Cd²⁺-Selective Fluorescence Enhancement of Bisquinoline Derivatives with 2-Aminoethanol Skeleton

Abstract

The development of fluorescent Cd²⁺ sensors requires strict selectivity over Zn²⁺ because of the high availability of Zn²⁺ in the natural environment. In this paper, bisquinoline-based fluorescent sensors with a 2-aminoethanol backbone were investigated. The weak coordination ability of quinoline compared to well-studied pyridine is suitable for Cd²⁺ selectivity rather than Zn²⁺. In the presence of 3 equiv. of metal ions, TriMeO-N,O-BQMAE (N,O-bis(5,6,7-trimethoxy-2-quinolylmethyl)-2-methylaminoethanol (3)), as well as its N,N-isomer TriMeO-N,N-BQMAE (N,N-bis(5,6,7-trimethoxy-2-quinolylmethyl)-2-methoxyethylamine (6)), exhibits Cd²⁺-selective fluorescence enhancement over Zn²⁺ in DMF-HEPES buffer (1:1, 50 mM HEPES, 0.1 M KCl, pH = 7.5) (I_Zn/I_Cd = 26–34%), which has similar selectivity in comparison to the corresponding ethylenediamine derivative TriMeOBQDMEN (N,N’-bis(5,6,7-trimethoxy-2-quinolylmethyl)-N,N’-dimethylethylenediamine) under the same experimental condition (I_Zn/I_Cd = 24%). The fluorescence mechanisms of N,O- and N,N-isomers of BQMAE are quite different, judging from the fluorescence lifetimes of their metal complexes. The Cd²⁺ complex with TriMeO-N,O-BQMAE (3) exhibits a long fluorescence lifetime similar to that of TriMeOBQDMEN via intramolecular excimer emission, whereas the Cd²⁺ complex with TriMeO-N,N-BQMAE (6) exhibits a short lifetime from monomer emission.

1. Introduction

Cadmium is a toxic metal and exists in vast areas of natural environments. Extensive use of cadmium in industry, including batteries and pigments, facilitates the exposure of cadmium to the air, water, and soils. As a result, cadmium-related health problems in humans are recent serious concerns [1,2]. In this context, the detection and quantification of cadmium in the environment, especially in water, are of urgent demand [3,4,5].

The fluorescence method is a powerful tool for the detection of metal ions and biologically important small molecules [6,7,8,9,10,11,12,13,14,15]. Its sensitivity, rapidness, and convenience allow us to visualize the target effectively. Specificity is the most important requisite for fluorescence sensing and owes to the rational molecular design of fluorescent probes. One of the most difficult and important problems in fluorescence detection is the discrimination between Cd²⁺ and Zn²⁺ because these two metal ions slightly differ in their ionic radii, which is only 21 pm [16,17,18,19]. The weak basicity of the quinoline nitrogen atom compared to that of pyridine offers a suitable coordination environment around the larger and softer Cd²⁺ center than Zn²⁺. Therefore, many quinoline-based fluorescent Cd²⁺ sensors have been extensively developed [20,21,22,23,24,25].

In our laboratory, several fluorescent probes for Zn²⁺ and Cd²⁺ based on the polyquinoline structure have been reported [26]. Considering the larger coordination number of Cd²⁺ in comparison to Zn²⁺, the hepta- and octadentate tetrakisquinoline derivatives such as TQOPEN [27] (N,N,N’,N’-tetrakis(2-quinolylmethyl)-3-oxa-1,5-pentanediamine) and TriMeOBAPTQ [28] (N,N,N’,N’-tetrakis(5,6,7-trimethoxy-2-quinolylmethyl)-2,2′-(N,N’-dimethylethylenediamino)dianiline) were developed. Recently, the tetradentate bisquinoline derivative TriMeOBQDMEN [29] (N,N’-bis(5,6,7-trimethoxy-2-quinolylmethyl)-N,N’-dimethylethylenediamine, Scheme 1) opened the door to a new field of easily accessible small molecular fluorescent sensors for Cd²⁺. The TriMeOBQDMEN exhibits Cd²⁺-selective fluorescence enhancement in DMF-H₂O (1:1) (I_Zn/I_Cd = 22% in the presence of 1 equiv. of metal ion). The fluorescent Cd²⁺/Zn²⁺ discrimination mechanism of TriMeOBQDMEN is not fully understood; however, the bis(µ-chloro) dicadmium complex was characterized using BQDMEN (Scheme 1) as a plausible fluorescent species [29]. In order to further develop the potential of tetradentate bisquinoline derivatives as Cd²⁺-specific fluorescent sensors, here we designed oxygen-containing N3O1 tetradentate bisquinoline derivatives of BQDMEN, namely, N,O-BQMAE (N,O-bis(2-quinolylmethyl)-2-methylaminoethanol (1)) and its methoxy-substituted analogs 2 and 3 (Scheme 1). These oxygen-containing ligands were expected to possess reduced metal binding affinity compared with N4 tetradentate ligands and, therefore, a library with a set of fluorescent ligands with a wide variety of thresholds to quantify the concentration of Cd²⁺ can be constructed. Moreover, N,N-BQMAE (N,N-bis(2-quinolylmethyl)-2-methoxyethylamine (4)) and its methoxy-substituted analogs 5 and 6 (Scheme 1) were also examined as a set of regioisomers of N,O-BQMAE derivatives to expand the compound library.

