Reactive Blue 4 (RB4), an anthraquinone dye (1-amino-4-[3-(4,6-dichlorotriazin-2-ylamino)-4-sulfophenylamino]anthraquinone-2-sulfonic acid), was obtained from Merck (Darmstadt, Germany). Demineralized water was used to prepare the solutions. Cu(NO₃)₂, KF, KCl, K₂CO₃, Na₂HPO₄, K₃PO₄, Na₂SO₄, KNO₃, NaC₂H₃O₂, KClO₄, K₂C₂O₄ and ascorbic acid (C₆H₈O₆) were purchased from Merck. Stock solution (10^–2 mol/L) of analytes was prepared by direct dissolution of their proper amount in deionized water. All other chemicals were of analytical grade and used as received. The buffer solution was prepared using 4-(2-hydroxyethyl)-1-piperazine ethane sulfonic acid (HEPES).

A Shimadzu 1601 PC UV-Vis double beam spectrometer (Kyoto, Japan) was used to record all UV-Vis spectra in quartz cuvettes 10.0 mm in diameter. A Bruker Vector 22 Fourier transform infrared spectrometer (Billerica, MA, USA) recorded the FTIR spectra using KBr pellet method. A digital Jenway 3510 digital pH meter (London, UK) calibrated with two standard buffer solutions was used to measure various pH values. A Hamilton syringe of 50 µL was used to deliver desired amounts of analyte solution into the cuvette.

The adjusted temperature in all titration experiments was 298.2 K. The 5·10^–5 mol/L solution of RB4 was prepared by the dilution of its 10^–2 mol/L stock solution in aqueous medium (10 mmol/L HEPES buffer solution, pH=7). Then the UV-Vis absorption spectra of RB4 were recorded by transferring 2.5 mL of this diluted solution to the quartz cuvette. For complexation study, a 10^–2 mol/L aqueous solution of the nitrate salt of Cu²⁺ was prepared. The detection of oxalate and its effect on the colour and absorption spectra of the RB4-Cu²⁺ solution were studied by preparing the oxalate solution and gradually adding it (1.76–96.0 µmol/L) with microlitre syringe to the solution containing RB4 complex with Cu²⁺. Then, the UV-Vis spectra were recorded. Subsequently, the interactions and the stoichiometry ratios (Yoe and Jones method (27) or the mole ratio method (28, 29)) were studied.

Test papers were fabricated by immersing ordinary filter papers into the aqueous solution of RB4 (0.1 mol/L) and Cu²⁺ (0.1 mol/L), respectively, dried at room temperature and then used for the determination of oxalate in aqueous solutions. The corresponding colour change was observed by the naked eye.

A simple oxalate colorimetric chemosensor was designed based on indicator displacement assay. For IDA, the indicator must bind to host reversibly and cause the change in spectroscopic signal. In this experiment, RB4 is a colorimetric indicator to form the RB4-Cu²⁺ complex. The interaction between RB4 and Cu²⁺ was investigated by UV-Vis spectroscopy. Fig. 1a shows that RB4 (50 µmol/L) in aqueous solution (10 mmol/L HEPES buffer solution, pH=7) exhibits a maximum absorption peak at approx. 607 nm as well as an obvious dark blue colour. Upon gradual addition of Cu²⁺ (1.97–122 µmol/L), the absorption intensity at 607 nm decreased and shifted to around 619 nm accompanied with a colour change from dark blue to deep sky blue (right inset, Fig. 1a). The absorbance changes versus the Cu²⁺ concentration increase is plotted in the left inset of Fig. 1a. A 1:1 binding of the formed complex between RB4 and Cu²⁺ was confirmed by Yoe and Jones method (Fig. S1; supplementary material available at: www.ftb.com.hr) (27-29). The binding constant (K_s) of the [RB4-
-Cu²⁺] complex was also estimated to be (4.46±0.12)·10⁵ L/mol by nonlinear curve fitting (31).

The colorimetric response of [RB4-Cu²⁺] complex towards oxalate was studied by the UV-Vis spectrometric titration experiment. When oxalate ion was added gradually (1.76–96.0 µmol/L) to the complex solution, the absorption band intensity increased at 607 nm (Fig.1b). The colour of the solution also gradually changed from deep sky blue to dark blue (right inset, Fig. 1b). Left inset in Fig. 1b is also the graph of the UV-Vis absorption changes of the RB4-Cu²⁺ complex as a function of the concentration of oxalate (mol/L). The absorption spectra and the obtained colour were identical with that of RB4. This colorimetric change indicates that [RB4-Cu²⁺] complex can be used for the determination of oxalate by the naked eye.

