 29.9–33.2 ppm, respectively), as did all the Yb(III) complexes at r.t. (YbL1: 112.5–133.5 ppm (D₂O), YbL2: 95.0–113.3 ppm (D₂O and CD₃OD), Figure 3).

The most downfield signals are assigned to the pseudo-axial cyclen ring protons in the SAP conformers [26,27]. In the unsymmetrical complexes discussed here, up to four signals are expected. This is the case in the carbostyril complexes, where the ¹H NMR spectra of both the Eu and Yb species have four distinct peaks in the 32–36/30–33 ppm and 112–133/95–113 ppm (tricarboxylate/triamide) regions, respectively. However, the spectra of the coumarin-appended complexes present more than four well-resolved signals in most cases, indicating the presence of more than one complex with SAP conformation (Figure 2 and Figure 3, bottom). A possible reason for this observation may be restricted rotation of the antenna due to steric clash between the 6-Me and amide C=O groups.

2.3. Cyclic Voltammetry

One of the processes that can quench the excited antenna is photoinduced electron transfer (PeT) to the Ln(III). PeT is often thermodynamically feasible for Eu and Yb, but not for Gd and Tb, as the latter two have very negative reduction potentials (E(Gd^III/Gd^II) = −3.9 V and E(Tb^III/Tb^II) = −3.7 V vs NHE, respectively [28]). The driving force for PeT is larger for Ln(III) with more positive reduction potentials [29]. There is ample evidence that the reduction potential of the Ln(III) depends on the ligand [30,31,32,33,34,35]. The replacement of neutral amide donors with −1 charged carboxylates in DO3A-type ligands has been found to shift the Eu(III) reduction potential to more negative values by ~95 mV/donor [36]. We expected that more subtle tuning may be achieved by the manipulation of the antenna linker. This hypothesis was investigated by cyclic voltammetry, the results are summarized in

Table 1. Cyclic voltammetry of Eu(III) complexes (positively charged complexes are with Cl-counterions) ^[a].

Compound	E1/2[b]	Epa[b]	Epc[b]	ΔE [b]
EuCl3	−384	−337	−431	94
EuL1aCar	−948 (–839 [c])	−766	−1131	365
EuL2aCar	−612 (–554 [d])	−437	−787	350
EuL2cCar	−564	−437	−691	254
EuL1dCou	−854	−744	−965	221
EuL2dCou	−565	−512	−617	105

^[a]E_1/2 is a half-wave potential, E_pa (E_pc) is anodic (cathodic) peak potential, ΔE is peak separation. ^[b] Values are in mV vs. NHE. Measured in H₂O (LiCl 0.1 M, pH ~6.5) with a sample concentration of 1 mM at a glassy C electrode using a SCE as reference electrode and a Pt wire counter electrode with a scan rate of 100 mV/s. ^[c] E_1/2 of the corresponding secondary amide complex EuL1b^Car, from [36]. ^[d] E_1/2 of the corresponding secondary amide complex EuL2b^Car, from [36].

.

Similarly to what was observed previously [36], the replacement of tricarboxylate by triamide ligand in coumarin complexes increases the Eu(III)/Eu(II) reduction potential from −854 mV vs NHE in EuL1d^Cou to −565 mV vs NHE in EuL2d^Cou, which corresponds to ~96 mV/donor (Table 1).

The introduction of a negatively-charged amide linker in EuL1a^Car and EuL2a^Car leads to a 109 mV and 58 mV decrease in the apparent Eu(III)/Eu(II) reduction potentials, respectively, compared to EuL1b^Car and EuL2b^Car with secondary amide-linked antennae (Figure 4, Table S3). The shift is most likely due to the additional negative charge on the linker, as the reduction potentials of EuL2b^Car and the ^tBu-protected ester-carrying EuL2c^Car are almost identical (−554 mV vs −564 mV vs NHE, respectively). The apparent reduction potential of EuL1d^Cou is only 15 mV lower than the secondary amide-linked EuL1b^Car. Hence, the influence of the neutral amide linker (^tBu-protected methylcarboxylate or Et) on the reduction potential is small (Figure S26), whereas the negatively charged arm leads to a better stabilization of Eu(III). It is possible that the replacement of the carbostyril heterocycle with a coumarin 2 also has a slight influence. Either way, the effects are small (Figure S27) and are in the measurement error range (~10 mV at a scan rate 100 mV/s [36]).

From the above data we can conclude that the electronic impact of the amide linker in EuL1a^Car can be ascribed to its negative charge, which contributes to the stabilization of the +3 oxidation state.

2.4. Photophysical Characterization

The UV-Vis absorption and emission properties of the complexes were studied in 10 mM PIPES buffer in H₂O or D₂O (pH or pD 6.5) at r.t. at nominally 10 µM concentrations. The new carbostyril-carrying complexes had UV-Vis absorption spectra that resembled those of the previously reported LnL1a^Car. Only small variations were seen between complexes of the same ligand with different Ln(III) ions (Figure 5, left), and the complexes had local absorption maxima at λ_abs = 328 nm. Replacing the tricarboxylate metal-binding framework with a triamide one essentially did not impact the complex absorptions (Figure S42). Coumarin-appended LnL1d^Cou and LnL2d^Cou similarly had superimposable absorption spectra (Figure 5, right), with λ_abs = 319 nm (Figure S43).

The triplet levels of the carbostyril and coumarin antennae were determined from the steady-state emission spectra of the Gd(III) complexes at 77 K upon excitation at λ_ex = 327 and 315 nm, respectively (Figure S44). The triplet excited state (T₁) of carbostyril was located at 23,000 cm⁻¹, while that of the coumarin was found at slightly lower energies, and 22,200 cm⁻¹. These values were determined from the 0-0 phonon transitions. The energies of both types of antennae were expected to be suitable for sensitization of Tb(III), Eu(III) and Sm(III), with excited states at 20,600 (⁵D₄), 17,300 (⁵D₀) and 18,000 (⁴G_5/2) cm⁻¹ [37], respectively. As the coumarin T₁ is within ~2000 cm⁻¹ of the Tb(III) excited state, back energy transfer from Tb(III) to T₁ was considered a possibility.

Antenna excitation at λ_ex = 327 or 315 nm yielded Ln(III)-centered emission (Ln = Eu, Tb and Sm) (Figure 6 and Figure 7,

Table 2. Antenna- and Ln(III)-based luminescence quantum yields, Ln(III) lifetimes and hydration states of LnL.