2. Results and Discussion

2.1. Synthesis of Ligands

The newly designed six ligands (1–6) based on a 2-aminoethanol skeleton were prepared according to Scheme S1 using corresponding amines and methoxy-substituted chloromethylquinolines. N-alkylation of 2-methylaminoethanol followed by O-alkylation using n-butyllithium affords 1–3 in quite low yields (6–19%) due to the unoptimized experimental procedure for the second step. The synthesis of 4–6 was conducted in one step, starting from 2-methoxyethylamine in excellent yields (96% to quantitative yield). Structure and purity of ligands 3–6 were supported by ¹H/¹³C NMR and elemental analysis. Structure of ligands 1 and 2 was confirmed by ¹H/¹³C NMR. The complicated coupling pattern at the aliphatic proton region of 2 in ¹H NMR spectrum is still under investigation.

2.2. Fluorescence Spectral Changes of N,O-BQMAE Derivatives 1–3

The metal ion-induced fluorescence spectral changes of N,O-BQMAE derivatives 1–3 were examined in DMF-HEPES buffer (1:1, 50 mM HEPES, 0.1 M KCl, pH = 7.5) (Figure 1a–f). Absorbance spectral changes and excitation spectrum for 3 are shown in Figure S1. The previously reported N4 tetradentate fluorescent Cd²⁺ sensor TriMeOBQDMEN [29] was also re-examined in DMF-HEPES buffer (1:1, 50 mM HEPES, 0.1 M KCl, pH = 7.5) in this study for strict comparison (Figure 1g,h) because all measurements in the previous report for TriMeOBQDMEN were performed in DMF-H₂O solution. As shown in Figure 1, only trimethoxy-substituted derivative 3 from the N,O-BQMAE derivatives exhibits distinct fluorescence enhancement upon addition of metal ions. The N,O-BQMAE (1) bearing unsubstituted quinolines does not respond to any metal ions except for a small response toward Cd²⁺, Zn²⁺, and Pb²⁺ (Figure 1b). The mono-methoxy derivative 6-MeO-N,O-BQMAE (2) only exhibits a Cu²⁺-induced fluorescence quenching (Figure 1d). The trimethoxy derivative TriMeO-N,O-BQMAE (3) affords apparent Cd²⁺-selective fluorescence enhancement over Zn²⁺ (I_Zn/I_Cd = 34%) (Figure 1f), which is a bit diminished in selectivity in comparison with TriMeOBQDMEN under the same experimental condition (I_Zn/I_Cd = 24%) (Figure 1h and

Table 1. Fluorescent and Thermodynamic Properties for Cd²⁺ and Zn²⁺ Complexes of BQMAE Derivatives 3–6 and TriMeOBQDMEN ^a.

Ligand	λex (nm)	λem (nm) b	IZn/ICdc	ϕCd	Kd for Cd2+ Complex (M)	Kd for Zn2+ Complex (M)
TriMeO-N,O-BQMAE (3)	338	470	34%	0.281 d	(1.4 ± 0.2) × 10−5	(8.2 ± 0.4) × 10−5
N,N-BQMAE (4)	317	366	87%	0.018 d	(8.1 ± 1.5) × 10−6	(5.2 ± 0.3) × 10−5
6-MeO-N,N-BQMAE (5)	337	404	51%	0.171 d	(6.4 ± 0.3) × 10−7	(1.6 ± 0.2) × 10−5
TriMeO-N,N-BQMAE (6)	340	465	26%	0.228 d	~10−7	(1.5 ± 0.2) × 10−5
TriMeOBQDMEN	335	458	24% (41% e)	0.293 c,e	(2.6 ± 1.0) × 10−7	(2.7 ± 0.5) × 10−6

^a In DMF-HEPES buffer (1:1, 50 mM HEPES, 0.1 M KCl, pH = 7.5) at 25 °C. ^b For Cd²⁺ complex. ^c In the presence of 3 equiv. of metal ions. ^d In the presence of 10 equiv. of metal ions. ^e In DMF-H₂O (1:1) [29].

).