The application of the Yoe and Jones method (27-29) by plotting the absorbance at 607 nm versus the molar ratio of C₂O₄^2– to Cu²⁺ confirmed a stoichiometry of 1:1 for the interaction of Cu²⁺ and C₂O₄^2– (Fig. 2) with the association constant K_s=(2.3±0.1)·10⁷ L/mol obtained by nonlinear curve fitting (Fig. S2) (31, 32). Two oxygen atoms from oxalate act as a bidentate ligand. Thus, the displacement process can be as follows:

As mentioned above, the binding constant of Cu²⁺ with oxalate is larger than that of Cu²⁺ with RB4. Therefore, the indicator displacement by oxalate is highly favourable due to higher affinity of oxalate towards Cu²⁺.

The selectivity of the proposed chemosensor towards oxalate ion in the presence of ascorbic acid and other anions was investigated. The solution system was selective towards oxalate anion over ascorbic acid (C₆H₈O₆), SO₄²ˉ, NO₃ˉ, CO₃²ˉ, HPO₄²ˉ, PO₄³ˉ, AcOˉ, ClO₄ˉ, Fˉ, Clˉ and Brˉ (Fig. 3). Only the oxalate anion is able to generate colour and spectral changes. When ascorbic acid and other anions (500 µmol/L) were added into the [RB4-Cu²⁺] complex solution, negligible changes in colour and UV-Vis spectrum were observed, showing that they do not interfere with the detection of the presence or amount of oxalate anion. Therefore, it was proven that this chemosensor has selectivity towards oxalate in water solution without the need for sophisticated instruments.

Scheme 1 shows the proposed IDA strategy for the determination of oxalate (C₂O₄²ˉ). This strategy is based on the competition of the analyte and the chromogenic indicator (RB4) for the interaction with Cu²⁺. As already mentioned, a 1:1 complex with the stability constant of (4.46±0.12)∙10⁵ L/mol formed after the addition of Cu²⁺ to the RB4 solution. There are some changes in the FTIR spectrum of the RB4-Cu²⁺ complex from 1570 to 1616 cm^–1 (B in Fig. 4). Obviously, the frequency of the N-H (primary amine) bending vibration weakened at 1570 cm^–1. At the same time, the frequency of the C=O stretching vibration also shifted from 1616 to 1603 cm^–1. Furthermore, one band was observed at 692 cm^–1, which is attributed to Cu-O stretching. The existence of these changes suggests that –NH₂ and C=O groups can participate in the coordination (Scheme 1).

In FTIR spectrum of RB4-Cu²⁺+C₂O₄^2– (C in Fig. 4), the existence of the frequencies of C=O, –NH₂, S=O, C–Cl stretching vibrations, C–N–C triazine bending vibration and N–H wag shows that these groups are free and do not participate in coordination. Furthermore, the bands observed at 1312 and 525 cm^–1 correspond to the stretching vibrations of C–O of oxalate and Cu–OCCO, respectively. The presence of these bands indicates the Cu–C₂O₄^2– association, which can imply better affinity of oxalate towards Cu²⁺ than that of RB4 (Scheme 1). The addition of C₂O₄^2– to the RB4-Cu²⁺ complex solution restored the colour and spectrum of RB4 (Fig. 5). C₂O₄^2– has a stronger binding capability with Cu²⁺ ion than with RB4 and can form a more stable CuC₂O₄, so after the addition of C₂O₄, the UV-Vis absorption spectrum and colour of RB4 were recovered and the colour and absorbance changes were reversible with the addition of Cu²⁺ and C₂O₄^2–.

The effective and suitable pH range (values between 3 and 10) for the selective determination of C₂O4^2– was studied. At lower pH values (<5) and acidic conditions, the colorimetric response decreased (Fig. S3), which may be due to protonation of the heteroatoms of RB4 and the prevention of the complex formation with Cu²⁺. The concentration of free oxalate is also low because all oxalate ions are protonated. Oxalate is the deprotonated form of oxalic acid (pK_a1=1.23 and pK_a2=4.19). In the acidic environment, the interaction of C₂O₄^2– with H⁺ causes a decrease in C₂O₄^2– and HC₂O₄^– and the production of H₂C₂O₄.

Under alkaline conditions (pH>8), the colorimetric response was also weak due to the interaction of Cu²⁺ with hydroxide ion and the lack of the RB4-Cu²⁺ complex formation. However, the absorption intensity of [RB4-Cu²⁺] complex solution showed significant changes only in the pH range of 5–7 upon the addition of oxalate (Fig. S3). Furthermore, the best colorimetric response of RB4-Cu²⁺ to oxalate was obtained at pH=7. The results show that under approximately physiological conditions, the proposed chemosensor can be employed to detect oxalate.