Complex	ΦL [%] [a]	ΦLn [%] [a]	τH2O [ms]	τD2O [ms]	q[b]
LaL1aCar	4.9	-	-	-	-
SmL1aCar	3.6 [c]	0.19 [c]	0.009	0.032	-
EuL1aCar[d]	2.5	8.9	0.66	2.17	1.0
GdL1aCar[d]	5.1	-	-	-	-
TbL1aCar[d]	4.5	45.1	1.81	2.92	0.8
YbL1aCar	4.5	-	-	-	-
LuL1aCar	4.7	-	-	-	-
LaL2aCar	4.1	-	-	-	-
SmL2aCar	3.4 [c]	0.21 [c]	0.010	0.032	-
EuL2aCar	0.77 [e]	2.7 [e]	0.51	2.05	0.9
GdL2aCar	4.3 [c]	-	-	-	-
TbL2aCar	4.1	36.6	1.50	2.70	1.2
YbL2aCar	4.3	-	-	-	-
LuL2aCar	4.5	-	-	-	-
LaL1dCou	0.66	-	-	-	-
SmL1dCou	0.64 [c]	0.059 [c]	0.010	0.031	-
EuL1dCou	0.62 [e]	2.2 [e]	0.62	1.96	1.0
GdL1dCou	0.66 [c]	-	-	-	-
TbL1dCou	0.62	1.7	0.47	0.61	-
YbL1dCou	0.66	-	-	-	-
LuL1dCou	0.64	-	-	-	-
LaL2dCou	0.66	-	-	-	-
SmL2dCou	0.64 [c]	0.067 [c]	0.012	0.032	-
EuL2dCou	0.58 [e]	1.5 [e]	0.54	2.01	0.8
GdL2dCou	0.63 [c]	-	-	-	-
TbL2dCou	0.67	1.0	0.29	0.45	-
YbL2dCou	0.57	-	-	-	-
LuL2dCou	0.61	-	-	-	-

Measurements were performed with [LnL] = 10 µM in 10 mM aqueous PIPES buffer solutions at pH (pD) 6.5. ^[a] Relative to quinine sulfate (Φ = 0.59) in H₂SO₄ (0.05 M) [39]. ^[b] Calculated using the Equation q = 5(1/τ_H2O – 1/τ_D2O – 0.06) for Tb, and q = 1.2(1/τ_H2O – 1/τ_D2O – 0.25 – n × 0.075), where n is the number of nearby N-H oscillators, for Eu [18]. ^[c] Mean for two independent measurements. ^[d] Data from [19]. ^[e] Mean for three independent measurements.

) with all four ligands. The emission spectra along with the assignment of the peaks are shown in Figure 6 and Figure 7. Varying amounts of residual antenna fluorescence were also seen at λ_em = 375 and 385 nm in the carbostyril and the coumarin-sensitized complexes, respectively. The ligand fluorescence quantum yield (Φ_L) of EuL2a^Car was 3.2 times lower than that of EuL1a^Car. In both Eu(III) triamide and tricarboxylate complexes carrying the coumarin antenna Φ_L was ~0.60%. The new Eu and Tb complexes had lower Ln(III) emission quantum yields (Φ_Ln) than the previously reported LnL1a^Car. In the case of LnL2a^Car the lower Φ_Ln is due to the replacement of the tricarboxylate metal-binding fragment with the triamide one. The latter places 6 N-H oscillators close to the Ln(III), which quenches some of the Ln(III) emission. PeT quenching is more efficient for more electron-rich antennae [20] and for more reducible Eu(III) ions [30,36]. As Eu(III) in EuL2a^Car is easier to reduce than in EuL1a^Car, PeT quenching might be more prominent in the former. This suggestion is supported by the significantly smaller Φ_L value of EuL2a^Car compared to that of EuL1a^Car. Φ_Ln of SmL1a^Car and SmL2a^Car were similar to those obtained for Sm(III) emitters sensitized by secondary amide-linked carbostyril antennae [20].

The coumarin-sensitized complexes were less emissive than their carbostyril analogues (Table 2) irrespective of the Ln(III). For Ln = Tb this may be due to the oxygen sensitivity of the excited complex. However, for Eu(III), the reason is less obvious. The coumarin T₁ is well-placed for energy transfer to the Eu(III) ⁵D₂ (21500 cm⁻¹) level. Based on the work of Latva, this antenna should provide excellent sensitization [38]. The Φ_L data do not indicate that there is either PeT quenching of the excited state or direct energy transfer from antenna singlet level (S₁), as the Φ_L for the coumarin-sensitized complexes are similar irrespective of the Ln(III) and the type of metal binding site.

Secondary amide-linked carbostyril complexes have previously been shown to sensitize Yb(III) emission [20]. The Yb(III) complexes of the new ligands with tertiary amide linkers were prepared, and their luminescence spectra in the near infrared were recorded. While the sensitivity of the instrument available to us has decreased considerably over the past years (Figures S56 and S57), the relative emission intensities of the previously reported complexes [20] could be compared to those of YbL1a^Car or YbL2a^Car and YbL1d^Cou or YbL2d^Cou (Figure S58). Only very weak emission was observed for YbL1d^Cou and YbL2d^Cou, much weaker than for either the secondary or the tertiary amide-linked carbostyril-sensitized species. As the most likely sensitization pathway in these emitters is via PeT-back electron transfer [30], these data are in line with an inefficient PeT in the coumarin-carrying complexes.

The Ln(III) luminescence lifetimes (τ) were measured using time-resolved emission spectroscopy. In PIPES-buffered aqueous solutions the triamide-based Eu(III) complexes had τ_H2O between 0.51–0.54 ms (

Table 3. Photophysical properties of EuL in PIPES-buffered H₂O and D₂O solutions ^[a].

Complex	τrad,Ln[b]	${{\mathit{\Phi }}_{\mathbf{L}\mathbf{n}}^{\mathbf{L}\mathbf{n}}}^{\left[\mathbf{c}\right]}$	ηsens[c]	ΦL,D2O[d]	ΦLn,D2O[d]	τrad,Ln (D2O) [e]	${{\mathit{\Phi }}_{\mathbf{L}\mathbf{n}}^{\mathbf{L}\mathbf{n}}\left({\mathbf{D}}_2\mathbf{O}\right)}^{\left[\mathbf{f}\right]}$	ηsens (D2O) [f]
EuL1aCar	5.36	12.2	72.8	2.6 (1.04)	32.3 (3.6)	5.39	40.3 (3.30)	80.3 (1.10)
EuL2aCar	5.11	10.0	27.6	0.83 (1.08)	11.8 (4.2)	5.16	39.7 (3.97)	29.8 (1.08)
EuL1dCou	5.26	11.8	17.6	0.66 (1.10)	8.5 (4.0)	5.31	36.9 (3.15)	23.0 (1.30)
EuL2dCou	5.03	10.7	14.1	0.60 (1.03)	6.3 (4.2)	5.11	39.3 (3.67)	16.0 (1.14)

^[a] [LnL] = 10 µM in 10 mM aqueous PIPES buffer H₂O or D₂O solutions at pH (pD) 6.5. ^[b] In ms, determined using Equation (1). ^[c] In %, determined using Equation (2). ^[d] In %, relative to quinine sulfate (Φ = 0.59) in H₂SO₄ (0.05 M) [39]. ^[e] In ms, determined using Equation (1). [f] In %, determined using Equation (2), in parentheses change relative to the solution in H₂O.

); the lifetimes of the tricarboxylates increased to 0.62–0.66 ms due to the absence of the quenching amide N-H oscillators. The Tb(III) lifetimes in TbL1a^Car and TbL2a^Car were 1.81 ms and 1.50 ms, respectively. In the coumarin-sensitized Tb(III) complexes τ_H2O was much shorter. This is likely due to the oxygen-sensitivity of these emissions, which also precluded the determination of the hydration states (q) of TbL1d^Cou and TbL2d^Cou. For the other Eu(III) and Tb(III) complexes, q was found to be ~1 using the Horrocks [17,40] and Beeby [18] method. This value is in accordance with the octadentate ligand structure and the nine-coordinate environment of Eu(III) and Tb(III). The Sm(III) lifetimes were unsurprisingly short, around 10 μs. These lifetimes are similar to what was previously obtained for Sm(III) emitters in DO3A-type ligands [20].

