The Design and Synthesis of Fluorescent Coumarin Derivatives and Their Study for Cu²⁺ Sensing with an Application for Aqueous Soil Extracts

Abstract

A series of fluorescent coumarin derivatives 2a–e were systematically designed, synthesized and studied for their Cu²⁺ sensing performance in aqueous media. The sensitivities and selectivities of the on-to-off fluorescent Cu²⁺ sensing signal were in direct correlation with the relative arrangements of the heteroatoms within the coordinating moieties of these coumarins. Probes 2b and 2d exhibited Cu²⁺ concentration dependent and selective fluorescence quenching, with linear ranges of 0–80 μM and 0–10 μM, and limits of detection of 0.14 μM and 0.38 μM, respectively. Structural changes of 2b upon Cu²⁺ coordination were followed by fluorescence titration, attenuated total reflection Fourier transform infrared spectroscopy (ATR-FTIR), mass spectrometry, and single crystal X-ray diffraction on the isolated Cu²⁺-coumarin complex. The results revealed a 1:1 stoichiometry between 2b and Cu²⁺, and that the essential structural features for Cu²⁺-selective coordination are the coumarin C=O and a three-bond distance between the amide NH and heterocyclic N. Probe 2b was also used to determine copper (II) levels in aqueous soil extracts, with recovery rates over 80% when compared to the standard soil analysis method: inductively coupled plasma-mass spectrometry (ICP-MS).

1. Introduction

Heavy metal contamination [1] in soil and water is a major concern due to its direct effect on the quality of drinking water, safe food production [2,3], and the detrimental impact on human health when present in the body in excessive amounts [4,5]. However, heavy metals are also important micronutrients for living organisms, thus are desired ingredients in soil and water [6]. Copper is known as one of the most abundant essential trace elements in human body [7,8], and is also required by many living organisms to maintain healthy physiological processes due to its redox-active nature [9]. The World Health Organization (WHO) sets the maximum allowable level of Cu²⁺ in drinking water at 30 μM and recommends that the population mean intake should not exceed 10–12 mg per day for adults [10]. Thus, the development of methods allowing for sensitive, selective, and real-time detection of bioavailable heavy metals, including Cu²⁺, in environmental samples is ever relevant.

Some of the most readily utilized sensing methods are based on molecular probes that display concentration dependent changes in their optical properties arising from the structural changes upon interaction with the analytes of interest [11]. Coumarins are attractive for use as fluorescent optical probes, due to their easy-to-tailor structure that allows for metal ion-selective coordination sites, as well as their exhibiting significant changes in their optical signals upon interaction with such target analytes [12]. When interacting with Cu²⁺, coumarins can undergo ligand to metal charge transfer, forming a Cu²⁺–coumarin complex. This can inhibit the intramolecular charge transfer (ICT) – the mechanism responsible for its fluorescence – within the coumarin. Thus, complex formation can result in the quenching of the coumarin’s fluorescence [13,14,15,16]. A broad range of coumarin-based fluorescent Cu²⁺-sensors have been reported [17]. For example, a ratiometric coumarin-hydrazone sensor exhibited Cu²⁺ selectivity in the presence of other transition metal ions, and its fluorescence emission at 574 nm was quenched by increasing amounts of Cu²⁺ present [18]. A reversible ratiometric Cu²⁺-sensor incorporating two covalently linked fluorophores (coumarin and 4-amino-7-sulfamoyl benzoxadiazole) displayed Cu²⁺-induced blue shift of its emission (from 555 to 460 nm) upon coordination of Cu²⁺ due to decreased intramolecular fluorescence resonance energy transfer (FRET) [19]. Despite the ever-increasing number of optical molecular probes reported for Cu²⁺ detection, improvements in selectivity and sensitivity, and a better understanding of the structural requirements when designing these Cu²⁺-coordinating molecular probes, remain of interest.

This work was aimed at the design, synthesis, and simultaneous evaluation of a series of structurally related coumarin derivatives to further understand the structural requirements towards optical Cu²⁺-sensors that have enhanced selectivity and sensitivity, especially when employed in mixed aqueous media. Previously reported work on picolyl-substituted coumarin-3-carboxylic acid derivatives 1a–c (Scheme 1a) [20] indicated that the distance and relative orientation of the pyridine nitrogen relative to the other coordinating moieties (the coumarin C=O and the amide NH) were essential features affecting the ability to coordinate Cu²⁺. Coumarin 1a displayed Cu²⁺-selective fluorescence quenching, as opposed to 1b–c which failed to do so. Herein, we designed a series of coumarin derivatives 2a–e (Scheme 1b) incorporating various spacers between the amide NH and a pyridine N (two-bond in 2a, three-bond in 2b, and four-bond in 2c), or benzoimidazole N in 2d, or ethylenediamine -NH₂ in 2e. Thus, information on the effect of (i) the different spacers between the Cu²⁺-coordinating heteroatoms in 2a–c, and (ii) the distribution and availability of the electrons on the nitrogen oriented three-bonds away from the amide NH (2b, 2d, and 2e) on the coordination of Cu²⁺ can be gained. The coumarin derivatives which showed Cu²⁺-dependent fluorescence quenching were further studied by attenuated total reflection Fourier transform infrared spectroscopy (ATR-FTIR), mass spectrometry, and single crystal X-ray diffraction to indicate changes in their structure before and after Cu²⁺-coordination. From the results obtained, candidate 2b was evaluated on aqueous extracts of real soil samples collected from the roadside in the western parts of Melbourne, Victoria, Australia; the results were compared to the standard, inductively coupled plasma mass spectrometry (ICP-MS) method.