The metal binding affinity (K_d) was calculated from the non-linear fitting of titration curves with Cd²⁺ and Zn²⁺, indicating that TriMeO-N,O-BQMAE (3) has two orders lower metal binding ability than TriMeOBQDMEN (Figures S3 and S4 and Table 1). This is because the coordination ability of the nitrogen atom is higher than that of oxygen, but two orders of difference are quite significant for such a small modification.

2.3. Fluorescence Spectral Changes of N,N-BQMAE Derivatives 4–6

The metal ion-induced fluorescence spectral changes of N,N-BQMAE derivatives 4–6 were next examined (Figure 2). Absorbance spectral changes and excitation spectrum for 6 are shown in Figure S2. Because of the synthetic inconvenience of N,O-BQMAE derivatives 1–3 and disappointing fluorescence response of 1 and 2 toward metal ions due to the weak metal binding ability of these compounds, the other N3O1 ligands with a bis(2-quinolylmethyl)amine moiety were explored. As shown in Figure 2, even in the unsubstituted quinoline derivative N,N-BQMAE (4), metal-induced fluorescence enhancement was clearly observed, which is largely different from the N,O-isomer 1 and 2. In the presence of a large excess of metal ions to assure the complete complexation, Zn²⁺ induces a slightly higher fluorescence intensity of 4 than Cd²⁺ (Figure S5). Other methoxy-substituted derivatives, including TriMeO-N,O-BQMAE (3) and TriMeOBQDMEN in the previous section, show complete metal binding in the presence of 3 equiv. of metal ions, and exhibit consistent preference of Cd²⁺ over Zn²⁺ (Figures S3, S4, S6 and S7). As the number of methoxy substituents increases, the fluorescence Cd²⁺ selectivity is improved (Figure 2b,d,f, and Table 1), and the metal binding affinity is enhanced (Figures S5–S7 and Table 1). The three trimethoxy-substituted derivatives, namely, TriMeO-N,O-BQMAE (3) (K_d(Zn) = ~10⁻⁴), TriMeO-N,N-BQMAE (6) (K_d(Zn) = ~10⁻⁵), and TriMeOBQDMEN (K_d(Zn) = ~10⁻⁶) exhibit approximately one order different metal binding affinities in this order (Table 1). Since the fluorescence metal ion selectivity of these compounds is essentially the same (I_Zn/I_Cd = ~30%), this ligand library is a nice set of fluorescent sensors with different metal binding affinities, which is useful for convenient determination of Cd²⁺ concentration.

2.4. Effect of pH and Estimation of LOD

The effect of pH on the fluorescence intensity of the Cd²⁺ complex with TriMeO-N,O-BQMAE (3) and TriMeO-N,N-BQMAE (6) was examined (Figure 3). Protonation to the ligand at low pH region and coordination of hydroxide ion to Cd²⁺ at high pH region prevent the binding of Cd²⁺ to the probes. The pH window for fluorescent detection of Cd²⁺ is wider for 6 (Figure 3b) than 3 (Figure 3a), reflecting the difference in metal binding affinity of the ligands (Table 1). For 6, fluorescence intensity is fairly stable between pH = 4 and pH = 11. The limit of detection (LOD) for 6 was estimated as low as 17 nM (Figure S8), which is comparable to the other tetradentate fluorescent Cd²⁺ probes in the literature [24,30,31]. This value (LOD = 17 nM) is lower than the environmental limit of water in Japan (3 ppb, 27 nM). Accordingly, the limit of quantitation (LOQ) is calculated to be 57 nM. The above properties of TriMeO-N,N-BQMAE (6) demonstrate the potential of this ligand for practical uses.

2.5. Fluorescence Lifetimes of N,O- and N,N-BQMAE Derivatives 3–6 and TriMeOBQDMEN

The fluorescence lifetimes (τ) for Cd²⁺ and Zn²⁺ complexes with N,O- and N,N-BQMAE derivatives 3–6 were measured in the presence of 2 equiv. of metal ions, in which most of the ligands form metal complexes. For comparison, TriMeOBQDMEN was also re-examined in DMF-HEPES buffer (1:1, 50 mM HEPES, 0.1 M KCl, pH = 7.5).

Table 2. Fluorescence Lifetimes for Cd²⁺ and Zn²⁺ Complexes of BQMAE Derivatives 3–6 and TriMeOBQDMEN ^a,b.