The reason for the low Φ_Ln of EuL1d^Cou and EuL2d^Cou was next investigated by an analysis of the Eu(III) luminescence spectra. Sensitized Ln(III) emission is the product of multiple processes: photon absorption by the antenna, energy transfer from the antenna to the Ln(III), and Ln(III) luminescence. For Eu(III), it is possible to obtain information on the processes leading to the excited Ln(III) using Equations (1) and (2) [41,42]. The overall luminescence intensity (Φ_Ln) is a product of the luminescent decay of the excited Ln(III), quantified by its intrinsic quantum yield (${\Phi }_{\mathrm{L}\mathrm{n}}^{\mathrm{L}\mathrm{n}}$), and of the sensitization efficiency (η_sens), i.e., the efficiency with which the Ln(III) excited state is populated. ${\Phi }_{\mathrm{L}\mathrm{n}}^{\mathrm{L}\mathrm{n}}$ is the ratio of the observed and radiative luminescence lifetimes (τ_obs and τ_rad, respectively). The latter for Eu(III) can be calculated from the luminescence spectrum using Equation (1), where A_MD,0 is the spontaneous emission probability for the ⁵D₀→⁷F₁ transition of Eu(III) in vacuo with a value of 14.65 s⁻¹, n is the refractive index of the medium (1.333 for H₂O [43] and 1.328 for D₂O [44]), and I_tot and I_MD are the total corrected Eu(III) emission spectrum (521–800 nm) and the ⁵D₀→⁷F₁ band (582–603 nm), respectively [42]. The obtained values are summarized in Table 3.

${\Phi }_{\mathrm{L}\mathrm{n}}^{\mathrm{L}\mathrm{n}}$ quantifies the extent to which the excited Ln(III) can emit in the given coordination environment; similar values were obtained for both types of complexes in H₂O and D₂O. Clearly, the difference in Φ_Eu for the four Eu(III) emitters is primarily down to differences in η_sens. The highest η_sens is obtained for EuL1a^Car, and the second highest for EuL2a^Car. The lower η_sens of EuL2a^Car is likely due to a combination of PeT and also possibly the quenching of the antenna by the amide N-H oscillators. The latter, along with solvent quenching of the antenna, can be eliminated by determining η_sens in D₂O (Table 3). Solvent deuteration does indeed result in a small increase in η_sens across the board. The differences in η_sens between the EuL1a^Car (80.3%) and EuL2a^Car (29.8%), on the one hand, and their coumarin analogues EuL1d^Cou (23.0%) and EuL2d^Cou (16.0%), on the other, remain large.

$$\frac{1}{{\tau }_{\mathrm{r}\mathrm{a}\mathrm{d}}}=A_{\mathrm{M}\mathrm{D},0}\times n^3\left(\frac{I_{\mathrm{t}\mathrm{o}\mathrm{t}}}{I_{\mathrm{M}\mathrm{D}}}\right)$$

(1)

$${\Phi }_{\mathrm{L}\mathrm{n}}={\eta }_{\mathrm{sens}}·{\Phi }_{\mathrm{L}\mathrm{n}}^{\mathrm{L}\mathrm{n}}={\eta }_{\mathrm{sens}}·\frac{{\tau }_{\mathrm{obs}}}{{\tau }_{\mathrm{rad}}}$$

(2)

The addition of external fluoride to a solution of EuL2b^Car was shown to increase Φ_Eu up to 7.6-fold [36]. Most of this increase could be ascribed to Eu(III) stabilization by the negatively charged F⁻ ligand, which reduced PeT quenching. This was apparent from the 3.35-fold larger η_sens in the presence of KF compared to η_sens without KF. To evaluate the cause of the low η_sens in EuL1d^Cou and EuL2d^Cou their photophysical properties in the presence of a large excess of KF (EuL-F) in PIPES-buffered H₂O (

Table 4. Photophysical properties of EuL-F in PIPES-buffered H₂O ^[a].

Complex	ΦL [%] [b]	ΦLn [%] [b]	τrad,Ln[c]	τobs [ms]	${{\mathit{\Phi }}_{\mathbf{L}\mathbf{n}}^{\mathbf{L}\mathbf{n}}}^{\left[\mathbf{d}\right]}$	ηsens[e]
EuL1aCar-F	2.5	9.5 (×1.07)	5.33	0.68	12.8 (×1.05)	73.4 (1.01)
EuL2aCar-F	2.1 (×2.73)	12.3 (×4.56)	4.86	1.01	20.7 (×2.07)	58.9 (2.13)
EuL1dCou-F	0.64 (×1.03)	2.3 (×1.10)	5.19	0.69	13.3 (×1.14)	17.3 (0.98)
EuL2dCou-F	0.65 (×1.12)	3.3 (×2.20)	4.81	1.01	21.0 (×1.96)	15.7 (1.12)

^[a] Formed by the addition of excess KF (0.1 M, 10⁴-fold excess) to a solution of EuL. [EuL] = 10 µM in 10 mM aqueous PIPES buffer solutions at pH 6.5. In parentheses fold increase compared to EuL without added KF. ^[b] Relative to quinine sulfate (Φ = 0.59) in H₂SO₄ (0.05 M) [39]. ^[c] In ms, determined using Equation (1). ^[d] In %, determined using Equation (2). ^[e] In %, determined using Equation (2).

) and D₂O (

Table 5. Antenna- and Eu(III)-based luminescence properties of EuL-F in PIPES-buffered D₂O solution ^[a].

Complex	ΦL [%] [b]	ΦLn [%] [b]	τrad,Ln[c]	τobs [ms]	${{\mathit{\Phi }}_{\mathbf{L}\mathbf{n}}^{\mathbf{L}\mathbf{n}}}^{\left[\mathbf{d}\right]}$	ηsens[e]
EuL1aCar-F	2.6 (1.04)	32.3 (3.4)	5.38	2.29	42.6 (3.33)	75.7 (1.03)
EuL2aCar-F	2.2 (1.05)	31.3 (2.6)	4.89	2.40	49.1 (2.37)	63.7 (1.08)
EuL1dCou-F	0.67 (1.06)	8.2 (3.6)	5.29	2.25	42.6 (3.20)	19.2 (1.11)
EuL2dCou-F	0.69 (1.08)	8.7 (2.6)	4.88	2.39	48.9 (2.33)	17.8 (1.13)

^[a] Formed by the addition of KF (0.1 M, 10⁴-fold excess) to a solution of EuL. [EuL] = 10 µM in 10 mM PIPES buffer solutions in D₂O at pD 6.5. In parentheses: fold increase relative to the solution in H₂O. ^[b] Relative to quinine sulfate (Φ = 0.59) in H₂SO₄ (0.05 M) [39]. ^[c] In ms, determined using Equation (1). ^[d] In %, determined using Equation (2). ^[e] In %, determined using Equation (2).

) solutions were determined.

KF addition increased Φ_L and Φ_Eu 2.73-fold and 4.56-fold in the case of triamide EuL2a^Car-F. This increase was somewhat smaller than seen for the secondary amide EuL2b^Car-F, suggesting that PeT quenching was smaller in former. The increase in Φ_L and Φ_Eu in the case of EuL1a^Car-F was negligible, either because PeT quenching in the absence of fluoride was already small, or because fluoride binding was ineffective due to the overall −1 charge of the complex. The absence of fluoride binding was supported by the fact that the EuL1a^Car-F and EuL1d^Cou-F both retained their water ligand and had q = 1 (Figures S63 and S65). EuL2a^Car-F and EuL2d^Cou-F on the other hand had q = 0, which is in accordance with the presence of a fluoride ligand instead of a water molecule (Table S28, Figures S64 and S66).