2. Results and Discussion

2.1. Design and Synthesis of Coumarin Derivatives 2a–e

In order to gain further information on the ideal relative atomic arrangement of the Cu²⁺-selective coordinating moiety within substituted carboxylic coumarins; a series of alkylamino-pyridines 3a–c, a benzimidazole 3d, and a linear chain mono-protected diaminoethylene 3e were selected and subjected to amide bond formation with coumarin-3-carboxylic acid using PyBOP as a coupling agent and triethylamine as a non-nucleophilic base (Scheme 2) [21]. Although the yields of 2a–e were moderate (

Table 1. Selected amines 3a–e, synthesized coumarin derivatives 2a–e, and their corresponding yields (%), absorption maxima, excitation and emission maxima, and Stokes shifts.

Probe	2a	2b	2c	2d	2e*
Yield %	26	23	25	26	30^
λabs (nm)	298	300	300	306	300
λex (nm)	280	303	303	325	303
λem (nm)	406	412	412	436	412
Δν (nm)	126	109	109	111	109

^*2e was obtained after removal of the protecting group of Boc-2e using TFA in DCM to give 2e (see Section 3.2.5); ^{^} overall yield.

) (presumably due to the potential side reactions resulting in undesired by-products) [22], the choice of this coupling agent allowed for a facile, one-step purification process upon completion of the reaction. Due to the formation of water-soluble by-products, the desired products 2a–d and Boc-2e were isolated after a simple extraction to a purity suitable for characterisation and analytical studies. Boc-2e was then subjected to deprotection using TFA, and then the TFA salt of 2e was neutralised with NaOH to obtain 2e. The resulting coumarins 2a–e were characterised by ¹H and ¹³C-NMR, IR, and high-resolution mass spectrometry (Figures S1–S20). Absorption, and fluorescence excitation and emission spectra for 2a–e were recorded and the corresponding maxima are listed in Table 1. The Stokes shifts (∆ = λ_em − λ_ex) were greater than 100 nm for each of these coumarin derivatives 2a–e. This is a desired property for optical molecular probes, facilitating enhanced sensitivity due to the lesser overlap between the selected wavelengths of the exciting light-source and the emission filter in the detector recording the sensing signal [23].

2.2. The Effects of pH and Concentration On The Fluorescence of 2a–e

In order to find the optimum conditions to evaluate the Cu²⁺-sensing performance, fluorescence intensities of 2a–e were recorded across the pH range of 3–13 (Figure 1a). The fluorescence of 2a–e between pH 3–11 showed no significant changes; however, above pH 11, sharp quenching was observed. Fluorescence intensities at various concentrations of dissolved coumarin 2a–e were also recorded within the range of 5–80 µM (Figure 1b). For the subsequent studies, 30 µM of dissolved 2a–e at pH 7 was chosen, as beyond that concentration, no further concentration dependent fluorescence was observed.

2.3. UV-Vis and Fluorescence Properties of 2a–e in the Presence of Cu²⁺

UV-Vis spectra of 2a–e in DMSO/HEPES buffer (v/v, 1/9) were recorded upon the addition of one equivalent of CuCl₂ (Figure S21). Coumarins 2a and 2c–e showed no changes in their absorbances, with only 2b showing a broadening of the absorption peak upon addition of CuCl₂. Recording of the fluorescence emission intensities of 2a–e in DMSO/HEPES buffer (v/v, 1/9) in the absence and presence of one equivalent of CuCl₂ (λ_ex according to Table 1) revealed that 2a, 2c, and 2e exhibited no optical response to Cu²⁺ (Figure 2). However, fluorescence quenching was observed for both 2b (λ_em = 412 nm) and 2d (λ_em = 436 nm), suggesting the coordination of Cu²⁺. In 2b, 2d, and 2e, the relative arrangement of the moieties that can donate electrons towards the copper are similar, yet despite this, only 2b and 2d exhibited optical responses upon interaction with the Cu²⁺. The lack of fluorescence quenching in 2e is proposedly due to the greater flexibility of the linear chain amino group that does not facilitate the formation of a metal-coumarin complex. Based on these results, a three-bond spacing and restricted flexibility between the heterocyclic nitrogen and the amide NH are deemed to be essential for the Cu²⁺-based fluorescence quenching, and consequently, 2b and 2d were further studied for their Cu²⁺-sensing performance is aqueous media.

