Highly Sensitive Sensor for the Determination of Riboflavin Using Thionine Coated Cadmium Selenide Quantum Dots Modified Graphite Electrode

In this paper, the electrochemical non-enzymatic detection of Riboflavin (RF) was proposed based on its catalytic reduction in a Thionine-coated Cadmium Selenide Quantum dots (TH@CdSe QDs)-modified paraffin wax-impregnated graphite electrode (PIGE) that was prepared using a novel approach. The synthesized TH@CdSe QDs were confirmed by UV-Vis spectroscopy, Confocal Raman Microscopy and High Resolution Transmission Electron Microscopy (HRTEM) studies. The electrochemical response of the TH@CdSe QDs-modified PIGE was studied by cyclic voltammetry. The voltammetric response of RF at the TH@CdSe QDs-modified PIGE showed higher current than the bare PIGE. Under optimum conditions, the electrocatalytic reduction currents of RF was found to be linearly related to its concentration over the range of 1.6 × 10−7 M to 1.4 × 10−4 M with a detection limit of 53 × 10−9 M (S/N = 3). The TH@CdSe QDs-modified PIGE was utilized as an amperometric sensor for the detection of RF in flow systems was performed by carrying out hydrodynamic and chronoamperometric experiments. The TH@CdSe QDs-modified PIGE showed very good stability and a longer shelf life. The applicability of the fabricated electrode was justified by the quantification of RF in commercial tablets.


Introduction
Vitamins play a vital role in biochemistry and metabolism as they are biologically active molecules. Among all vitamins, Riboflavin (RF), also called vitamin B 2 (7,8-dimethyl-10-ribityl-isoalloxazine), is water soluble and it is important for the growth, development and function of the cells in the body by converting carbohydrates, fats and proteins into energy [1][2][3]. It participates in redox reactions in animal species as a hydrogen transporter in glucose, protein and fat metabolism, hemoglobin synthesis and ocular visual function maintenance. As RF is not formed in the body, it is necessary to take it through diet such as milk, dairy products, meat, eggs and green vegetables. A deficiency in RF causes anemia, digestive problems, skin disorders, catarrh and cellular growth [4,5]. Thus, a determination of RF is needed to monitor human health. In the literature, various methods were utilized to detect RF, which includes fluorescence [6], MSPE/UPLC [7], electrochemiluminescence [8], Raman spectroscopy [9] and voltammetry [10,11]. The use of miniaturized and portable sensors is only possible because electrochemical instrumentation is also miniaturized and portable [12]. Among these techniques, voltammetric detection is a promising and suitable alternative method due to its low cost of operation, fastness, high sensibility and flexibility [13].

Characterization Techniques
The absorption spectra of the QDs were obtained using a UV-visible spectrophotometer model Agilent Diode Array spectrophotometer (Santa Clara, CA, USA). A quartz cuvette was used for the UV-visible spectrophotometer and the path length of the quartz cuvette is 10 mm. Raman spectra were recorded with a RAMAN-11i system (Nanophoton Corp., Osaka, Japan) for measuring the vibrational modes of the molecules. Surface topography, size and shape of the QDs were studied using HRTEM (Technai G2-model T-30 S-twin, FEI, Hillsboro, OR, USA).

Electrochemical Measurements
The electrochemical experiments were carried out using a CHI 660B electrochemical workstation (CH Instruments, Austin, TX, USA) coupled with an IBM personal computer using a conventional three-electrode cell at room temperature. A platinum electrode was used as an auxiliary electrode and a saturated calomel electrode as a reference electrode. TH@CdSe QDs-modified PIGE was used as a working electrode.

Fabrication of TH@CdSe QDs Modified PIGE
L-Cysteine-capped CdSe QDs were synthesized using the reported procedure [32]. Briefly, CdCl 2 (320 µM) and L-Cysteine ethyl ester hydrochloride (3.2 mM) was mixed with 20 mL aqueous solution in the three-necked round bottom flask, and the pH of solution was adjusted to 11 using NaOH (1.0 M). Subsequently, the addition of NaHSe (20.0 mL) was prepared by adding 0.32 mmol of selenium metal and 0.81 mmol of NaBH 4 in an N 2 atmosphere. The mixture was stirred for 30 min and then refluxed for four hours under N 2 atmosphere. After four hours an orange-red colloidal CdSe QDs was formed and stored in an air-tight glass container. In the next step, an optimized volume of 30.0 mL of prepared CdSe QDs was mixed with 5.0 mL of saturated solution of TH and was stirred continuously for four hours. The solution was then centrifuged, and the residue was washed many times with distilled water. The weakly absorbed dye molecules were removed during washing and the resulting precipitate was dried and it is used for further studies. The TH@CdSe QDs were dispersed in propanol by ultrasonication. About 5.0 µL of TH@CdSe QDs was drop casted at the PIGE and allowed to dry under room temperature. After it dried, the TH@CdSe QDs-modified PIGE was utilized for sensing applications. For comparing the electrochemical performance, bare PIGE, TH-modified PIGE and CdSe QDs modified-PIGE were fabricated.