Ligand	Metal Ion	λem (nm)	BPF (nm) c	τ (nsec) d
TriMeO-N,O-BQMAE (3)	—	—	460	1.2 (17%), 4.3 (74%), 14.5 (9%)
	Cd2+	470	460	2.6 (5%), 11.5 (14%), 29.7 (81%)
	Zn2+	484	460	2.1 (14%), 7.5 (34%), 22 (52%)
N,N-BQMAE (4)	—	—	370	1.2 (14%), 5.5 (40%)
	Cd2+	364	370	5.4 (20%)
	Zn2+	370	370	1.0 (43%), 5.3 (22%)
6-MeO-N,N-BQMAE (5)	—	—	400	2.7 (45%), 6.0 (46%)
	Cd2+	404	400	1.1 (5%), 3.0 (1%), 6.2 (94%)
	Zn2+	408	400	3.4 (27%), 6.8 (66%)
TriMeO-N,N-BQMAE (6)	—	—	460	1.2 (34%), 4.0 (24%), 16.8 (42%)
	Cd2+	465	460	1.6 (2%), 8.5 (7%), 17.7 (91%)
	Zn2+	480	490	1.2 (12%), 7.4 (33%), 16.5 (54%)
TriMeOBQDMEN	—	—	460	1.1 (29%), 4.3 (44%), 21.4 (27%)
	Cd2+	472	460	3.5 (1%), 14.6 (5%), 31.8 (91%)
	Zn2+	487	490	1.7 (2%), 7.8 (9%), 20.2 (88%)

^a In DMF-HEPES buffer (1:1, 50 mM HEPES, 0.1 M KCl, pH = 7.5). ^b In the presence of 2 equiv. of metal ions. ^c Bandpath filter used (±10 nm). ^d Components with extremely short lifetimes (<1 nsec) were omitted.

summarizes the results.

As discussed previously for TriMeOBQDMEN based on the investigation in DMF-H₂O solution [29], the fluorescent Cd²⁺ selectivity of this ligand is explained by the difference in fluorescence lifetime in Cd²⁺ and Zn²⁺ complexes, in which a very long (τ = ~30 nsec) lifetime was observed exclusively for the Cd²⁺ complex. From the crystal structure of the Cd²⁺ complex with BQDMEN ([Cd₂(µ-Cl)₂(BQDMEN)₂]²⁺), this long lifetime was assigned to be an intramolecular excimer emission from the highly stacked two quinoline rings with interligand interaction. Interestingly, in DMF-HEPES buffer (1:1, 50 mM HEPES, 0.1 M KCl, pH = 7.5), this long fluorescence lifetime was preserved for the TriMeOBQDMEN-Cd²⁺ complex (τ = 32 nsec) (Table 2). Here again, the Zn²⁺ complex with TriMeOBQDMEN exhibits a shorter lifetime (τ = 20 nsec) than Cd²⁺. These results strongly support that the fluorescence lifetime is a critical measure for the mechanism of fluorescence and structure of the metal complexes, which is independent of the solvent system used for measurements.

Similar distinct difference in fluorescence lifetimes of Cd²⁺ and Zn²⁺ complexes was found for TriMeO-N,O-BQMAE (3) (τ_Cd = 30 nsec and τ_Zn = 22 nsec). This is because the ligand structure of N,O-BQMAE resembles that of BQDMEN (Scheme 1) and rationally suggests the specific formation of a bis(µ-chloro) dinuclear cadmium complex for 3. On the other hand, the N,N-BQMAE derivatives did not exhibit any differences in fluorescence lifetimes between Cd²⁺ and Zn²⁺ complexes. Even in the trimethoxy-substituted TriMeO-N,N-BQMAE (6), the lifetimes for Cd²⁺ and Zn²⁺ complexes are completely the same (τ_Cd = 18 nsec and τ_Zn = 17 nsec), and these values are too short to consider the excimer formation discussed above. These values (τ = ~20 nsec) are in good agreement with those for monomer complexes such as TriMeOTQTACN [28] (N,N’,N’’-tris(5,6,7-trimethoxy-2-quinolylmethyl)-1,4,7-triazacyclononane) (τ_Cd = 21 nsec and τ_Zn = 23 nsec) measured in MeOH-HEPES buffer (1:1, 50 mM HEPES, 0.1 M KCl, pH = 7.5). As a result, TriMeO-N,O-BQMAE (3) and TriMeO-N,N-BQMAE (6) adopt a different fluorescence mechanism possibly due to the different complex structure. Slightly diminished fluorescence quantum yield of 6 (ϕ_Cd = 0.23) in comparison to those of 3 (ϕ_Cd = 0.28) and BQDMEN (ϕ_Cd = 0.29) (Table 1) may reflect such differences. Although extensive trials of crystallization of Cd²⁺ complexes of 1–6 have been unsuccessful so far, the proton NMR indicates the coordination of quinoline and methoxy moieties to the metal center for N,N-BQMAE (4)-Cd²⁺ complex, affording the chemical shift changes (Figure S9). The characteristic peak splitting of the methylene signal around 4.4–4.7 ppm also supports the coordination of 2-aminomethylquinoline moiety to the metal center.

