1 Introduction

Designing specific layered double hydroxides (LDHs) with tailor-made properties to suit the desired applications have drawn great attention due to their potential applications in biotechnology, energy storage, separation science, adsorbents, photochemistry, electrochemistry, etc.1,2,3,4,5,6,7,8,9,–10 Generally, LDHs are synthetic lamellar solids with metal cations (trivalent and divalent) and anions (n-valent). Their general formula can be given as [M 2+1−y M 3+y (OH)2]y+ [(An−)y/n, y H2O].11 LDHs can be synthesized by conventional methods like ion-exchange, co-precipitation, and hydrothermal methods.12,13 However, these methods are not efficient in controlling the morphology, surface area, and particle size of the LDHs.13 Due to this limitation, the application of LDHs as catalysts, adsorbents, and electrode modifiers are limited.14 To overcome this drawback and to explore their electrocatalytic properties, a new Ni–Fe LDH which is synthesized using a template based on Fe(III) metal-organic framework (MOF) is reported. This new Ni–Fe LDH is envisaged to exhibit high surface area, high adsorption capacity, and enhanced catalytic activity. Kojic acid (KA), otherwise known as 5-hydroxy-2-(hydroxymethyl)-4-pyrone (based on IUPAC recommended nomenclature) is globally used as an antioxidant, food additive, or preservative.15,16,–17 KA is generally extracted from the fungal genus Aspergillus.15,16,–17 However, its safety assessment is controversial due to carcinogenicity, teratogenicity, and embryotoxicity.18,19 Therefore, monitoring of KA concentration levels is very important due to its human health issues, food safety, and quality control aspects.18,20 Real-time monitoring/determination of KA can be achieved by several techniques like ion-pair liquid chromatography, capillary electrophoresis, fluorimetry and electrochemical methods.21,22,23,–24 Among these methods, the electrochemical method generally receives great attention since it can offer a sensitive, selective, simple, and cost-effective route for many target analyte determinations.25,26,–27 Accordingly, in this work, we report an electrochemical method for the quantitative assay of KA using a MOF templated Ni–Fe LDH for the first time.

2 Experimental

2.1 Materials and reagents

N,N-dimethylformamide (DMF), poly(vinyl alcohol), nickel acetate tetrahydrate and benzene-1,4-dicarboxylic acid were supplied by S.D. Fine Chemicals, India. Potassium ferrocyanide, potassium ferricyanide, potassium chloride, K2HPO4, KH2PO4, and ferric nitrate were procured from Qualigens, India. Kojic acid (KA) was obtained from Himedia, India. Stock solutions of K2HPO4 and KH2PO4 were used for the preparation of 0.1 M pH 7.0 phosphate buffer. Triple distilled water was used for electrochemical experiments.

Fe-MIL-88 and Ni–Fe LDH were synthesized hydrothermally as described before in our previous publication.14 The synthesis of Fe-MIL-88 is carried out using a mixture of benzene-1,4-dicarboxylic acid and ferric nitrate in DMF-ethanol-water mixture. The mixture was ultra-sonicated, stirred, refluxed at 80 °C under the stirring condition for 12 h, and finally subjected to centrifugation (8000 rpm). The obtained precipitate (Fe-MIL-88) was washed multiple times with a mixture of ethanol-water and then dried for 24 h at 60 °C. Synthesis of Ni–Fe LDH was accomplished by mixing the required amounts of Fe-MIL-88 (0.02 g) and nickel acetate tetrahydrate (0.2 M) in DMF. This mixture was then subjected to ultra-sonication and refluxed with stirring at 85 °C for 4 h. Further, DMF and water were added to the mixture and refluxed at 120 °C for an additional 12 h. Then the mixture was cooled to room temperature and centrifuged to get the Ni–Fe LDH. The formed Ni–Fe LDH was filtered, washed with a large amount of water, and dried for 5 h at 120 °C.14,28