UV-Visible Spectroscopy
UV-vis absorption spectroscopy is a very simple method and applicable to explore the interaction between CdSe QDs and TH. Figure 1 exhibits the absorption changes observed upon the addition of various concentrations of CdSe QDs on 150 µM cationic dye TH. With increasing CdSe QDs concentration, the absorption band of TH decreases and a new absorption band appears at shorter wavelengths. TH has a characteristic absorption maximum at 600 nm with a shoulder at 565 nm in aqueous solution [33]. Upon increasing the concentration of CdSe QDs, a new absorption peak has seen with a maximum at 520 nm. The decrease in the absorption band of the TH monomer confirms the formation of H-type aggregates with CdSe QDs. The presence of an isobestic point also authenticates the existence of two species that are responsible for the observed absorption changes [34]. Thus, with the addition of CdSe QDs, TH molecules are adsorbed on the surface of CdSe QDs due to the electrostatic attractive force between the cationic dye molecules and the negatively charged L-cysteine capped CdSe QDs surfaces aggregates, which causes a blue shift (H-type) in the absorption band [35,36]. This causes the formation of a close-packing arrangement of the dye molecules on the negatively charged particle surface [37], and thus increases the concentration of TH on particle surfaces. aggregates, which causes a blue shift (H-type) in the absorption band [35,36]. This causes the formation of a close-packing arrangement of the dye molecules on the negatively charged particle surface [37], and thus increases the concentration of TH on particle surfaces.

Confocal Raman Microscopy
The formation of TH@CdSe QDs was also confirmed with confocal Raman studies. Figure 2 displays the confocal Raman spectra of (a) TH and (b) TH-coated CdSe QDs. The Raman bands obtained for TH (a) are in good agreement with the early reports [38,39]. The Raman spectrum of Thionine displays characteristics peaks at 480 cm −1 , 531 cm −1 , 1484 cm −1 and 1626 cm −1 , corresponding to C-S-C stretching, asymmetric ring stretching, C-N-C bending and H-plane bending vibrations. In addition, The Raman spectrum of TH@CdSe QDs (b) are considerably similar to the vibrational bands obtained for TH, indicating no significant change in the structural features of TH upon interaction with CdSe QDs. The increase in the intensity of Raman bands of TH-coated CdSe QDs was attributed to the presence of CdSe QDs. From the spectrum, it is inferred that there exists strong electrostatic interactions between the positively charged dye molecules and negatively charged CdSe QDs.

Confocal Raman Microscopy
The formation of TH@CdSe QDs was also confirmed with confocal Raman studies. Figure 2 displays the confocal Raman spectra of (a) TH and (b) TH-coated CdSe QDs. The Raman bands obtained for TH (a) are in good agreement with the early reports [38,39]. The Raman spectrum of Thionine displays characteristics peaks at 480 cm −1 , 531 cm −1 , 1484 cm −1 and 1626 cm −1 , corresponding to C-S-C stretching, asymmetric ring stretching, C-N-C bending and H-plane bending vibrations. In addition, The Raman spectrum of TH@CdSe QDs (b) are considerably similar to the vibrational bands obtained for TH, indicating no significant change in the structural features of TH upon interaction with CdSe QDs. The increase in the intensity of Raman bands of TH-coated CdSe QDs was attributed to the presence of CdSe QDs. From the spectrum, it is inferred that there exists strong electrostatic interactions between the positively charged dye molecules and negatively charged CdSe QDs.
aggregates, which causes a blue shift (H-type) in the absorption band [35,36]. This causes the formation of a close-packing arrangement of the dye molecules on the negatively charged particle surface [37], and thus increases the concentration of TH on particle surfaces.