Short response time is also another important characteristic of chemosensors in practical applications. We tested the response of the proposed chemosensor within 60 min. The maximum response for the determination of oxalate was obtained in a satisfactorily short time (less than 1 min) (Fig. S4). This fast response can provide a new real-time method for oxalate determination.

Molecular chemosensors represent an advancement of the molecular logic gate in sensing chemistry. Here, we constructed an IMPLICATION molecular logic gate based on the proposed colorimetric indicator displacement assay (Fig. S5). Since the significant characteristic of the proposed chemosensor was its optical signal response, Cu²⁺ and C₂O₄^2– as two inputs, and the colour and the UV-Vis absorbance at 607 nm as outputs were defined. Numbers 0 and 1 indicate the absence and presence of the inputs, respectively. In the output, 1 and 0 define the dark blue and deep sky blue colours, respectively. The presence and/or absence of C₂O₄^2– and Cu²⁺ inputs were defined by four pairs (0/0, 1/0, 0/1 and 1/1). Only one pair gives the output of 0 with an obvious deep sky blue colour and the spectral decrease due to [RB4-Cu²⁺] complex formation if the Cu²⁺ input is 1 and C₂O₄^2– 0 (Fig. S5a-5c). Thus, only with Cu²⁺ and without C₂O₄^2– input, the output of IMPLICATION is a deep sky blue (Fig. S5d). This logic function (Fig. S5) can be potential for oxalate determination.

Recently, the development of colorimetric methods along with the detection by the naked eye has been one of the subjects of interest in the analytical applications. There is an enormous interest in the colorimetric assay based on test paper as a simple, effective, fast, and low-cost sensing technology (5, 33). The paper-based test strips fabricated according to the instruction given in Materials and Methods section were used for colorimetric experiments. The evaluation of the practicability of the proposed colorimetric strategy is possible with these experiments. The proposed colorimetric IDA assay was repeated with the paper-based test strips and the primary colour recovery after the addition of oxalate to [RB4-Cu²⁺] was observed (A in Fig. S6). Test papers immersed in different concentrations of oxalate from 0.005 to 0.1 mol/L demonstrated different colours. This experiment can indicate the applicability of fast, simple and effective paper-based analytical strategy for oxalate determination (B in Fig. S6). A selectivity experiment was also performed using paper-based test strips. Upon the addition of anions to the [RB4-Cu²⁺] complex, only oxalate showed an obvious colour change compared with other available anions (C in Fig. S6c). The obtained results can verify the applicability of the proposed colorimetric strategy for ‘in-the-field’ oxalate measurement.

The analytical features of the colorimetric response of the present chemosensor for determination of oxalate ion were investigated. The calibration curves, attributed to the colorimetric determination of oxalate ion, were constructed using least square regression (Fig. S7 and Fig. S8). According to corresponding calibration curves, limit of detection (LOD) and quantitation (LOQ) were calculated using Eq. 1 and Eq. 2, respectively:andwhere δ is standard deviation, and S is slope.

The absorbance signal of RB4-Cu²⁺ increases linearly in the oxalate concentration range of 1.76–49.4 µmol/L (Fig. S7). The corresponding linear function is:

The limit of detection (3 δ/S) and the limit of quantification (10 δ/S) were calculated to be 0.62 and 2.07 µmol/L, respectively. These results exhibit excellent sensing performance of the sensor towards oxalate.

Fig. S8 shows the linear relationship of the absorbance signal with Cu²⁺ concentration coordinated with RB4 for the determination of oxalate in the range of 31 to 100.71 µmol/L with the regression equation:

The detection limit (3 δ/S) of Cu²⁺ was calculated to be 1.96 µmol/L.

Since the determination of oxalate ion concentration in food and human body fluids is important, the analytical performance of the proposed chemosensor was examined by analyzing urine, mushroom and spinach samples, which were prepared as described in the materials and methods section. The colorimetric responses of the IDA chemosensor to these samples were evaluated in three replicate measurements. The results are summarized in Table 1.

The analysis of the results of the three samples by the proposed method and the reference method (29) showed good agreement with each other. The RSD values and the range of recoveries (97.9–101.2%) suggested that the substances in these real samples have no serious interference for the detection of oxalate. Therefore, this chemosensor has potential applications in oxalate detection in real samples.

Table 2 compares the limit of detection, dynamic linear range and analytical parameters of our method with the colorimetric determination of oxalate (2, 34-36). Regarding the simplicity of the operation, rapid response and the ability of easy detection, the proposed chemosensor can serve as a probe for the determination of oxalate.