2.2 Apparatus

X-ray diffraction (XRD) patterns (BRUKER D8 X-ray diffractometer with Cu Kα line (0.159 nm), X-ray photoelectron spectroscopy (XPS, AMICUS spectrometer, UK), scanning electron microscopy (SEM) (SEM VEGA 3 TESCAN at 30 kV) and transmission electron microscopy (TEM) (TECNAI 20 G2 FEI microscope; 120 kV) analyses were carried out for the characterization of the materials. Cyclic voltammetry, amperometry, and chronoamperometry measurements were accomplished with CHI-660C electrochemical workstation (CH Instruments, USA). Glassy carbon (GC) or modified GC, platinum wire and saturated calomel electrode i.e. Hg/Hg2Cl2/KCl(sat.) were used as working, counter, and reference electrodes, respectively. GC electrode surface is modified either with Fe-MIL-88 or Ni–Fe LDH. The colloid of the respective materials (0.1 wt%) with poly(vinyl alcohol) (0.01 wt%) were prepared and typically 20 µL of the colloid was drop coated and dried for 4 h to get the modified GC electrodes. The modified electrodes were represented as GC/Fe-MIL-88 or GC/Ni–Fe LDH. It should be noted that the present work reports the construction of Fe-MIL-88 or Ni–Fe LDH film prepared on GC electrodes. However, carbon paste electrodes (CPEs) using the Fe-MIL-88 or Ni–Fe LDH materials were already constructed, characterized, and demonstrated for the efficient water oxidation in alkaline medium.14

3 Results and Discussion

3.1 Characterization of Ni–Fe LDH

Detailed characterization of Fe-MIL-88 and Ni–Fe LDH were elaborately discussed based on the results of TEM, XRD patterns, and XPS in our earlier study.14 The powder XRD data are provided in supplementary information, Table S1 (Supplementary Information).14 The powder XRD patterns of Fe-MIL-88 show peaks at the 2θ values of 8.9°, 16.2°, and 17.3° which can be attributed to (002), (103), and (200) planes of Fe-MIL-88, respectively (Table S1, Supplementary Information).14 Ni–Fe LDH shows additional peaks at the 2θ values 11.7°, 23.7°, 34.8°, 39.5°, 59.5°, and 61.0° which are ascribed to various crystal planes of Ni–Fe LDH, respectively (Table S1, Supplementary Information).14 Spindle shaped morphology of Fe-MIL-88 and porous layered structure of Ni–Fe LDH are revealed from the SEM and TEM analyzes (Figure S1, Supplementary Information).14 The XPS of Fe 2p1/2 (726.1 eV) and Fe 2p3/2 (712.6 eV) in both Fe-MIL-88 and Ni–Fe LDH materials confirm the +3 oxidation state of Fe (Table S2, Supplementary Information,).14 Similarly, the XPS of Ni 2p3/2 (856.7 and 851.8 eV) and Ni 2p1/2 (872.3 eV) in Ni–Fe LDH confirm the presence of +2 and +3 oxidation states of Ni (Table S2, Supplementary Information).14 The FT-IR spectrum14 of Ni–Fe LDH displays peaks at 1107 and 876 cm−1 which can be corroborated to the deformation of –OH. A peak around 876 cm−1 is also observed14 owing to the out of plane deformation of carbonate species present in the Ni–Fe LDH.14 Electrochemical impedance analyzes of CPEs based on these materials were already reported.14 In the present study, electrochemical impedance analyzes are carried out with the Fe-MIL-88 and Ni–Fe LDH film modified electrodes to understand the electron transfer activities within the film. Aqueous ferrocyanide/ferricyanide redox couple (10.0 mM K3[Fe(CN)6] and 10.0 mM K2[Fe(CN)6]) is used as a probe and the analyzes are carried out with 0.1 M KCl by applying 0.2 V at GC/Fe-MIL-88 and GC/Ni–Fe LDH electrodes (Figure 1). The cyclic voltammetry (CV) responses of the probe at GC/Fe-MIL-88 and GC/Ni–Fe LDH are also recorded (inset (i) of Figure 1). Both GC/Fe-MIL-88 and GC/Ni–Fe LDH electrodes show redox peaks corresponding to [Fe(CN)6]3−/4− redox process. GC/Fe-MIL-88 shows a slightly low peak current with a ΔEp value of 140 mV while the GC/Ni–Fe LDH electrode shows a high peak current with a ΔEp value of 180 mV. These results suggest that the electrical conductivity inside the Ni–Fe LDH film is greater than the Fe-MIL-88 film (based on high current obtained at GC/Ni–Fe LDH electrode). Nyquist plots demonstrate that two different physicochemical processes take place at the electrode-solution interface which may be referred as charge transfer phenomena inside the sheets of Ni–Fe LDH or Fe-MIL-88 and a heterogeneous electron transfer process between the modified electrode and the electrolyte solution.29 A suitable Randle’s equivalent circuit is identified (inset (ii) of Figure 1) which relies on the experimentally observed data. As per the Randle’s equivalent circuit, it possesses three resistance components, the resistance of electrolyte (R1), charge transfer resistance (Rct), and resistive part of the mass transfer (Rp), mass transfer resistance (Zd), the capacitance of the film (Cdl) and two constant phase elements (Q1 and Q2). The Ni–Fe LDH displays a semicircle with a lesser diameter than the Fe-MIL-88 indicating lower Rct value than that of the Fe-MIL-88. This low Rct leads to a fast electron transfer process at GC/Ni–Fe LDH electrode. Presence of Ni3+ in the Ni–Fe LDH and/or the broad interlayer space of Ni–Fe LDH may cause increased conductivity within the film.14,30