Firstly, fluorescence titrations of 2b and 2d (30 µM) with CuCl₂ (0–80 µM) in DMSO/HEPES buffer were carried out (Figure 3) to identify the linear range and the limit of detection (LOD). Both 2b and 2d responded with gradually quenched fluorescence intensities at their emission maxima, 412 nm and 436 nm, respectively, during the addition of increasing amounts of CuCl₂. The fluorescence quenching of 2b by Cu²⁺ was linear (R² = 0.995) within the Cu²⁺ concentration range of 0–80 µM, which is a significant improvement from the previously reported probe 1a, whose response was linear only within the range of 0–25 μM [20]. The calculated limit of detection (LOD) was 0.14 µM (LOD = 3σ/slope, where σ is standard deviation of the blank measurements) [24]. Both the linear range and the LOD were satisfactory to fulfil the sensing requirements for monitoring copper levels in drinking water, as the maximum allowed level is specified at 30 µM by the Australian Drinking Water Guidelines [25]. Meanwhile, 2d (Figure 3c–d) exhibited a rather narrow linear range of 0–10 µM, and a calculated association constant of 7.83 × 10⁵ M⁻¹ (versus that of 2b at 6.85 × 10⁴ M⁻¹) [26], and a LOD of 0.38 µM. Furthermore, a shift of the emission maxima from 436 nm towards 405 nm was observed in the presence of Cu²⁺ at concentrations >10 µM. Thus, 2d may only be applicable for Cu²⁺-sensing within the range of 0.4–10 µM. Also, the greater association constant shows higher affinity of 2d to coordinate Cu²⁺, which is suggested to be indicative to the greater availability of the heterocyclic N lone pair in 2d [27], hence the faster coordination of Cu²⁺.

For both 2b and 2d, the fluorescence quenching mechanism can be explained by the previously reported ligand to metal charge transfer [28]. The fluorescence quenching results herein showed that a decrease of the emission intensity occurs beyond the addition of one equivalent (30 μM) of Cu²⁺ to both 2b and 2d (Figure 3a,c). This phenomenon was previously studied via femtosecond time-resolved fluorescence, showing evidence that the charge transfer only contributes to 29% of the overall fluorescence quenching [20].

To further assess the time-frame of developing coordination and charge transfer between Cu²⁺ and the probes, the changes in the fluorescence intensity as functions of time were recorded for 2b and 2d. After the addition of two equivalents of Cu²⁺ (ensuring the full quenching of fluorescence), and a subsequent mixing step taking 3.5 sec, the fluorescence intensity values were recorded at 25 ℃ (Figure 4) and fitted with modified first order kinetics according to (Equation (1)) [29] to identify the time points where the fluorescence was quenched by the Cu²⁺:

$$\ln \left(\mathrm{I}\right)=\mathrm{k}\times \mathrm{t}+\mathrm{b}$$

(1)

where I (a.u.) was fluorescence intensity of the coumarins, t (s) was time, and k (s⁻¹) was the rate constant of the first-order model. The linear range of ln(I) versus time (s) function for 2b and 2d is shown in Figure 4, with R² values of 0.99 and 0.98, respectively. To determine the time frame within which the fluorescent intensities of 2b and 2d were quenched by the copper, firstly, the average quenched fluorescence intensity was identified, as shown in Figure S22. Then, based on the linear fitted equations (Figure 4), the quenching time points were calculated to be 7.5 sec and 6.9 sec for 2b and 2d, respectively. Those results show the faster quenching of 2d compared to 2b, suggesting greater availability of the heterocyclic N lone pair in 2d [27] resulting in faster coordination.

One of the most indispensable qualities of a fluorescent probe is its selectivity towards a target analyte over other commonly encountered species. To evaluate the selectivity of 2b and 2d (30 μM) towards Cu²⁺ (30 μM), a variety of metal ions (30 μM) including Al³⁺, Ca²⁺, Cd²⁺, Co²⁺, Fe³⁺, Mn²⁺, Pb²⁺, Hg²⁺, Ni²⁺, Cu⁺, Mg²⁺, Li⁺, and Zn²⁺ were pre-mixed with the probes and fluorescence emissions were examined with and without the addition of Cu²⁺. In the presence of one equivalent of the potentially interfering cations, both 2b and 2d exhibited selectivity towards Cu²⁺ (Figure S23). When considering real life assays, the copper may be present at a significantly lower concentration compared to other metal ions; therefore, fluorescence quenching in the presence of excess amounts (20 equivalents) of interfering metal ions was also investigated for 2b and 2d (Figure 5). Fluorescence intensity prior the addition of the target analyte Cu²⁺ was quenched by more than 10% by only Fe³⁺ and Cd²⁺ for 2b (Table S1), and Co²⁺, Fe³⁺, Ni²⁺, and Hg²⁺ for 2d (Table S2). The inferior selectivity observed in the case of 2d suggests that the heterocyclic N and the availability of its lone pair have an essential role in the coordination of the metal ion species.

Another desirable property of a sensing system is the potential for repeated recovery of the sensor, which results in an extended life-cycle and applicability. Cu²⁺ is known for its high affinity towards chelating agents, such as L-cysteine [30], EDTA [31], and dopamine [32]. In this work, EDTA was used to record fluorescence recovery of both 2b and 2d. Upon the alternating addition of 1 eq. of Cu²⁺, then EDTA, the fluorescence levels of 2b was recovered up to approximately 80% of its original intensity through four recovery cycles (Figure 6). These results indicated that 2b could be employed in reusable optical sensing devices for Cu²⁺ detection.