At the present stage, the details in the fluorescent Cd²⁺/Zn²⁺ discrimination mechanism of TriMeO-N,N-BQMAE (6) is still unknown, but the dynamic equilibrium of mononuclear ([Cd(L)Cl]⁺) and µ-chloro-bridged dinuclear ([Cd₂(µ-Cl)₂(L)₂]²⁺) cadmium complexes needs to be considered. To answer these critical questions, related projects using other ligand systems with different skeletons are now in progress in our laboratory.

3. Experimental

3.1. Materials and Methods

All reagents and solvents used for the preparation of probe molecules were obtained from commercial sources and used as received. DMF (N,N-dimethylformamide) used for spectroscopic measurements was of spectral grade (Dojin, Spectrosol). All aqueous solutions were prepared using Milli-Q water (Millipore, Merck, Germany). ¹H/¹³C NMR spectra were recorded on a JEOL (Akishima, Japan) AL-400 (400/100 MHz) spectrometer and referenced to internal Si(CH₃)₄ or solvent signal. Elemental analyses were recorded on J-Science (Kyoto, Japan) JM-10 micro corder. UV-vis and fluorescence spectra were measured on a Jasco (Hachioji, Japan) V-660 spectrophotometer and Jasco (Hachioji, Japan) FP-6300 spectrofluorometer, respectively. Fluorescence quantum yields were measured on a HAMAMATSU photonics (Hamamatsu, Japan) C9920-02 absolute PL quantum yield measurement system. Fluorescence lifetimes were measured on a HORIBA fluorescence lifetime system Tempro equipped with a 370, 400, 460, or 490 nm bandpath filter. TriMeOBQDMEN [29] was prepared as described previously. CAUTION: Perchlorate salts of metal complexes with organic ligands are potentially explosive. All due precautions should be taken.

3.2. Synthesis of N,O-Bis(2-quinolylmethyl)-2-methylaminoethanol (N,O-BQMAE (1))

The mixture of 2-methylaminoethanol (37.6 mg, 0.501 mmol), 2-chloromethylquinoline (89.8 mg, 0.506 mmol), potassium carbonate (71.4 mg, 0.517 mmol), and potassium iodide (84.5 mg, 0.509 mmol) in dry acetonitrile (15 mL) was refluxed for 23 h. After the reaction mixture was cooled to room temperature, solvent was removed under reduced pressure, and organic materials were extracted with dichloromethane–water. The organic layer was dried and evaporated to afford N-(2-quinolylmethyl)-2-methylaminoethanol as yellow oil (106 mg, 0.490 mmol, 98%).

This material (200 mg, 0.925 mmol) was dissolved in THF (4 mL) and cooled on ice, then n-butyllithium (1.6 M solution in hexane, 869 µL, 1.39 mmol) was added slowly through a syringe. After stirring for 30 min on ice, 2-chloromethylquinoline (205 mg, 1.15 mmol) in THF (2 mL) was added slowly through a syringe. The reaction mixture was warmed to room temperature and refluxed for 2 h. After the reaction, solvent was removed under reduced pressure, and organic materials were extracted with dichloromethane–water. The organic layer was dried and evaporated, and the residue was purified by alumina column chromatography using chloroform containing 0.1% methanol (R_f = 0.25) as an eluent. The product was further purified by recycling GPC to afford N,O-bis(2-quinolylmethyl)-2-methylaminoethanol (N,O-BQMAE (1)) as yellow oil (21.3 mg, 0.0596 mmol, 6%).

¹H NMR (CDCl₃, 400 MHz): δ (ppm) 8.03–8.15 (m, 4H, quinoline-H4 and H8), 7.78–7.82 (m, 2H, quinoline-H5), 7.67–7.72 (m, 3H, quinoline-H3 and H6), 7.62 (d, J = 8.3 Hz, 1H, quinoline-H3), 7.49–7.54 (m, 2H, quinoline-H7), 4.84 (s, 2H, OCH₂Ar), 3.94 (s, 2H, NCH₂Ar), 3.77 (t, J = 5.6 Hz, 2H, OCH₂CH₂), 2.85 (t, J = 5.6 Hz, 2H, NCH₂CH₂), 2.41 (s, 3H, NCH₃).

¹³C NMR (CDCl₃, 100 MHz): δ (ppm) 160.2, 259.3, 147.6, 147.5, 136.7, 136.3, 129.5, 129.3, 129.0, 128.9, 127.6, 127.5, 126.2, 126.1, 121.1, 119.3, 74.6, 69.1, 64.8, 57.0, 43.3.