Figure 1
figure 1

Nyquist plots for the GC/Fe-MIL-88 (a) and GC/Ni–Fe LDH (b). Inset (i) represents the cyclic voltammograms of GC/Fe-MIL-88 (a) and GC/Ni–Fe LDH (b) in a mixture of 10.0 mM K3[Fe(CN)6], 10.0 mM K2[Fe(CN)6], and 0.1 M KCl. Inset (ii) shows the suitable Randle’s equivalent circuit.

Full size image

3.2 Electrochemical determination of KA

Cyclic voltammograms are recorded in phosphate buffer (0.1 M, pH 7.0) with and without KA (100 µM) using GC, GC/Fe-MIL-88, and GC/Ni–Fe LDH electrodes (Figure 2). No significant redox behavior is noticed in the absence of KA at all the electrodes (inset of Figure 2). However, in the presence of KA, a well-defined voltammetric peak is observed due to the efficient electro-oxidation of KA at all the three electrodes (Figure 2a′–2c′). At bare GC and GC/Fe-MIL-88 electrodes (Figure 2a′ and 2b′), KA is oxidized at 0.79 V with a low current. At GC/Ni–Fe LDH electrode, KA is oxidized at 0.75 V with a high current. Thus, GC/Ni–Fe LDH exhibits low oxidation peak potential and high peak current for KA electro-oxidation in comparison to bare GC and GC/Fe-MIL-88 electrodes which indicate the efficient electrocatalytic activity of GC/Ni–Fe LDH. This electrocatalytic activity may be due to the involvement of Ni3+ present in Ni–Fe LDH.

Figure 2
figure 2

Cyclic voltammograms of bare GC (a and a′), GC/Fe-MIL-88 (b and b′), and GC/Ni–Fe LDH (c and c′) in 0.1 M phosphate buffer (pH 7.0) in the presence (a′, b′, and c′) and absence (a, b, and c) of 100 µM KA at 20 mV s−1 scan rate.

Full size image

Efficient electrocatalytic oxidation of KA at the GC/Ni–Fe LDH electrode may lead to a sensitive determination of KA which can be verified by amperometry using the GC/Ni–Fe LDH electrode. The amperometry responses arising from successive additions of KA are recorded (Eapplied: 0.6 V) in 0.1 M pH 7.0 phosphate buffer (Figure 3A). The calibration curve from the amperometry responses is drawn for KA determination which shows two linear responses (Figure 3B). The first linear response is spread over the KA concentration range from 1.0 µM to 1.5 mM and the next linear range is observed from 1.5 to 4.5 mM. The sensitivity and detection limit for the KA determination are found to be 0.032 µA µM−1 cm−2 and 0.73 µM, respectively. On increasing the concentration above 1.5 mM of KA, the rate of the electrocatalytic reaction may become slow due to saturation of the active catalytic sites31 which results in a decreased sensitivity of the second linear region of the calibration plot. The sensitivity of the present method is comparable to many of the recently reported methods.15,17,32,33,34,35,36,–37 Although certain methods show slightly high sensitivity, they exhibit a narrow linear calibration range.15,35 Thus, the present method demonstrates higher sensitivity and wider linear calibration range than the previous methods for the determination of KA.15,35,37 Therefore, the present method is comparable or superior with previously reported methods for KA determination as shown in Table 1. 32,33,34,35,36,–37

Figure 3
figure 3

(A) Amperometric response of GC/Ni–Fe LDH with successive additions of KA (1, 5, 10, 20, 30, 50, 100, 200, 300, 500, 1000, 1500, 2000, 2500, 3000, 3500, 4000 and 4500 µM) into 0.1 M phosphate buffer. Expanded view from 1 to 30 µM additions of KA is shown in the inset. (B) Calibration curve showing the two linear ranges (1–1500 µM and 1500–4500 µM) for KA determination with error bars.