A summary of the sensing properties for 2b, 2d, and the previously published probe 1a, can be found in

Table 2. A comparison of sensing properties among 2b, 2d, and 1a.

	2b	2d	1a*
Limit of detection (μM)	0.14	0.38	0.5
Linear range (μM)	0–80	0–10	0–50
Interference	Fe3+, Cd2+	Co2+, Fe3+, Ni+, Hg2+	No data included
Association constant (M−1)	6.85 × 104	7.83 × 105	1.17 × 105

* Values taken from literature [20].

. It is apparent that 2b shows improvements both in its LOD and linear range for Cu²⁺-sensing in DMSO/HEPES buffer.

2.4. The Coordination of Cu²⁺ by Coumarin 2b

The fluorescence titration of coumarins 2a–e showed that 2b and 2d had affinities to coordinate Cu²⁺, with 2b having a suitable linear range and selectivity for practical application. Hence, the interaction between 2b and Cu²⁺ was further studied. Job plot, solid state ATR-FTIR, single crystal X-ray diffraction, and mass spectra were recorded on 2b and the isolated 2b–Cu²⁺ complex. The Job plot analysis, based on the fluorescence recorded by titrating 2b with Cu²⁺ (Figure 7), revealed a 1:1 stoichiometry between 2b and Cu²⁺. This corresponds with the mass spectra depicting a peak for 2b-Cu²⁺ complex at m/z 383.03 (Figure S24). (Note that the corresponding peak is due to the association of the complex with the solvent used, acetonitrile.) However, tridentate chelation [33] of Cu²⁺ by 2b in a 1:2 stoichiometry is apparent by a m/z 622.09 peak in the mass spectra (Figure S24); upon increasing Cu²⁺ concentration this form transitions into a 1:1 complex.

Infrared spectra were also recorded on the solids isolated from an acetonitrile solution of 2b in the absence and presence of added Cu²⁺ (Figure 8 and Figure S25). Notably, the solid samples isolated from acetonitrile were a mixture of 2b, and 2b–Cu²⁺. By overlapping the normalised IR spectra, the characteristic changes in the intensities of particular peaks have confirmed the previously reported observation, that the key Cu²⁺-coordinating moieties are the coumarin C=O, the amide NH, and the pyridine N (

Table 3. Characteristic ATR-FTIR peaks and their intensities (in brackets) for 2b before and after coordination of Cu²⁺.

	ν(C=O) Coumarin	ν(C=O) Amide	ν(C=N) Pyridine	ν(C=N) Pyridine	ν(N-H) Amide	ν(C-N) Amide	ν(C=C) Pyridine	ν(C=C) Pyridine
2b	1706 (0.74)	1648 (0.47)	1608 (0.47)	1601 (0.47)	1564 (0.62)	1514 (0.85)	1480 (0.44)	1451 (0.52)
+0.5 eq. Cu2+	1707 (0.30)	1653 (0.75)	1609 (0.59)	1595 (0.54)	1566 (0.49)	1535 (0.36)	1483 (0.34)	1451 (0.40)
+1eq. Cu2+	1707 (0.15)	1653 (0.71)	1610 (0.69)	1595 (0.51)	1567 (0.46)	1535 (0.57)	1483 (0.36)	1451 (0.39)

).

Firstly, the defined peak at 3301 cm⁻¹ corresponding to the NH stretch in 2b became broader and less intense upon the coordination of Cu²⁺ (Figure S25). The most apparent changes occurred within the region of 1680–1750 cm⁻¹ originating from alterations of the coumarin C=O bond within 2b. Upon coordinating the Cu²⁺, the coumarin C=O peak at 1706 cm⁻¹ became significantly less intense, suggesting the involvement of the oxygen lone pair electrons in the coordinative bond. Additionally, the amide C=O stretch shifted from 1648 in 2b to 1653 cm⁻¹ in 2b-Cu²⁺ along with increased peak intensity. That suggests a lesser role of the carbonyl moiety in the potential amide resonance structures, hence the restoration of the amide C=O double bond, resulting in a stronger dipole, thus absorption at higher wavenumbers. Further changes were observed in the peaks presenting at about 1608 cm⁻¹, depicted as C=N stretches of the pyridine ring. Both the amide NH and C–N peaks at 1564 cm⁻¹ decreasing in intensity, and the transition of the 1514 cm⁻¹ peak of free ligand 2b into 1550 cm⁻¹ was observed (see shoulder of 2b-Cu²⁺) upon coordination with the Cu²⁺. Comprehensive loss of those peaks was not expected for two reasons: firstly, the studied solids were a mixture (as discussed above); secondly, the peak at 1564 cm⁻¹ originates from both the vibration of amide NH and the stretch of amide C–N bonds. Furthermore, upon addition of Cu²⁺, the peak at 1451 cm⁻¹ was reduced in intensity which suggested changes of the possible resonance structures in the pyridine ring.