Fluoride binding increased Φ_Eu in EuL2d^Cou-F 2.2-fold. This increase was ascribable to an improved intrinsic quantum yield (1.96-fold increase), which in turn was caused by the removal of the inner-sphere water molecule (q = 0). Changes in Φ_L and η_sens were negligible, which suggests that the low Φ_Eu in EuL2d^Cou was not due to PeT quenching.

When fluoride addition was carried out in D₂O, ~2.6-fold higher Φ_Eu was recorded for the triamide complexes compared to what was obtained in H₂O. For the tricarboxylates, the increase was on average 3.5-fold. These changes are consistent with the removal of outer-sphere X-H oscillators, as shown by a comparable increase in the intrinsic quantum yields. The greater Φ_Eu increase for EuL1 further supports the hypothesis that fluoride does not bind to these Eu(III) centers, and thus, X-H oscillators are only removed by deuteration rather than ligand exchange.

Having excluded PeT quenching as the major cause of the low η_sens, poor energy transfer efficiency and inefficient population of the antenna feeding levels remain as possible alternatives. While these were not investigated in depth, we note that the steady-state and time-resolved emission spectra of Eu(III) complexes at 77 K did not contain the triplet emission bands from the antennae (Figure S71). This indicates that T₁ was quenched in these complexes, possibly by energy transfer to Eu(III).

3. Materials and Methods

3.1. General Procedures

¹H NMR (400 MHz), ¹³C NMR (100 MHz) and ¹⁹F NMR (376 MHz) spectra were recorded on a JEOL 400 MHz instrument (JEOL RESONANCE Inc., Tokyo, Japan). Chemical shifts were referenced to residual solvent peaks and are given as follows: chemical shift (δ, ppm), multiplicity (s, singlet; br, broad; d, doublet, t, triplet; q, quartet; m, multiplet), coupling constant (Hz), integration. LC-MS analysis was carried out using an analytical Dionex UltiMate 3000 HPLC instrument (Dionex Softron GmbH, Germering, Germany) coupled to a Thermo Finnigan LCQ DECA XP MAX mass spectrometer (Thermo ELECTRON CORPORATION, San Jose, CA, USA). HR-ESI-MS analyses were performed at the Organisch Chemisches Institut WWU Münster, Germany or at the Stenhagen Analyslab AB, Mölndal, Sweden. All compounds displayed the expected isotope distribution pattern. Anhydrous CH₂Cl₂ was obtained by distillation from CaH₂ under an Ar atmosphere.

Compounds 1 [19], 6 [19], L1a^Car [19], LnL1a^Car (Ln = Eu, Gd, Tb) [19] and L1d^Cou [20] were synthesized following literature methods. All other chemicals were from commercial sources (Sigma Aldrich, St. Louis, MO, USA or Fluorochem, Hadfield, UK) and used as received.

3.2. Paramagnetic ¹H NMR

¹H NMR spectra of Eu complexes were recorded at 400 MHz using the following parameters: cooling for 5 min until the temperature stabilizes at 0 ± 0.1 °C for samples measured in CD₃OD and at 10 ± 0.1 °C for samples measured in D₂O; relaxation delay: 1 s; number of scans: 128; number of points: 131,072; range: −60 to 60 ppm. For Yb complexes measured at r.t. the number of points were 524,288 and the range was from −240 to 240 ppm.

3.3. Chromatography

Preparative chromatography was carried out on silica gel [Normasil 60 chromatographic silica media (40–63 micron)] and aluminum oxide [activated, neutral, Brockmann Activity I, Sigma-Aldrich (Sigma Aldrich, St. Louis, MO, USA)]. Thin layer chromatography was performed on silica-coated (60G F254) glass plates from Merck and aluminum oxide coated with 254 nm fluorescent indicator aluminum plates from Sigma-Aldrich. Samples were visualized by UV-light (UVP LLC, Upland, CA, USA) (254 and 365 nm).

HPLC-analysis was performed on a Dionex UltiMate 3000 system (Dionex Softron GmbH, Germering, Germany) using a Phenomenex Gemini^® C18 TMS end-capped 150 mm × 4.6 mm HPLC column with HPLC water (0.05% formic acid): CH₃CN (0.05% formic acid) eluent system using the methods: (a) 0–8 min: 10→20% and 8–12 min: 20% iso and 12–16 min 20→90% CH₃CN, 0.5 mL/min; (b) 0–8 min: 10% iso and 8–12 min: 10%→50% and 12–16 min 50%→90% CH₃CN, 0.25 mL/min. UV-Vis (UltiMate 3000 Photodiode Array Detector (Dionex Softron GmbH, Germering, Germany)) and ESI-MS detections (Thermo Finnigan LCQ DECA XP MAX (Thermo ELECTRON CORPORATION, San Jose, CA, USA)) were used. Semi-preparative HPLC was performed on Dionex UltiMate 3000 system (Dionex Softron GmbH, Germering, Germany) using a Phenomenex Gemini^® C18 TMS end-capped 150 mm × 30 mm HPLC column with water (0.05% formic acid): MeOH (0.05% formic acid) eluent system with the same UV-detection. The method utilized for semi-preparative purification was the following: 0–6 min: 14% iso and 6–9 min: 95% iso & 9–12 min: 14% iso MeOH, 25 mL/min.

3.4. Electrochemistry

Cyclic voltammograms (CV) were obtained in an argon atmosphere at room temperature (~20 °C) using an AUTOLAB PGSTAT 100 potentiostat, or an AUTOLAB PGSTAT 204N potentiostat, equipped with a 3 mm glassy carbon (GC) working electrode, a Pt wire auxiliary electrode, and a saturated calomel electrode (SCE) as a reference. The solution was stirred between each measurement. The solution was allowed to equilibrate for 10 s at the start potential before starting the measurements. A step potential of −0.9 mV was used for 50, 100 and 200 mV/s scan rates, and of −2 mV was used for 500 and 1000 mV/s scan rates. For measurements in aqueous media the supporting electrolyte was LiCl (0.1 M), in the case of non-aqueous (DMF) solutions it was TBAPF₆ (0.1 M).

General procedure for CV measurements in water: a solution of LiCl (0.1 M) was prepared and pH was set to ~6.5 by addition of NaOH (0.1 M) or HCl (0.1 M). This solution was added to the electrochemical cell, allowed to stir and purged with argon for 10 min prior to each measurement. The working electrode was polished with 0.05 µm alumina on a polishing pad, washed with water and ethanol and dried with air. The three electrodes (GC working electrode, Pt wire auxiliary electrode, and SCE reference electrode) were inserted into the cell setup and a background scan was recorded with a scan rate of 100 mV/s, and four sweeps. A lack of oxygen redox signal verified that oxygen had been removed below detectable levels. The Eu complex (1 mM) was added in the solution, and the pH of the resulting solution was adjusted to ~6.5 (Table S1) by addition of NaOH (0.1 M) or HCl (0.1 M). The resulting solution was stirred and purged with argon for 10 min. Scans were recorded at various scan rates (50 to 1000 mV/s) with four sweeps for each measurement. The voltammograms obtained at various scan rates are shown in Figures S28–S34. The anodic and cathodic peak current intensities (I_pa and I_pc, respectively) were plotted vs. the square root of scan rate and fit to a linear regression to ensure that the electron transfer was heterogenous.