Full size image
Table 1 An assessment of key analytical factors for the electrochemical KA determination.
Full size table

3.3 Chronoamperometry of KA oxidation

The oxidation of KA is further studied by chronoamperometry to understand the kinetics of the GC/Ni–Fe LDH electrode process and shown in Figure 4. From the slope of the plot shown in the inset of Figure 4, the value of the catalytic rate constant (KC) is calculated to be 1.96 × 105 M−1 s−1 using equation 1.17

$$ \frac{{I_{cat} }}{{I_{L} }} = \sqrt {K_{C} C_{b} \pi t} $$
(1)

where Icat (catalytic current) is the anodic oxidation currents measured in the presence of different concentrations of KA and IL (limiting current) is the current obtained in the absence of KA. Cb is the bulk concentration of KA and t is the time. The value of KC for KA oxidation agrees practically with other reported values.17

Figure 4
figure 4

Chronoamperograms of GC/Ni–Fe LDH in 0.1 M phosphate buffer (pH 7.0) with varying concentrations of KA (a–e; 0, 100.0, 200.0, 400.0 and 600.0 μM, respectively). Inset shows the Icat/IL for 100 μM of KA.

Full size image

Based on the above studies and in line with the literature,38 the mechanism of electrocatalytic oxidation of KA by the Ni–Fe LDH may be postulated as displayed in equations 2–4.

$$ {\text{Ni}}\left( {\text{II}} \right) - {\text{Fe}}\;{\text{LDH}} \rightleftarrows {\text{Ni}}\left( {\text{III}} \right) - {\text{Fe}}\,{\text{LDH}} + {\text{e}}^{ - } $$
(2)
(3)
(4)

In the given mechanism, Ni(II)-Fe LDH is electrochemically oxidized to Ni(III)-Fe LDH and KA is chemically oxidized to KA(ox) by the electrogenerated Ni(III)-Fe LDH involving one electron and one proton. The oxidized KA (KA(ox)) undergoes further reaction to produce 6,6′-bis(5-hydroxy-2-(hydroxymethyl)-4H-pyran-4-one as the final product (equation 4).39

3.4 Reproducibility, stability, and interference studies

The reproducibility of the GC/Ni–Fe LDH electrode is analyzed by determining 100 µM KA three times. The relative standard deviation for the three determinations is calculated as 5.6% which reveals an acceptable reproducibility. The storage stability of GC/Ni–Fe LDH is tested by keeping the electrode under the room temperature condition for 7 days and comparing the CV response of the first and seventh days for the determination of 100 µM KA. On the seventh day, GC/Ni–Fe LDH exhibits 95.0% of the initial signal (first-day current response), which reveals reasonable stability of the GC/ Ni–Fe LDH.

The selectivity of the GC/Ni–Fe LDH electrode towards the determination of 100 µM KA is analyzed in the presence of several possible interferents (Figure 5). The influence of possible coexisting common salts (2.0 mM of each salt like KNO3, MgCl2, CaCl2, NaHCO3, Ni(NO3)2 and HgCl2) and biomolecules (1.0 mM of each compound like caffeine, glucose, aspartic acid, alanine, and glutamic acid) in the determination of KA is investigated (Figure 5).34 Many of the added interferents exhibit negligible influence (less than ±5%) during the determination of 100 µM KA. Few interferents show a signal change slightly higher than ± 5%, however, less than 10% indicating the high selectivity of this method. Thus, GC/Ni–Fe LDH can be used for the reliable quantitative determination of KA.

Figure 5
figure 5

Percentage signal alteration for the determination of KA in the presence of possible interferents at GC/Ni–Fe LDH electrode in 0.1 M phosphate buffer (pH 7.0) at 20 mV s−1.

Full size image

3.5 Determination of KA in different tomato sauces

The realistic use of the sensor is established by analyzing KA present in the commercially available tomato sauces, A (Tombo), B (Kissan), and C (Maggi pichkoo). The samples (sauce A, B, and C) were diluted with water and analyzed using GC/Ni–Fe LDH electrode in 0.1 M phosphate buffer. Tomato sauce may contain Fe3+, ascorbic acid, oxalic acid, gallotannic acid, etc. However, they are estimated to be present in a very small amount. Therefore, these interferents cannot affect the response due to KA. The amperometry current response is measured for all three sauces (alone and also spiked with standard KA). The results are illustrated in Table 2. Reasonable recovery justifies the use of the GC/Ni–Fe LDH electrode for the estimation of KA in real samples. The results are comparable with the previously reported methods.21,40

Table 2 Determination of KA present in different brands of tomato sauce samples.
Full size table

4 Conclusions

Fe-MIL-88 is used as a template to synthesize Ni–Fe LDH. Based on the efficient electrocatalytic oxidation of KA by Ni–Fe LDH, the quantitative determination of KA is realized. The performance characteristics of the GC/Ni–Fe LDH electrode with those of the earlier reports revealed that the proposed sensor exhibits low overpotential for the oxidation of KA. The quantitative determination of KA is not affected even in the presence of many possible interfering compounds. It could be useful for the direct determination of KA in real samples without any pretreatments.