Confocal Raman Microscopy
The formation of TH@CdSe QDs was also confirmed with confocal Raman studies. Figure 2 displays the confocal Raman spectra of (a) TH and (b) TH-coated CdSe QDs. The Raman bands obtained for TH (a) are in good agreement with the early reports [38,39]. The Raman spectrum of Thionine displays characteristics peaks at 480 cm −1 , 531 cm −1 , 1484 cm −1 and 1626 cm −1 , corresponding to C-S-C stretching, asymmetric ring stretching, C-N-C bending and H-plane bending vibrations. In addition, The Raman spectrum of TH@CdSe QDs (b) are considerably similar to the vibrational bands obtained for TH, indicating no significant change in the structural features of TH upon interaction with CdSe QDs. The increase in the intensity of Raman bands of TH-coated CdSe QDs was attributed to the presence of CdSe QDs. From the spectrum, it is inferred that there exists strong electrostatic interactions between the positively charged dye molecules and negatively charged CdSe QDs.

Surface Morphology
The HRTEM images of CdSe QDs before and after the addition of TH are shown in Figure 3. The HRTEM image of CdSe QDs (a) shows the spherical shape with the size Micro 2023, 3 690 ranges between 2-8 nm. After the addition of TH, (b) it was found that the CdSe QDs exhibits similar particle shape and size, with some aggregation. The formation of TH@CdSe QDs further supports the above absorption spectra changes.

Surface Morphology
The HRTEM images of CdSe QDs before and after the addition of TH are shown in Figure 3. The HRTEM image of CdSe QDs (a) shows the spherical shape with the size ranges between 2-8 nm. After the addition of TH, (b) it was found that the CdSe QDs exhibits similar particle shape and size, with some aggregation. The formation of TH@CdSe QDs further supports the above absorption spectra changes.

Electrochemical Characterization of TH@CdSe QDs Modified PIGE
The electrochemical characterization of TH@CdSe QDs-modified PIGE was carried out using cyclic voltammetry in 0.1 M NH4NO3 and purged nitrogen. The CVs of bare PIGE (a), CdSe QDs-modified PIGE (b) TH-modified PIGE (c) and TH@CdSe QDsmodified PIGE (d) in 0.1 M NH4NO3 at a scan rate of 50 mV/s is shown in Figure 4. No peak was observed for the bare PIGE and CdSe QDs-modified PIGE in the potential range studied. Compared with bare PIGE, the background current for CdSe QDs-modified PIGE is apparently larger, which indicates that the electrode surface area is significantly enhanced after being modified with CdSe QDs. The TH-modified PIGE shows the cathodic and anodic peaks at potentials of −0.32 V and −0.21 V, respectively, and the peakto-peak separation (∆Ep) is found to be 0.11 V. On the other hand, TH@CdSe QDs-modified PIGE exhibited a pair of well-defined redox peaks with cathodic and anodic peak potentials of −0.2260 V and −0.1547 V, respectively. The redox peaks correspond to the two-electron process of immobilized TH [40] and the ∆Ep was found to be 0.0713 V. The decrease in the ∆Ep values together with the increase in the peak current indicate the reversibility of the electrode process, and this is attributed to the synergistic effect of CdSe QDs [41,42].

Electrochemical Characterization of TH@CdSe QDs Modified PIGE
The electrochemical characterization of TH@CdSe QDs-modified PIGE was carried out using cyclic voltammetry in 0.1 M NH 4 NO 3 and purged nitrogen. The CVs of bare PIGE (a), CdSe QDs-modified PIGE (b) TH-modified PIGE (c) and TH@CdSe QDs-modified PIGE (d) in 0.1 M NH 4 NO 3 at a scan rate of 50 mV/s is shown in Figure 4. No peak was observed for the bare PIGE and CdSe QDs-modified PIGE in the potential range studied. Compared with bare PIGE, the background current for CdSe QDs-modified PIGE is apparently larger, which indicates that the electrode surface area is significantly enhanced after being modified with CdSe QDs. The TH-modified PIGE shows the cathodic and anodic peaks at potentials of −0.32 V and −0.21 V, respectively, and the peak-to-peak separation (∆E p ) is found to be 0.11 V. On the other hand, TH@CdSe QDs-modified PIGE exhibited a pair of well-defined redox peaks with cathodic and anodic peak potentials of −0.2260 V and −0.1547 V, respectively. The redox peaks correspond to the two-electron process of immobilized TH [40] and the ∆E p was found to be 0.0713 V. The decrease in the ∆E p values together with the increase in the peak current indicate the reversibility of the electrode process, and this is attributed to the synergistic effect of CdSe QDs [41,42].