General procedure for CV measurements in DMF: a sample of TBAPF₆ (194 mg) was dissolved in 5 mL of DMF (0.1 M) and purged with argon for 10 min. After detecting blank signal without oxygen redox events, the CVs were recorded as it is described in the procedure for aqueous media, with 1 mM concentration of Eu complex. At the end of each experiment a sample of Ferrocene (Fc) was added at the tip of the spatula into the electrochemical cell to adjust potentials according to Fc⁰/Fc⁺ redox events vs SCE which was then shifted according to the difference vs NHE [45]. The cyclic voltammograms of increasing scan rates are displayed in Figures S35–S41.

3.5. UV-Vis Absorption and Emission Spectroscopy

All measurements were performed in PIPES-buffered HPLC water or D₂O at pH 6.5 or pD 6.5. [LnL] was nominally 10 µM; however, small quantities of Ln salts may diminish this. Glycerol was of 99.9+% purity. Quartz cells with 1 cm optical pathlengths were used for the room temperature measurements. The absorbance spectra were measured by a Varian Cary 100 Bio UV-Visible spectrophotometer (VARIAN AUSTRALIA PTY LTD, Mulgrave, Victoria, Australia). The emission and excitation spectra, lifetimes, time-resolved spectra and quantum yields were recorded on a Horiba FluoroMax-4P (HORIBA Jobin Yvon, Edison, NJ, USA). All emissions were corrected by the wavelength sensitivity (correction function) of the spectrometer. All measurements were performed at room temperature unless stated otherwise.

Quantum yields were measured at room temperature, using quinine sulfate (QS) in H₂SO₄ 0.05 M (Φ_ref = 0.59) as reference [39] in Equation (3). Quantum yields were calculated according to (3), with Φ_s the quantum yield of the sample, Φ_ref the quantum yield of the reference, I the integrated corrected emission intensity of the sample (s) and of the reference (ref), f_A the absorption factor of the sample (s) and of the reference (ref) at the excitation wavelength and n the refractive indexes of the sample (s) and of the reference (ref). The concentration of the complexes was adjusted to obtain an absorbance around the maxima of the antennae matching that of the QS fluorescence standard. The excitation wavelength where the absorption factors of the samples and of the reference were the same was chosen (i.e., where the absorptions are identical). The corrected emission spectra of the sample and reference standard were then measured under the same conditions over the 330–800 nm (320-800 nm for carbostyril complexes) spectral range as well as blank samples containing only the solvent (i.e., PIPES-buffered aqueous solutions). The appropriate blanks were subtracted from their respective spectra, and the antenna fluorescence and Ln(III) luminescence were separated by fitting the section of the antenna emission overlapping the Ln(III) emission with an exponential decay or with a scaled emission spectrum from the corresponding Gd(III) complexes. The quantum yields were then calculated according to (3). The given relative error on the quantum yields (δΦ = ΔΦ/Φ, where ΔΦ is the absolute error) take into account the accuracy of the spectrometer and of the integration procedure [δ(I_s/I_ref) < 2%], an error of 0.59 ± 0.01 on the quantum yield of the reference QS [δ(Φ_ref) < 2%], an error on the ratio of the absorption factors [δ(f_Aref/f_As) < 5%, relative to the fixed absorption factor of the reference QS] and an error on the ratio of the squared refractive indexes [δ(n_s²/n_ref²) < 1%, < 0.25% around 1.333 for H₂O [43] and 1.328 for D₂O [44] on each individual refractive index], which sums to a total estimated relative error that should be δΦ_s < 10%. A limit value of 10% is thus chosen.

$$\Phi =\frac{I_{\mathrm{s}}}{I_{\mathrm{ref}}}\times \frac{f_{\mathrm{Aref}}}{f_{\mathrm{As}}}\times \frac{{\left(n_{\mathrm{s}}\right)}^2}{{\left(n_{\mathrm{ref}}\right)}^2}\times {\Phi }_{\mathrm{ref}}$$

(3)

Low temperature measurements were done in quartz capillaries (0.2 cm optical pathlength) at 77 K by immersion in a liquid N₂-filled quartz Dewar and with addition of glycerol (1 drop) to the solutions (9 drops) measured at room temperature.

Lifetimes in the millisecond range were recorded 0.05 ms after pulsed excitation at the excitation maxima (λ_ex) of either 315 (coumarin) or 327 nm (carbostyril) by measuring the decay of the lanthanide main emission peak (i.e., Sm 600 nm, Eu 615 nm and Tb 545 nm). The increments after the initial delay were adjusted between 0.2–20 μs depending on the lifetime in order to have a good sampling of the decay. The obtained data were fitted by single and double exponential decay models in OriginPro 9 (OriginLab Corporation, Northampton, MA, USA), and the most reliable value was chosen according to the adjusted R² value and the shape of the residuals. A relative error of 10% is typically found among a series of measurements on the same sample.

Hydration numbers (q) were obtained by measuring the lifetimes of the same quantity of complex in a PIPES buffered solution in H₂O and in D₂O and fitting the difference according to the model of Horrocks et al. [17], and Beeby et al. [18].

The NIR emission and excitation spectra were recorded on a Horiba Jobin Yvon Fluorolog3-22 instrument (HORIBA Jobin Yvon, Edison, NJ, USA), and were automatically corrected for wavelength dependent instrument sensitivity.

3.6. Synthetic Procedures and Characterization Data

4. Known compound [22], new procedure. Coumarin 2 (4, 2.00 g, 9.21 mmol) and triethylamine (3.85 mL, 27.6 mmol, 3.0 equiv.) were dissolved in 28.2 mL DMF. The solution was cooled to 0 °C and chloroacetyl chloride (3.19 mL, 36.8 mmol, 4.0 equiv.) was added dropwise while stirring. After 1 h at 0 °C the solution was allowed to warm to r.t. The reaction proceeded for additional 2 h, after which TLC analysis showed full conversion of the starting material. The solution was poured into water (100 mL), which resulted in the formation of a light brown suspension. The mixture was filtered, and the filter cake was washed with water, and the solid was dried in vacuo. The product was obtained as a light brown solid (2.54 g, 94%). ¹H NMR (400 MHz, CDCl₃) δ ppm 7.55 (s, 1H), 7.15 (s, 1H), 6.35 (s, 1H), 4.19 (dq, J = 14.5, 7.0 Hz, 1H), 3.73 (d, J = 2.0 Hz, 2H), 3.26 (dq, J = 14.0, 7.0 Hz, 1H), 2.46 (s, 3H), 2.33 (s, 3H), 1.17 (t, J = 7.0 Hz, 3H); ¹³C NMR (101 MHz, CDCl₃) δ (ppm) = 165.7, 160.2, 152.2, 151.4, 142.4, 132.2, 127.4, 120.6, 117.8, 116.3, 44.2, 41.7, 18.8, 17.6, 12.7; HR-ESI-MS obsd 316.07095, calcd 316.07109 [(M + Na)⁺, M = C₁₅H₁₆NO₃Cl].