Ultimately, single crystal X-ray diffraction was recorded on the crystal grown from an aqueous acetonitrile solution containing 2b and CuCl₂ in a 1:1 molar ratio (Figure 9, Figure S26, Tables S1 and S2), as well as single crystal of 2b itself. The as-observed structure of 2b-Cu²⁺ complex is depicting a 1:1 stoichiometric ratio, and that the coumarin O(3), the amide N(2), the pyridine N(1), and a residual chlorine Cl(1) are the coordinating moieties [20]. This four-bond coordination was previously explained by the sp³ hybridization of Cu²⁺ when dissolved in acetonitrile water mixture [34]. According to the structure exhibited in the X-ray diffraction, the hydrogen of amide N(1) is displaced during the coordination process. Moreover, the bond length of coumarin C=O increased from 1.21 Å to 1.35 Å upon Cu²⁺ coordination, which falls between the conjugated and single bond length [35,36] (

Table 4. Experimental and literature references of bond lengths (Å).

	In 2b	In 2b-Cu2+	Single bond length
Coumarin C=O	1.21	1.35	1.43

). This correlates with the IR results, translating into the diminishing C=O IR peak at 1707 cm⁻¹ upon increasing Cu²⁺ concentration (Table 3).

2.5. The Evaluation of 2b on Soil Samples

The evaluation of 2b for Cu²⁺-sensing in soil was conducted using a water extraction [37] of roadside soil samples collected in the western parts of Melbourne, Victoria, Australia. Fluorescence titrations generated using 2b were compared against the industry standard of ICP-MS quantification [38]. ICP-MS quantifies elemental total copper content, while use of 2b measures dissolved Cu²⁺; comparison is still feasible, as it is accepted within the field that following sample preparation, most copper species will be present as Cu²⁺ [37,39,40]. The standard curve of 2b was recorded (Figure S27) and background fluorescence of soil extracts was also tested (Figure S28), allowing for correction when calculating the Cu²⁺ concentration by 2b. Recovery of Cu²⁺, directly from the water extraction without further sample manipulation was >80% for all three soil samples tested (

Table 5. Determination and recovery values of Cu²⁺ in soil samples using both ICP-MS (μM) and 2b (μM).

Sample	Cu2+ Concentration by ICP-MS (μM)	Fluorescence of Solution of 2b at λex = 412 nm (a.u.)	Cu2+ Concentration by Using 2b (μM)	Recovery
1	1.23 ± 0.13	173.54 ± 0.69	1.03 ± 0.13	83.7%
2	1.36 ± 0.21	175.11 ± 0.36	1.27 ± 0.05	93.4%
3	1.71 ± 0.27	178.48 ± 0.69	1.75 ± 0.09	102.3%

).

3. Materials and Methods

3.1. Materials and Instruments

Benzotriazol-1-yloxy tripyrrolidinophosphonium hexafluoro-phosphate (PyBOP) was purchased from Oakwood Chemical (West Columbia, South Carolina, SC, USA). Coumarin-3-carboxylic acid (99%), 2-amino pyridine (99%), 2-(2ʹ-pyridyl)ethylamine (95%), and trifluoroacetic acid (99%) were purchased from Sigma-Aldrich (St. Louis, MO, USA). 2-(Aminomethyl)pyridine (99%) was obtained from Acros Organic (Belgium, WI, USA). 2-(Aminomethyl)benzimidazole dihydrochloride (98%) was purchased from Alfa Aesar (Ward Hill, USA). N-Boc-ethylenediamine (98%) was purchased from Combi-blocks (SanDiego, CA, USA). All chemicals were used as received without further purification. IR spectra were recorded on Nicolet 6700 FT-IR (Thermo Scientific, Waltham, MA, USA). UV-Vis and fluorescence spectra were measured on Varian Cary Eclipse 5G UV-Vis-NIR spectrometer and Varian Cary Eclipse fluorescence spectrophotometer (CMUSA), respectively. NMR were recorded on Bruker Ascend (Billerica, MA, USA) at 400 MHz for ¹³C and 100 MHz for ¹H, respectively. Mass spectra were obtained on Thermo Scientific Q Exactive spectrometer in high resolution electrospray mode. Single crystal X-ray diffraction was recorded on Bruker Axs Apex Duo Single crystal XRD instrument. ICP-MS was conducted on Agilent Technologies 7700x ICP-MS Analyser (Santa Clara, CA, USA). CLARIOstar plate reader was used for kinetics study.

3.2. General Synthetic Method for the Preparation of 2a–e

Coumarin derivatives 2a–e were synthesised by the adaptation of a previously reported procedure with an additional step of deprotection for Boc-2e [20]. Coumarin-3-carboxylic acid (1 eq.), a selected amine (1.1 eq.), and PyBOP (1 eq.) were stirred at room temperature in acetonitrile. Upon completion, the reaction was quenched with brine (30 mL/5 mmol of coumarin-3-carboxylic acid) and extraction was conducted with dichloromethane (30 mL/5 mmol of coumarin-3-carboxylic acid). The organic layer was then washed with 10% citric acid, then 10% NaHCO₃, following water and brine. The organic residues were dried over Na₂SO₄. Finally, the solvent was removed under vacuum. The residues were triturated with ethyl acetate to give the desired product as a solid.