HRMS (ESI) m/z: [1 + Na]⁺ calcd. for C₂₃H₂₃N₃ONa 380.17388. Found 380.17026.

3.3. Synthesis of N,O-Bis(6-methoxy-2-quinolylmethyl)-2-methylaminoethanol (6-MeO-N,O-BQMAE (2))

The mixture of 2-methylaminoethanol (62.8 mg, 0.836 mmol), 6-methoxy-2-chloromethylquinoline (154 mg, 0.742 mmol), potassium carbonate (103 mg, 0.745 mmol), and potassium iodide (142 mg, 0.855 mmol) in dry acetonitrile (20 mL) was refluxed for 23 h. After the reaction mixture was cooled to room temperature, solvent was removed under reduced pressure, and organic materials were extracted with dichloromethane–water. The organic layer was dried and evaporated to afford N-(6-methoxy-2-quinolylmethyl)-2-methylaminoethanol as yellow oil in quantitative yield.

This material (183 mg, 0.742 mmol) was dissolved in THF (4 mL) and cooled on ice, then n-butyllithium (1.6 M solution in hexane, 798 µL, 1.28 mmol) was added slowly through a syringe. After stirring for 30 min on ice, 6-methoxy-2-chloromethylquinoline (177 mg, 0.852 mmol) in THF (2 mL) was added slowly through a syringe. The reaction mixture was warmed to room temperature and refluxed for 2 h. After the reaction, solvent was removed under reduced pressure, and organic materials were extracted with dichloromethane–water. The organic layer was dried and evaporated, and the residue was purified by alumina column chromatography using chloroform containing 0.3% methanol (R_f = 0.3) as an eluent. The product was further purified by recycling GPC to afford N,O-bis(6-methoxy-2-quinolylmethyl)-2-methylaminoethanol (6-MeO-N,O-BQMAE (2)) as brown oil (54.8 mg, 0.131 mmol, 18%).

¹H NMR (CDCl₃, 400 MHz): δ (ppm) 7.95–8.00 (m, 3H, quinoline-H4 and H8), 7.89 (d, J = 8.3 Hz, 1H, quinoline-H8), 7.29–7.37 (m, 4H, quinoline-H3 and H7), 7.04 (d, J = 2.9 Hz, 1H, quinoline-H5), 7.01 (d, J = 2.9 Hz, 1H, quinoline-H5), 4.68 (t, J = 7.3 Hz, 1H, OCH₂Ar), 4.0–4.5 (br., 1H, OCH₂Ar), 3.92 (s, 3H, OCH₃), 3.90 (s, 3H, OCH₃), 3.56–3.76 (m, 4H, NCH₂Ar and OCH₂CH₂), 2.91–2.98 (m, 1H, NCH₂CH₂), 2.61–2.65 (m, 1H, NCH₂CH₂), 2.36 (s, 3H, NCH₃).

¹³C NMR (CDCl₃, 100 MHz): δ (ppm) 158.1, 157.6, 157.3, 157.0, 143.8, 143.5, 134.9, 134.6, 130.8, 130.1, 128.1, 127.6, 122.8, 122.0, 121.9, 105.1, 105.1, 68.6, 58.7, 55.53, 55.48, 55.42, 38.5, 38.1.

HRMS (ESI) m/z: [2 + H]⁺ calcd. for C₂₅H₂₈N₃O₃ 418.21307. Found 418.22029.

3.4. Synthesis of N,O-Bis(5,6,7-trimethoxy-2-quinolylmethyl)-2-methylaminoethanol (TriMeO-N,O-BQMAE (3))

The mixture of 2-methylaminoethanol (59.0 mg, 0.786 mmol), 5,6,7-trimethoxy-2-chloromethylquinoline (199 mg, 0.743 mmol), potassium carbonate (105 mg, 0.760 mmol), and potassium iodide (125 mg, 0.753 mmol) in dry acetonitrile (20 mL) was refluxed for 23 h. After the reaction mixture was cooled to room temperature, solvent was removed under reduced pressure, and organic materials were extracted with dichloromethane–water. The organic layer was dried and evaporated to afford N-(5,6,7-trimethoxy-2-quinolylmethyl)-2-methylaminoethanol as yellow oil in quantitative yield.

This material (220 mg, 0.718 mmol) was dissolved in THF (4 mL) and cooled on ice, then n-butyllithium (1.6 M solution in hexane, 675 µL, 1.08 mmol) was added slowly through a syringe. After stirring for 30 min on ice, 5,6,7-trimethoxy-2-chloromethylquinoline (195 mg, 0.728 mmol) in THF (2 mL) was added slowly through a syringe. The reaction mixture was warmed to room temperature and refluxed for 2 h. After the reaction, solvent was removed under reduced pressure, and organic materials were extracted with dichloromethane–water. The organic layer was dried and evaporated, and the residue was purified by alumina column chromatography using chloroform containing 0.3% methanol (R_f = 0.4) as an eluent. The product was further purified by recycling GPC to afford N,O-bis(5,6,7-trimethoxy-2-quinolylmethyl)-2-methylaminoethanol (TriMeO-N,O-BQMAE (3)) as brown oil (73.2 mg, 0.136 mmol, 19%).