L2c^Car. A sample of 1 (188 mg, 0.354 mmol) was dissolved in DMF (5.1 mL) and DIPEA (308 µL, 1.77 mmol, 5.0 equiv.) was added, followed by the addition of 2-bromoacetamide (161 mg, 1.17 mmol, 3.3 equiv.). The reaction mixture was let to stir overnight at r.t. When LCMS analysis showed full conversion to the product, DMF was removed by co-evaporation with toluene. The concentrated reaction mass was suspended in an equal mixture of THF:Et₂O and an off-white precipitate was formed which was filtered and washed with diethyl ether (209 mg, 83%). ¹H NMR (400 MHz, DMSO-d6) δ ppm 1.34–1.51 (s, 9H), 2.10–2.48 and 2.53–4.55 (m, 32 H (29H + 3H solvent residuals), 4.67 (s, 2H), 6.52 (s, 1H), 6.98–7.93 (m, 9H), 11.84 (s, 1H); ¹³C NMR (101 MHz, DMSO-d6) δ ppm 27.7, 48.6, 48.7, 50.1, 50.4, 51.5, 51.8, 52.1, 52.2, 53.6, 54.5, 54.7, 55.6, 56.3, 58.2, 70.2, 81.4, 81.6, 113.8, 117.0, 119.7, 120.9, 125.8, 139.7, 143.1, 146.8, 161.6, 167.9, 170.5, 170.7, 173.2; RP-HPLC t_R = 1.72, 3.55 min, method (a) from general procedures; ESI-MS obsd 702.57, calcd 702.39 (M + H)⁺; HR-ESI-MS obsd 370.6766, calcd 370.6782 [(M + K + H)²⁺, M = C₃₃H₅₁N₉O₈].

L2a^Car. Method A. A sample of L2c^Car (100 mg, 0.142 mmol) was dissolved in a 1:1 mixture of CH₂Cl₂ and TFA (3.2 mL) and this solution was let to stir for overnight at r.t. After full conversion was observed the solvents were removed on the rotary evaporator (co-evaporation with toluene). The crude product was purified on semi-preparative HPLC using the method described in the general procedures to yield a brownish-white solid (78 mg, 85%). For characterization see Method B.

Method B. An excess of TFA (0.76 mL) was added into the vial containing L2c^Car (48 mg, 0.068 mmol). The sticky solid of L2c^Car was sonicated with TFA until complete dissolving. The resulting solution was transferred into a 5 mL round bottom flask and CH₂Cl₂ was added (0.76 mL). The formed beige suspension was stirred overnight at r.t. The next day CH₂Cl₂ was removed under low pressure and TFA was removed via evaporation in a mixture with toluene (5 mL). The remaining oily residue was dissolved in 0.5 mL of MeOH and a large excess of Et₂O (5 mL) was added. The precipitated product was filtered, dissolved in MeOH and the solution was dried under vacuum to afford beige solid (44 mg, 99%). RP-HPLC t_R = 2.35 min, method (b) from general procedures; ¹H NMR (400 MHz, D₂O) δ ppm 2.70–4.75 (m, 31H), 6.61 (s, 1H), 7.33 (dd, J₁ = 8.5 Hz, J₂ = 2.0 Hz, 1H), 7.44 (d, J = 2.0 Hz, 1H), 7.81 (d, J = 8.5 Hz, 1H); ¹³C NMR (101 MHz, D₂O) δ ppm 47.5–52.5, 51.7, 54.0, 54.7, 55.2, 58.4, 70.3, 115.2, 118.5, 118.7, 122.3, 126.3, 138.4, 142.4 (br), 148.8, 164.1, 165.0–176.0 (br), 172.8; ESI-MS obsd 646.51, calcd 646.33 (M + H)⁺; HR-ESI-MS obsd 646.3316, calcd 646.3313 [(M + H)⁺, M = C₂₉H₄₃N₉O₈].

L2d^Cou. Compound 6 (400 mg, 0.932 mmol, 1.0 equiv.), 2-bromoacetamide (514 mg, 3.73 mmol, 4.0 equiv.) and K₂CO₃ (773 mg, 5.59 mmol, 6.0 equiv.) were suspended in MeCN (9.3 mL) and stirred at 60 °C for 2.5 days. The solids were filtered off and the filtrate was concentrated under reduced pressure. The isolated solid was the target product together with an excess of potassium carbonate. Therefore, the solid mixture was dissolved in a minimal amount of MeOH and filtrated from K₂CO₃. After evaporation of the solvent the product was obtained as a light brown solid (442 mg, 79%). ¹H NMR (400 MHz, CD₃OD) δ ppm 7.75 (s, 1H), 7.29 (s, 1H), 6.34 (s, 1H), 4.09 (dq, J = 14.0, 7.1 Hz, 1H), 3.95–3.64 (m, 3H), 3.27 (s, 3H), 3.18–2.39 (m, 20H), 2.32 (s, 3H), 2.08–2.00 (m, 2H), 1.12 (t, J = 7.0 Hz, 3H); ¹³C NMR (101 MHz, CD₃OD) δ ppm 176.1, 171.6, 162.5, 161.4, 158.9, 154.6, 153.3, 144.0, 129.0, 121.6, 118.6, 116.2, 64.6, 62.9, 62.4, 58.3, 44.8, 44.4, 17.5, 13.0; HR-ESI-MS obsd: 623.32682. calcd: 623.32760 [(M + Na)⁺, M = C₂₉H₄₄N₈O₆].

3.7. General Procedure for Ln (III) Complexation

A sample of the appropriate ligand (1 equiv.) in the mixture with the corresponding anhydrous LnCl₃ (1.05 equiv. for LnL1-2a,c^Car and 1.50 equiv. for LnL1-2d^Cou) were dissolved in H₂O:EtOH equal mixture (0.05 M). For the synthesis of LnL1a^Car only water (0.02 M) was used due to the low solubility of L1a^Car in ethanol. The reaction mixtures were stirred at 55 °C for 24 h. The completion of the complexation was observed via TLC analysis. The purification of LnL1a^Car was done with column chromatography on neutral alumina (MeCN:H₂O:NH₄OH (30% aqueous solution), 80:20:3 drops→50:50:9 drops) of the reaction mixtures yielding colorless complexes. The complexes of LnL2a^Car were purified via semi-preparative HPLC using the method described in the general procedures. The coumarin complexes LnL1-2d^Cou and LnL2c^Car were used for characterization without further purification. The isolated compounds contain traces of initial lanthanide chlorides.

LaL1a^Car. 8 mg (65%). ¹H NMR (400 MHz, D₂O, 2 isomers in 1:0.05 ratio, major isomer reported) δ ppm 1.96–4.61 ppm (m, 30.45H both isomers + 3H MeOH), 3.51 (s, 3H), 4.84 (s, 2H), 6.78 (s, 1H), 7.39 (dd, J₁ = 8.5 Hz, J₂= 2.1 Hz, 1H), 7.54 (s, 1H), 7.93 (d, J = 8.5 Hz, 1H); ESI-MS obsd 783.12, calcd 783.15 [(M)⁻, M = C₂₉H₃₆N₆O₁₁La]; λ_em = 375 nm (λ_ex = 327 nm).

SmL1a^Car. 9 mg (72%). ESI-MS obsd 796.18, calcd 796.16 [(M)⁻, M = C₂₉H₃₆N₆O₁₁Sm]; λ_em = 376, 563, 568, 593, 601, 608, 644, 651, 656, 704, 713 nm (λ_ex = 327 nm).

YbL1a^Car. 12 mg (93%). ESI-MS obsd 818.17, calcd 818.18 [(M)⁻, M = C₂₉H₃₆N₆O₁₁Yb]; λ_em = 375 nm (λ_ex = 327 nm).