3.2.1. Synthesis of 2-oxo-N-(pyridin-2-yl)-2H-chromene-3-carboxamide (2a)

Yield of 2a: 26%; white solid; FTIR (neat, cm⁻¹) C=O 1701, N–H 3261 and 1565, C=N 1670, C–N 1534; ¹H-NMR (400 MHz, DMSO-d₆, ppm): δ 7.20 (1H, t, J = 6.1 Hz, CH-4ʹ), 7.49 (1H, t, J = 7.6 Hz, CH-7), 7.57 (1H, d, J = 8.4 Hz, CH-8), 7.81 (1H, t, J = 7.8 Hz, CH-6), 7.89 (1H, t, J = 7.8 Hz, CH-5ʹ), 8.07 (1H, d, J = 7.8 Hz, CH-5), 8.27 (1H, d, J = 8.3 Hz, CH-3ʹ), 8.39 (1H, d, J = 4.9 Hz, CH-6ʹ), 9.05 (1H, s, CH-4), 11.17 (1H, s, NH). ¹³C-NMR (100 MHz, CDCl₃, ppm): δ 115.0, 116.9, 118.5, 118.7, 120.4, 125.5, 130.1, 134.7, 138.3, 148.5, 149.4, 151.3, 154.8, 159.9, 161.3; LRMS ASAP [2a + H⁺] 267.07; HRMS ASAP, [2a + H⁺] calculated m/z C₁₅H₁₁N₂O₃ 267.0725, found = 267.0764.

3.2.2. Synthesis of 2-oxo-N-(pyridin-2-ylmethyl)-2H-chromene-3 carboxamide (2b)

Yield of 2b 23%; white solid; FTIR (neat, cm⁻¹) C=O 1706, N–H 3301 and 1564, C=N 1648, C–N 1514; ¹H-NMR (400 MHz, DMSO-d₆, ppm): δ 4.65 (2H, d, J = 5.6 Hz, CH₂), 7.28 (1H, t, J = 6.2 Hz, CH-4ʹ), 7.39 (1H, d, J = 7.8 Hz, CH-3ʹ), 7.46 (1H, t, J = 7.5 Hz, CH-7), 7.53 (1H, d, J = 8.4 Hz, CH-8), 7.77 (2H, m, CH-5ʹ and 6), 8.01 (1H, d, J = 7.2 Hz, CH-5), 8.55 (1H, d, J=4.9 Hz, CH-6ʹ), 8.91 (1H, s, CH-4), 9.44 (1H, t, J = 5.5 Hz, NH); ¹³C-NMR (100 MHz, CDCl₃, ppm): δ 45.5, 116.7, 118.5, 118.7, 121.8, 122.4, 125.3, 129.9, 134.1, 136.8, 148.5, 149.5, 154.6, 156.8, 161.3, 161.8. LRMS ASAP [2b + H⁺] 281.09; HRMS ASAP [2b]⁻ calculated m/z C₁₆H₁₂N₂O₃ 280.0881, found = 280.0847.

3.2.3. Synthesis of 2-oxo-N-(2-(pyridin-2-yl) ethyl)-2H-chromene-3-carboxamide (2c)

Yield of 2c: 25%; white solid; FTIR (neat, cm⁻¹): C=O 1706, N–H 3299 and 1566, C=N 1654, C–N 1516; ¹H-NMR (500 MHz, DMSO-d₆, ppm): δ 3.00 (2H, t, J = 7.0 Hz, CH₂-7ʹ), 3.74 (2H, q, J = 6.6 Hz, CH₂-8ʹ), 7.24 (1H, t, J = 4.8 Hz, CH-4ʹ), 7.31 (1H, d, J = 7.8 Hz, CH-3′), 7.43 (1H, t, J = 6.0 Hz, CH-7),7.51 (1H, d, J = 8.4 Hz, CH-8), 7.70–7.76 (2H, m, 5ʹ and 6), 7.98 (1H, d, J = 6.2 Hz, CH-5), 8.52 (1H, d, J = 3.8 Hz, CH-6ʹ), 8.87 (1H, s, CH-4), 8.92 (1H, t, J = 5.6 Hz, NH); ¹³C-NMR (100 MHz, DMSO-d₆, ppm): δ 36.9, 38.8, 116.2, 118.5, 118.9, 121.7, 123.3, 125.2, 130.3, 134.1, 136.6, 147.6, 149.2, 153.9, 159.0, 160.4, 161.0. LRMS ASAP [2c + H⁺] 295.10; HRMS ASAP [2c + H⁺] calculated m/z C₁₇H₁₅N₂O₃ 295.1038, found = 295.1079.