Effect of Supporting Electrolyte
The effect of varying the supporting electrolytes on the electrochemical performance of the modified electrode was evaluated. Figure 5 compares the CV response of TH@CdSe

Effect of Supporting Electrolyte
The effect of varying the supporting electrolytes on the electrochemical performance of the modified electrode was evaluated. Figure 5 compares the CV response of TH@CdSe QDs-modified PIGE in 0.1 M solutions of Ba(NO 3 ) 2 , BaCl 2 , NH 4 NO 3 , NH 4 Cl, KNO 3 , KCl, NaCl, NaNO 3 and PBS. It was observed that the modified electrode exhibits a relatively better CV response in NH 4 NO 3 . The ammonium ions are favorable for the proton transfer (redox reaction) of Thionine molecules in the electrode surface compared to other K + , Na + or Ba 2+ . Hence, 0.1 M NH 4 NO 3 was chosen as the background electrolyte for further electrochemical studies.

Effect of Supporting Electrolyte
The effect of varying the supporting electrolytes on the electrochemical performance of the modified electrode was evaluated. Figure 5 compares the CV response of TH@CdSe QDs-modified PIGE in 0.1 M solutions of Ba(NO3)2, BaCl2, NH4NO3, NH4Cl, KNO3, KCl, NaCl, NaNO3 and PBS. It was observed that the modified electrode exhibits a relatively better CV response in NH4NO3. The ammonium ions are favorable for the proton transfer (redox reaction) of Thionine molecules in the electrode surface compared to other K + , Na + or Ba 2+ . Hence, 0.1 M NH4NO3 was chosen as the background electrolyte for further electrochemical studies.

Effect of Scan Rate and pH
The electrochemical behaviour of the modified electrode at different scan rates from 10-100 mV/s was studied using cyclic voltammetry and the results are shown in Figure 6a. It was observed that on increasing the scan rate from 2-100 mV/s in 0.1 M NH 4 NO 3 , the peak currents also increased. When the peak currents were plotted against the scan rate, there was a linear increase, which suggests that the redox process is surface-confined (Figure 6b) [43].
The electrochemical rugosity (Γ) or surface concentration of TH was estimated [44] according to the Equation (1): where Q is the integrated charge passed through the modified electrode, A is the area of the electrode, and n, F have their usual meaning. The Γ of the modified electrode was calculated to be 2.77 × 10 −10 mol/cm 2 at a scan rate of 50 mV/s. The increase in the surface concentration could be attributed to the large surface area provided by the CdSe QDs, where more TH molecules were adsorbed effectively. The electrochemical behaviour of the modified electrode was also investigated by CV in 0.1 M NH 4 NO 3 in the pH range of 4-9. It was found that the current response of the TH@CdSe QDs-modified PIGE increases from 4.0 to 7.0 and reaches a maximum response at a pH of 7.0. Then, the peak current starts to decrease. Thus, a working pH of 7.0 was chosen for further measurements.
The electrochemical behaviour of the modified electrode at different scan rates from 10-100 mV/s was studied using cyclic voltammetry and the results are shown in Figure  6a. It was observed that on increasing the scan rate from 2-100 mV/s in 0.1 M NH4NO3, the peak currents also increased. When the peak currents were plotted against the scan rate, there was a linear increase, which suggests that the redox process is surface-confined (Figure 6b) [43]. The electrochemical rugosity (Γ) or surface concentration of TH was estimated [44] according to the Equation (1): where Q is the integrated charge passed through the modified electrode, A is the area of the electrode, and n, F have their usual meaning. The Γ of the modified electrode was calculated to be 2.77 × 10 −10 mol/cm 2 at a scan rate of 50 mV/s. The increase in the surface concentration could be attributed to the large surface area provided by the CdSe QDs, where more TH molecules were adsorbed effectively.
The electrochemical behaviour of the modified electrode was also investigated by CV in 0.1 M NH4NO3 in the pH range of 4-9. It was found that the current response of the TH@CdSe QDs-modified PIGE increases from 4.0 to 7.0 and reaches a maximum response at a pH of 7.0. Then, the peak current starts to decrease. Thus, a working pH of 7.0 was chosen for further measurements.