LuL1a^Car. 10 mg (77%). ¹H NMR (400 MHz, D₂O, 2 isomers in 1:0.05 ratio, major isomer reported) δ ppm 2.25–4.15 ppm (m, 30.45H both isomers + 3H MeOH), 3.51 (s, 3H), 4.85 (s, 2H), 6.79 (s, 1H), 7.43 (d, J = 8.5 Hz, 1H), 7.57 (s, 1H), 7.93 (d, J = 8.5 Hz, 1H); ESI-MS obsd 819.19, calcd 819.19 [(M)⁻, M = C₂₉H₃₆N₆O₁₁Lu]; λ_em = 375 nm (λ_ex = 327 nm).

LaL2a^Car. 31 mg (quant.). ¹H NMR (400 MHz, D₂O, 2 isomers in 1:0.2 ratio, major isomer reported) δ ppm 1.97–3.23 and 3.28–4.75 ppm (m, 37.2H both isomers), 3.53 (s, 3H), 6.77 (s, 1H), 7.34 (d, J = 9.0 Hz, 1H), 7.48 (s, 1H), 7.91 (d, J = 8.0 Hz, 1H); ESI-MS obsd 783.20, calcd 783.22 [(M)⁺, M = C₂₉H₄₂N₉O₈La]; λ_em = 373 nm (λ_ex = 327 nm).

SmL2a^Car. 23 mg (quant.). ESI-MS obsd 796.24, calcd 796.24 [(M)⁺, M = C₂₉H₄₂N₉O₈Sm]; λ_em = 374, 562, 567, 593, 600, 607, 643, 650, 654, 704, 712 nm (λ_ex = 327 nm).

EuL2a^Car. 13 mg (47%). RP-HPLC t_R = 1.17 min, method (a) from general procedures; ESI-MS obsd 398.72, calcd 398.62 (M)²⁺; HR-ESI-MS obsd 398.61797, calcd 398.61801 [(M)²⁺, M = C₂₉H₄₂N₉O₈Eu]; λ_em = 375, 579, 589, 594, 615, 622, 653, 682, 687, 695, 699, 752, 760 nm (λ_ex = 327 nm).

GdL2a^Car. 6 mg (44%). RP-HPLC t_R = 1.18 min, method (a) from general procedures; ESI-MS obsd 401.13, calcd 401.12 (M)²⁺; HR-ESI-MS obsd 401.11986, calcd 401.11965 [(M)²⁺, M = C₂₉H₄₂N₉O₈Gd]; λ_em = 375 nm (λ_ex = 327 nm).

TbL2a^Car. 9 mg (65%). RP-HPLC t_R = 1.18 min, method (a) from general procedures; ESI-MS obsd 401.80, calcd 401.62 (M)²⁺; HR-ESI-MS obsd 401.61996, calcd 401.61994 [(M)²⁺, M = C₂₉H₄₂N₉O₈Tb]; λ_em = 374, 487, 542, 545, 581, 587, 620, 640, 650, 667, 680 nm (λ_ex = 327 nm).

YbL2a^Car. 14 mg (74%). ESI-MS obsd 817.19, calcd 817.25 [(M − H)⁻, M = C₂₉H₄₂N₉O₈Yb]; λ_em = 374 nm (λ_ex = 327 nm).

LuL2a^Car. 20 mg (quant.). ¹H NMR (400 MHz, D₂O, 2 isomers in 1:0.2 ratio, major isomer reported) δ ppm 2.44–3.00 and 3.30–4.17 ppm (m, 37.2H both isomers), 3.51 (s, 3H), 6.79 (s, 1H), 7.38 (d, J = 8.0 Hz, 1H), 7.55 (s, 1H), 7.93 (d, J = 8.5 Hz, 1H); ESI-MS obsd 818.31, calcd 818.25 [(M − H)⁻, M = C₂₉H₄₂N₉O₈Lu]; λ_em = 373 nm (λ_ex = 327 nm).

LaL2c^Car. 4.7 mg (92%). Partial characterization due to the complex being unstable. ¹H NMR (400 MHz, CD₃OD, mixture of LaL2a^Car and LaL2c^Car in 1:0.45 ratio, LaL2c^Car peaks reported) δ ppm 1.30 (s, 4.05 H), 2.10-3.27 and 3.57–4.35 ppm (m, 37.7H both species), 3.51 (s, 4.35H), 4.76 (s, 2.9H), 6.76 (s, 0.45H), 7.41 (m, 0.45H), 7.58 (s, 0.45H), 7.89 (d, J = 8.0 Hz, 0.45H); λ_em = 374 nm (λ_ex = 327 nm).

SmL2c^Car. 5.1 mg (98%). Partial characterization due to the complex being unstable. λ_em = 374, 562, 567, 593, 600, 607, 643, 650, 654, 704, 711 nm (λ_ex = 327 nm).

EuL2c^Car. 6.4 mg (91%). Partial characterization due to the complex being unstable. RP-HPLC t_R = 1.28 min, method (a) from general procedures; ESI-MS obsd 449.62, calcd 449.14 (M + Na − H)²⁺, M = C₃₃H₅₁N₉O₈Eu; λ_em = 374, 579, 589, 594, 615, 623, 653, 682, 687, 695, 699, 752, 761 nm (λ_ex = 327 nm). HRMS shows only the hydrolyzed complex EuL2a^Car. The LCMS obtained after overnight stirring shows 1:2 ratio of hydrolyzed:^tBu species (Figure S1).

GdL2c^Car. 4.9 mg (70%). Partial characterization due to the complex being unstable. RP-HPLC t_R = 1.27 min, method (a) from general procedures; ESI-MS obsd 451.70, calcd 451.64 (M + Na − H)²⁺, M = C₃₃H₅₁N₉O₈Gd; λ_em = 374 nm (λ_ex = 327 nm). HRMS shows only the hydrolyzed complexes GdL2a^Car. The LCMS obtained after overnight stirring shows 1:2 ratio of hydrolyzed:^tBu species (Figure S2).

TbL2c^Car. 3.8 mg (54%). Partial characterization due to the complex being unstable. RP-HPLC t_R = 1.27 min, method (a) from general procedures; ESI-MS obsd 452.57, calcd 452.14 (M + Na − H)²⁺, M = C₃₃H₅₁N₉O₈Tb; λ_em = 373, 487, 545, 545, 582, 587, 620, 640, 650, 667, 680 nm (λ_ex = 327 nm). HRMS shows only the hydrolyzed complexes TbL2a^Car. The LCMS obtained after overnight stirring shows 1:2 ratio of hydrolyzed:^tBu species (Figure S3).

YbL2c^Car. 5 mg (98%). Partial characterization due to the complex being unstable. λ_em = 374 nm (λ_ex = 327 nm).

LuL2c^Car. 4.8 mg (94%). Partial characterization due to the complex being unstable. ¹H NMR (400 MHz, CD₃OD, mixture of LaL2a^Car and LaL2c^Car in 1:0.45 ratio, both species reported) δ ppm 1.29 (s, 2.25 H), 2.47–3.08 and 3.57–4.00 ppm (m, 32.5H both species), 3.51 (s, 3.75H), 4.76 (s, 2.5H), 6.76–6.79 (s, 1.25H), 7.37 (dd, J₁ = 8.5 Hz, J₂ = 2.0 Hz, 1.25H), 7.58 (s, 1.25H), 7.91–7.98 (d, J = 8.0 Hz, 1.25H); λ_em = 375 nm (λ_ex = 327 nm).