3.2.4. Synthesis of N-((1H-benzo[d]imidazol-2-yl) methyl)-2-oxo-2H-chromene-3-carboxamide (2d)

Yield of 2d: 26%; orange solid; FT-IR (neat, cm⁻¹): C=O 1691, N–H 3395 and 1566, C=N 1653, C–N 1544; ¹H-NMR (400 MHz, DMSO-d₆, ppm): δ 4.94 (2H, s, CH₂), 6.63 (1H, t, J = 7.6 Hz, CH_{benzoimidazol}) 6.80 (1H, d, J = 8.0 Hz, CH_{benzoimidazol}), 6.96 (1H, t, J = 7.4 Hz, CH_{benzoimidazol}), 7.46–7.57 (3H, m, CH-8, 7 and benzoimidazol), 7.78 (1H, t, J = 7.9 Hz, CH-6), 8.03 (1H, d, J = 7.7 Hz, CH-5), 8.95 (1H, s, CH-4), 10.10 (s, 1H, NH_amide); ¹³C (100 MHz, DMSO-d₆, ppm): δ 52.6, 116.3, 116.3, 116.5, 117.0, 118.6, 119.9, 123.4, 124.5, 125.4, 126.3, 130.5, 134.4, 141.5, 147.4, 154.7, 159.9, 163.3; LRMS ASAP [2d + H₂O + H⁺] 338.34; HRMS ASAP [2d]⁻ calculated m/z C₁₈H₁₃N₃O₃ 319.0990, found = 319.0955.

3.2.5. Synthesis of N-(2-aminoethyl)-2-oxo-2H-chromene-3-carboxamide (2e)

The crude product received by the reaction of coumarin-3-carboxylic acid and 3e following the general procedure was purified by column chromatography on silica gel using 50% ethyl acetate in hexane to give desired product Boc-2e (yield 46%). Boc-2e (0.50 g, 1.5 mmol) was treated with trifluoroacetic acid (TFA, 5 mL) over an ice bath for 2 h. The volatiles were removed under vacuum, and the residues were taken up in water (10 mL). The mixture was neutralized with 1 M aqueous NaOH and extracted with DCM (50 mL), twice. The organic residues were washed with a saturated solution of Na₂CO₃ (aq) (50 mL), then brine (50 mL), and afterwards, dried over Na₂SO₄, filtered, and concentrated under reduced pressure to give 2e (0.23 g, 1.0 mmol, 66%) as a white solid. Overall yield of 2e: 30%; white solid; FT-IR (neat, cm⁻¹): C=O 1678, N–H 3318 and 1566, C–N 1512; ¹H-NMR (400 MHz, DMSO-d₆, ppm): δ 2.98 (2H, t, J = 6.1 Hz, CH₂-2ʹ), 3.55 (2H, t, J = 6.1 Hz, CH₂-1ʹ) 7.46 (1H, t, J = 7.6 Hz, CH-6), 7.54 (1H, d, J = 8.4 Hz, CH-8), 7.74-7.89 (3H, m, CH-7 and NH₂), 8.01 (1H, d, J = 7.8 Hz, CH-5), 8.89 (1H, s, CH-4), 8.93 (1H, t, J = 6.0 Hz, NH). ¹³C-NMR (100 MHz, DMSO-d₆, ppm): δ 37.2, 38.5, 116.3, 118.5, 118.9, 125.4, 130.5, 134.4, 147.8, 154.0, 160.2, 162.1; LRMS ASAP [2e + H⁺] 233.09; HRMS ASAP [2e]⁻ calculated m/z C₁₂H₁₂N₂O₃ 232.0880, found = 232.0846.

3.2.6. Synthesis of 2b-Cu²⁺ Complex for IR Study

Coumarin 2b was dissolved in acetonitrile, and then an aqueous solution of 0.5 or 1 equivalent of CuCl₂ was added and left to stir until fully dissolved. The volatiles were reduced under vacuum to give the desired Cu²⁺ complexes used without further purification to record IR and mass spectra.

3.2.7. The preparation of a Single Crystal of 2b and 2b–Cu²⁺ Complex

2b was dissolved in acetonitrile first and the crystal was grown under vapour diffusion into EtOH. One equivalent of 2b and CuCl₂ were dissolved in a solution of acetonitrile and water. The single crystal was grown under vapour diffusion into EtOH.

3.3. Optical Properties of 2a–e

CuCl₂ stock solution (1.00 mM) was prepared in aqueous solution of DMSO/HEPES buffer (20 mM, pH = 7) (v/v, 1/9), and stock solutions of 2a–e (0.9 mM) were prepared in DMSO/HEPES buffer (20 mM, pH = 7) (v/v, 1/9).

3.3.1. Fluorescence Intensity of 2a–e

Fluorescence intensity of 2a–e was recorded at pH 7 at concentrations in the range of 5–80 μM, diluted from the stock solution using DMSO/HEPES buffer on Varian Cary Eclipse spectrophotometer, with excitation and emission slit widths of 5 nm at their corresponding excitation maxima.

3.3.2. pH Dependence of Fluorescence Intensity

Fluorescence intensities of 2a–e (30 μM) with slit widths of 5 nm were recorded across the pH range of 3–13 at their corresponding excitation maxima.