¹H NMR (CDCl₃, 400 MHz): δ (ppm) 8.34 (d, J = 8.8 Hz, 1H, quinoline-H4), 8.31 (d, J = 8.3 Hz, 1H, quinoline-H4), 7.53 (d, J = 8.8 Hz, 1H, quinoline-H3), 7.47 (d, J = 8.3 Hz, 1H, quinoline-H3), 7.24 (s, 1H, quinoline-H8), 7.21 (s, 1H, quinoline-H8), 4.78 (s, 2H, OCH₂Ar), 4.06 (s, 6H, OCH₃), 4.00 (s, 6H, OCH₃), 3.98 (s, 6H, OCH₃), 3.89 (s, 2H, NCH₂Ar), 3.77 (t, J = 5.6 Hz, 2H, OCH₂CH₂), 2.83 (t, J = 5.6 Hz, 2H, NCH₂CH₂), 2.40 (s, 3H, NCH₃).

¹³C NMR (CDCl₃, 100 MHz): δ (ppm) 158.4, 155.9, 155.7, 146.9, 145.3, 140.4, 130.9, 130.6, 118.8, 118.4, 118.1, 117.0, 103.9, 130.7, 74.5, 68.9, 64.7, 61.5, 61.2, 56.9, 56.0, 43.2.

Anal. Calcd for C₂₉H_36.4N₃O_7.7 (3·0.7H₂O): H, 6.67; C, 63.31; N, 7.64. Found: H, 6.36; C, 63.16; N, 7.36.

HRMS (ESI) m/z: [3 + H]⁺ calcd. for C₂₉H₃₆N₃O₇ 538.25532. Found 538.25830.

3.5. Synthesis of N,N-Bis(2-quinolylmethyl)-2-methoxyethylamine (N,N-BQMAE (4))

The mixture of 2-methoxyethylamine (22.4 mg, 0.298 mmol), 2-chloromethylquinoline (88.6 mg, 0.499 mmol), and potassium carbonate (69.1 mg, 0.500 mmol) in dry acetonitrile (10 mL) was refluxed for 21 h. After the reaction mixture was cooled to room temperature, solvent was removed under reduced pressure, and organic materials were extracted with dichloromethane–water. The organic layer was dried and evaporated to afford N,N-bis(2-quinolylmethyl)-2-methoxyethylamine (N,N-BQMAE (4)) as a yellow-white solid in quantitative yield.

¹H NMR (CDCl₃, 400 MHz): δ (ppm) 8.13 (d, J = 8.7 Hz, 2H, quinoline-H4), 8.04 (d, J = 8.3 Hz, 2H, quinoline-H8), 7.76–7.81 (m, 4H, quinoline-H3 and H5), 7.68 (dt, J = 1.0, 8.3 Hz, 2H, quinoline-H6), 7.50 (t, J = 7.4 Hz, 2H, quinoline-H7), 4.11 (s, 4H, NCH₂Ar), 3.56 (t, J = 5.9 Hz, 2H, OCH₂CH₂), 3.27 (s, 3H, OCH₃), 2.89 (t, J = 5.9 Hz, 2H, NCH₂CH₂).

¹³C NMR (CDCl₃, 100 MHz): δ (ppm) 160.5, 147.5, 136.2, 129.3, 128.9, 127.4, 127.3, 126.0, 121.0, 71.0, 61.8, 58.6, 53.7.

Anal. Calcd for C₂₃H_23.4N₃O_1.2 (4·0.2H₂O): H, 6.53; C, 76.51; N, 11.64. Found: H, 6.55; C, 76.56; N, 11.72.

HRMS (ESI) m/z: [4 + Na]⁺ calcd. for C₂₃H₂₃N₃ONa 380.17388. Found 380.18173.

3.6. Synthesis of N,N-Bis(6-methoxy-2-quinolylmethyl)-2-methoxyethylamine (6-MeO-N,N-BQMAE (5))

The mixture of 2-methoxyethylamine (38.3 mg, 0.510 mmol), 6-methoxy-2-chloromethylquinoline (215 mg, 1.04 mmol), and potassium carbonate (138 mg, 0.998 mmol) in dry acetonitrile (20 mL) was refluxed for 21 h. After the reaction mixture was cooled to room temperature, solvent was removed under reduced pressure, and organic materials were extracted with dichloromethane–water. The organic layer was dried and evaporated to afford N,N-bis(6-methoxy-2-quinolylmethyl)-2-methoxyethylamine (6-MeO-N,N-BQMAE (5)) as yellow oil (208 mg, 0.498 mmol, 96%).