LaL1d^Cou. 27 mg (quant.); ¹H NMR (400 MHz, D₂O, 2 isomers in 1:0.33 ratio, major isomer reported) δ ppm 1.16 (t, J = 7.0 Hz, 3H), 2.36 (s, 3H), 2.50 (s, 3H) 2.82–3.90 ppm (m, 33.25H both isomers + 3H MeOH), 4.19 (dq, J₁ = 15.0 Hz, J₂ = 7.5 Hz, 1H), 6.46 (s, 1H), 7.48 (s, 1H), 7.86 (s, 1H); HR-ESI-MS m/z: obsd: 762.16330, calcd: 762.16252 [(M + Na)⁺, M = C₂₉H₃₈N₅O₉La]; λ_em = 387 nm (λ_ex = 315 nm).

SmL1d^Cou. Known compound [20], new procedure. 16 mg (quant.); HR-ESI-MS m/z: obsd: 775.17683, calcd: 775.17601 [(M + Na)⁺, M = C₂₉H₃₈N₅O₉Sm]; λ_em = 383, 563, 568, 592, 601, 608, 650 nm (λ_ex = 315 nm).

EuL1d^Cou. Known compound [20], new procedure. 33 mg (quant.); HR-ESI-MS m/z: obsd: 754.19657, calcd: 754.19573 [(M + H)⁺, M = C₂₉H₃₈N₅O₉Eu]; λ_em = 387, 579, 588, 594, 614, 623, 653, 682, 688, 694, 700 nm (λ_ex = 315 nm).

GdL1d^Cou. Known compound [20], new procedure. 18 mg (quant.); HR-ESI-MS m/z: obsd: 781.18182, calcd: 781.18098 [(M + Na)⁺, M = C₂₉H₃₈N₅O₉Gd]; λ_em = 387 nm (λ_ex = 315 nm).

TbL1d^Cou. Known compound [20], new procedure. 18 mg (quant.); HR-ESI-MS m/z: obsd: 782.18214, calcd: 782.18152 [(M + Na)⁺, M = C₂₉H₃₈N₅O₉Tb]; λ_em = 389, 487, 542, 545, 583, 587, 620, 651, 668, 682 nm (λ_ex = 315 nm).

YbL1d^Cou. 33 mg (quant.); HR-ESI-MS m/z: obsd: 797.19648, calcd: 797.19553 [(M + Na)⁺, M = C₂₉H₃₈N₅O₉Yb]; λ_em = 388 nm (λ_ex = 315 nm).

LuL1d^Cou. Known compound [20], new procedure. 26 mg (quant.); ¹H NMR (400 MHz, D₂O, 2 isomers in 1:0.1 ratio, major isomer reported) δ ppm 1.16 (t, J = 7.0 Hz, 3H), 2.36 (s, 3H), 2.50 (s, 3H) 2.57–3.72 ppm (m, 27.5H both isomers + 3H MeOH), 4.17 (m, 1H), 6.46 (s, 1H), 7.48 (s, 1H), 7.87 (s, 1H); HR-ESI-MS m/z: obsd: 798.19742, calcd: 798.19694 [(M + Na)⁺, M = C₂₉H₃₈N₅O₉Lu]; λ_em = 388 nm (λ_ex = 315 nm).

LaL2d^Cou. 24 mg (quant.); ¹H NMR (400 MHz, D₂O, 2 isomers in 1:0.05 ratio, major isomer reported) δ ppm 1.16 (m, 3H), 2.33 (s, 3H), 2.50 (s, 3H) 2.7–4.27 ppm (m, 27.3H both isomers + 3H MeOH), 6.48 (s, 1H), 7.41 (s, 1H), 7.87 (s, 1H); HR-ESI-MS m/z: obsd: 369.11778, calcd: 369.11790 [(M − H)²⁻, M = C₂₉H₄₄N₈O₆La]; λ_em = 387 nm (λ_ex = 315 nm).

SmL2d^Cou. 16 mg (quant.); HR-ESI-MS m/z: obsd: 375.62442, calcd: 375.62464 [(M − H)²⁻, M = C₂₉H₄₄N₈O₆Sm]; λ_em = 385, 563, 567, 593, 600, 606, 648 nm (λ_ex = 315 nm).

EuL2d^Cou. 32 mg (quant.); HR-ESI-MS m/z: obsd: 375.12529, calcd: 376.12547 [(M − H)²⁻, M = C₂₉H₄₄N₈O₆Eu]; λ_em = 387, 579, 593, 615, 623, 653, 682, 688, 700 nm (λ_ex = 315 nm).

GdL2d^Cou. 17 mg (quant.); HR-ESI-MS m/z: obsd: 378.62684, calcd: 378.62711 [(M − H)²⁻, M = C₂₉H₄₄N₈O₆Gd]; λ_em = 386 nm (λ_ex = 315 nm).

TbL2d^Cou. 17 mg (quant.); HR-ESI-MS m/z: obsd: 379.12710, calcd: 379.12740 [(M − H)²⁻, M = C₂₉H₄₄N₈O₆Tb]; λ_em = 387, 487, 545, 582, 587, 620, 650, 667, 680 nm (λ_ex = 315 nm).

YbL2d^Cou. 32 mg (quant.); HR-ESI-MS m/z: obsd: 386.63411, calcd: 386.63439 [(M − H)²⁻, M = C₂₉H₄₄N₈O₆Yb]; λ_em = 387 nm (λ_ex = 315 nm).

LuL2d^Cou. 25 mg (quant.); ¹H NMR (400 MHz, D₂O, 2 isomers in 1:0.15 ratio, major isomer reported) δ ppm 1.17 (t, J = 7.0 Hz, 3H), 2.36 (s, 3H), 2.51 (s, 3H) 2.57–3.06 and 3.20–4.25 ppm (m, 29.9H both isomers + 9H 3MeOH), 6.48 (s, 1H), 7.43 (d, J = 23.0 Hz, 1H), 7.88 (d, J = 10.0 Hz, 1H); HR-ESI-MS m/z: obsd: 387.13483, calcd: 387.13511 [(M − H)²⁻, M = C₂₉H₄₄N₈O₆Lu]; λ_em = 388 nm (λ_ex = 315 nm).

4. Conclusions

Octadentate ligands carrying 4-methoxymethylcarbostyril or coumarin 2 sensitizing antennae mounted on a DO3A ligand binding site were prepared and characterized. Paramagnetic ¹H NMR spectroscopy indicated that the coumarin-carrying ligands existed as a mixture of SAP rotamers, possibly due to the steric clash between the antenna 6-Me group and the linker carbonyl. Only one species was seen for the carbostyrils with H in the 6-position. The Eu(III)/Eu(II) reduction potentials were found to be quite unaffected by the nature of the amide linker (secondary vs tertiary) between the antenna and the metal binding site. The Eu(III) oxidation state was nevertheless stabilized by a negatively charged methylcarboxylate substituent on the tertiary amide linker.

Ln(III) (Ln = Eu, Tb, Sm) emission was detected from both coumarin and carbostyril-sensitized complexes. The coumarin antenna was unexpectedly ineffective in sensitizing Ln(III) emission. In the case of the Eu(III) complexes, the low overall quantum yield appears to be caused by low sensitization efficiency. While a large part of the excitation energy was eliminated due to PeT in the triamide carbostyril complex, this quenching pathway did not seem prominent in the coumarin 2-sensitized +3-charged species. This was indicated by the lack of improvement in Eu(III) sensitization efficiency upon the stabilization of the +3 oxidation state via fluoride binding. These results underscore the importance of understanding the structures of emitters in solution and show that superficially similar complexes can be subject to different quenching processes.