3.3.3. The Optical Properties of 2a–e in the Presence of 1 eq. of Cu²⁺

UV-Vis and fluorescence spectra were recorded for 2a–e (30 μM) and 2a–e (30 μM) in the presence of CuCl₂ (30 μM) in DMSO/HEPES buffer (20 mM, pH = 7) in the 250–600 nm region for the UV-Vis and with slit widths of 5 nm for fluorescence studies.

3.3.4. Fluorescence Spectra of 2b and 2d in the Presence of Cu²⁺ at Various Concentrations

Fluorescence spectra of 2b and 2d (30 μM) were recorded in the presence of CuCl₂ (0–80 μM) in DMSO/HEPES buffer (20 mM, pH = 7) with slit widths of 5 nm at their corresponding excitation maxima, respectively. The kinetics study was conducted using 100 μL of 60 μM solutions of compound 2b and 2d in DMSO/HEPES buffer (20 mM, pH = 7, v/v, 1/9) with the automatic pipetting of 100 μL of 120 μM Cu²⁺ solution to form the complexes of 2b–Cu²⁺ and 2d–Cu²⁺ at 25 °C. After 3.5 s shaking time, the changes in time of the fluorescence for 2b–Cu²⁺ and 2d–Cu²⁺ complexes were recorded on CLARIOstar plate reader.

3.3.5. Interference Studies

The fluorescence intensities of 2b and 2d (30 μM) were recorded in the presence of following metal chlorides: Al³⁺, Ca²⁺, Cd²⁺, Co²⁺, Fe³⁺, Mn²⁺, Pb²⁺, Hg²⁺, Ni²⁺, Fe²⁺, Cu⁺, Mg²⁺, and Zn²⁺ (20 eq., 600 μM) in DMSO/HEPES buffer (20 mM, pH = 7), both in the absence and presence of CuCl₂ (30 μM). Firstly, an aliquot of selected potentially interfering metal ions (60 μM, 1 mL) and 2b or 2d (60 μM, 1 mL) was mixed together to reach a final concentration of 30 μM. The fluorescence of this solution was recorded. Then, a stock solution of CuCl₂ (60 μL) was added to those, and optical properties were again measured. Fluorescence spectra were recorded with slit widths of 5 nm at their corresponding excitation maxima.

3.3.6. Fluorescence Signal Recovery Study

The fluorescence intensity of 2b (30 μM) was recorded in DMSO/HEPES buffer (20 mM, pH = 7) upon the alternating addition of 1 equivalent of CuCl₂, followed by 1 equivalent of ethylene diamine tetraacetic acid (EDTA) in repeat cycles.

3.3.7. Cu²⁺ Sensing in Aqueous Soil Extracts

Soil samples were collected from the roadside in the western parts of Melbourne, Victoria, Australia. Soil samples (2.5 g) were dispersed in water (25 mL) and shaken for one week to extract water-soluble copper. The extracts were filtered through a 0.45 mm pore-size nylon filter membrane. Each extract was analysed for their copper content via ICP-MS [37] and by recording the fluorescence quenching using 60 µM solution of 2b (1 mL in DMSO/HEPES, v/v, 2:8) mixed with the soil extract (and 1 mL of water) for the quantitation of Cu²⁺ in soil using a pre-recorded standard curve.

4. Conclusions

A series of five coumarin derivatives were systematically designed, synthesized, and evaluated for their potential use for Cu²⁺-sensing in aqueous media. Coumarins 2b and 2d exhibited selectivity towards Cu²⁺ in the presence of other multivalent metal ions. The linear ranges of the sensing signal versus Cu²⁺ concentration were 0–80 μM for 2b, and 0–10 μM for 2d, falling into the relevant and desired concentration range for environmental sensing techniques. Limits of detection in aqueous media were 0.14 μM and 0.38 μM for 2b and 2d, respectively, satisfactory for the potential application of these molecular optical probes. Probe 2b exhibited superior applicable sensing performance towards Cu²⁺ when compared to the other coumarins presented in this work, hence it was further studied to gain understanding on its interaction with Cu²⁺. Solid state ATR-FTIR, single crystal X-ray diffraction, and mass spectrometry, revealed a 1:1 stoichiometric complex as the responsible species for the displayed quenching of the fluorescence of 2b. The formation of this coordination complex resulted in a Cu²⁺-concentration dependent sensing signal, while offered good recovery of the probe signal based on the fluorescence recovery study using EDTA. As a proof-of-concept for its applicability, 2b provided results in the Cu²⁺ quantitation of soil extracts which were comparable with the standard ICP-MS method. Furthermore, by assessing structurally related coumarins 2a–e (owning slightly different Cu²⁺-coordinating moieties), we gained valuable information on the required size of the coordination-site, relative spatial and atomic arrangement, and electron distribution of the coordinating heteroatoms. The fluorescence responses of 2a–c suggested that the size of the coordination site has a significant impact on the Cu²⁺ coordination. Finally, the fluorescence titration of 2b and 2d suggested that the lone pair’s availability of the heteroatoms at the same relative position also affects the sensing properties, both in coordination speed and selectivity. These results will be used in our future work for the systematic design and development of metal-ion coordinating optical sensing probes, owning enhanced selectivity and sensitivity.