¹H NMR (CDCl₃, 400 MHz): δ (ppm) 8.01 (d, J = 8.8 Hz, 2H, quinoline-H4), 7.94 (d, J = 9.3 Hz, 2H, quinoline-H8), 7.70 (d, J = 8.3 Hz, 2H, quinoline-H3), 7.33 (dd, J = 2.9, 9.2 Hz, 2H, quinoline-H7), 7.05 (d, J = 2.9 Hz, 2H, quinoline-H5), 4.05 (s, 4H, NCH₂Ar), 3.92 (s, 6H, OCH₃), 3.55 (t, J = 5.6 Hz, 2H, OCH₂CH₂), 3.26 (s, 3H, OCH₃), 2.87 (t, J = 5.6 Hz, 2H, NCH₂CH₂).

¹³C NMR (CDCl₃, 100 MHz): δ (ppm) 157.8, 157.4, 134.4, 135.1, 130.3, 128.2, 121.8, 121.4, 105.1, 70.9, 61.5, 58.6, 55.4, 53.6.

Anal. Calcd for C₂₅H₂₈N₃O_3.5 (5·0.5H₂O): H, 6.62; C, 70.40; N, 9.85. Found: H, 6.42; C, 70.15; N, 9.80.

HRMS (ESI) m/z: [5 + Na]⁺ calcd. for C₂₅H₂₇N₃O₃Na 440.19501. Found 440.20330.

3.7. Synthesis of N,N-Bis(5,6,7-trimethoxy-2-quinolylmethyl)-2-methoxyethylamine (TriMeO-N,N-BQMAE (6))

The mixture of 2-methoxyethylamine (27.3 mg, 0.363 mmol), 5,6,7-trimethoxy-2-chloromethylquinoline (201 mg, 0.751 mmol), and potassium carbonate (103 mg, 0.745 mmol) in dry acetonitrile (15 mL) was refluxed for 21 h. After the reaction mixture was cooled to room temperature, solvent was removed under reduced pressure, and organic materials were extracted with dichloromethane–water. The organic layer was dried and evaporated to afford N,N-bis(5,6,7-trimethoxy-2-quinolylmethyl)-2-methoxyethylamine (TriMeO-N,N-BQMAE (6)) as yellow oil in quantitative yield.

¹H NMR (CDCl₃, 400 MHz): δ (ppm) 8.31 (d, J = 8.3 Hz, 2H, quinoline-H4), 7.62 (d, J = 8.8 Hz, 2H, quinoline-H3), 7.20 (s, 2H, quinoline-H8), 4.04 (s, 10H, NCH₂Ar and OCH₃), 3.98 (s, 6H, OCH₃), 3.97 (s, 6H, OCH₃), 3.57 (t, J = 5.6 Hz, 2H, OCH₂CH₂), 3.29 (s, 3H, OCH₃), 2.87 (t, J = 5.6 Hz, 2H, NCH₂CH₂).

¹³C NMR (CDCl₃, 100 MHz): δ (ppm) 159.9, 155.7, 146.9, 145.2, 140.4, 130.5, 118.7, 118.1, 103.9, 71.0, 61.7, 61.5, 61.1, 58.7, 56.0, 53.7.

Anal. Calcd for C_29.3H_35.3Cl_0.9N₃O₇ (6·0.3CHCl₃): H, 6.20; C, 61.37; N, 7.33. Found: H, 6.05; C, 61.66; N, 7.34.

HRMS (ESI) m/z: [6 + Na]⁺ calcd. for C₂₉H₃₅N₃O₇Na 560.23727. Found 560.24627.

4. Conclusions

The newly designed six bisquinoline-based N3O1 tetradentate ligands with 2-aminoethanol backbone were examined as a fluorescent Cd²⁺ sensor. This ligand library includes N,O-BQMAE derivatives 1–3 and N,O-BQMAE derivatives 4–6, which differ in the position of two quinoline moieties and the number of methoxy substituents on the quinoline rings. Judging from the moderately good fluorescent Cd²⁺ selectivity over Zn²⁺ (I_Zn/I_Cd = ~30%) and wide range of metal binding affinity (two orders differences in K_d), the trimethoxy derivatives TriMeO-N,O-BQMAE (3) and TriMeO-N,N-BQMAE (6), as well as previously reported TriMeOBQDMEN, are regarded as a potential set of fluorescent ligands with different metal binding affinities useful for convenient quantification of Cd²⁺ in solution.