Electrocatalytic Behavior of TH@CdSe QDs Modified PIGE for the Reduction in RF
The electrocatalytic activity of the TH@CdSe QDs-modified PIGE for the reduction in RF was examined by cyclic voltammetric experiments in 0.1M NH4NO3 in the absence and presence of RF. The reduction in RF under similar conditions with the bare electrode was also tested. Figure 7A shows the CV response of the bare electrode (curve a and b) and TH@CdSe QDs-modified PIGE (curve c and d) in the absence and in the presence of 6.25 × 10 −5 M RF, respectively. As can be seen from the figure, the current response for the reduction in RF at the bare electrode was poor, but the modified electrode showed a better and considerable increase in the current response at −0.52 V. The oxidation and reduction potentials of RF are −0.44 V and −0.52 V, respectively, and show a quasi-reversible system.

Electrocatalytic Behavior of TH@CdSe QDs Modified PIGE for the Reduction in RF
The electrocatalytic activity of the TH@CdSe QDs-modified PIGE for the reduction in RF was examined by cyclic voltammetric experiments in 0.1M NH 4 NO 3 in the absence and presence of RF. The reduction in RF under similar conditions with the bare electrode was also tested. Figure 7A shows the CV response of the bare electrode (curve a and b) and TH@CdSe QDs-modified PIGE (curve c and d) in the absence and in the presence of 6.25 × 10 −5 M RF, respectively. As can be seen from the figure, the current response for the reduction in RF at the bare electrode was poor, but the modified electrode showed a better and considerable increase in the current response at −0.52 V. The oxidation and reduction potentials of RF are −0.44 V and −0.52 V, respectively, and show a quasi-reversible system. The presence of CdSe QDs in the modified electrode influenced the electrocatalytic activity in terms of peak potential and increased current response. Under optimal conditions, the cyclic voltammograms of the modified electrode were studied in the presence of an increasing concentration of RF. The TH@CdSe QDs-modified PIGE showed increased cathodic peak currents in the range of 1.6 × 10 −7 M to 1.4 × 10 −4 M (R 2 = 0.9940) with a detection limit of 53 × 10 −9 M (S/N = 3) with various additions of RF using the cyclic voltammetry technique ( Figure 7B). The above result indicates that the TH@CdSe QDsmodified PIGE exhibits an improved analytical performance for the determination of RF. The mechanism for electrocatalytic reduction in RF can be represented as shown in Scheme 1. The linear working range and detection limit obtained with the present sensor are better or comparable with earlier reports, as shown in Table 1.
The effect of solution pH on the electrochemical behavior of TH@CdSe QDs-modified PIGE in the presence of 1.6 × 10 −5 M RF was investigated by CV at various pHs (4-9) in 0.1 M NH 4 NO 3 and the current response was plotted against pH, as shown in Figure 8. It was observed that the modified electrode exhibits maximum and a stable peak current response at pH 7.0. Hence, the determination of RF is well favored at the modified electrode at a physiological pH, and it is suitable for real time applications.
increased cathodic peak currents in the range of 1.6 × 10 −7 M to 1.4 × 10 −4 M (R 2 = 0.9940) with a detection limit of 53 × 10 −9 M (S/N = 3) with various additions of RF using the cyclic voltammetry technique ( Figure 7B). The above result indicates that the TH@CdSe QDsmodified PIGE exhibits an improved analytical performance for the determination of RF. The mechanism for electrocatalytic reduction in RF can be represented as shown in Scheme 1. The linear working range and detection limit obtained with the present sensor are better or comparable with earlier reports, as shown in Table 1. with a detection limit of 53 × 10 −9 M (S/N = 3) with various additions of RF using the cyclic voltammetry technique ( Figure 7B). The above result indicates that the TH@CdSe QDsmodified PIGE exhibits an improved analytical performance for the determination of RF. The mechanism for electrocatalytic reduction in RF can be represented as shown in Scheme 1. The linear working range and detection limit obtained with the present sensor are better or comparable with earlier reports, as shown in Table 1.

Hydrodynamic and Chronoamperometric Studies
Hydrodynamic voltammetric studies were performed to optimize the working potential of RF determination in flow systems. The cathodic current was measured at constant intervals of applied potential, for 3.2 × 10 −5 M RF in the potential range of 0 to −0.8 V. The curves a and b of Figure 9 show the HDVs of the TH@CdSe QDs-modified electrode in the presence and absence of RF, respectively. As can be seen from the figure, the current response for the reduction in RF starts at a potential of −0.4 V and exhibits a maximum response at −0.55 V. Hence, a potential of −0.6 V was chosen as the working potential for the amperometric determination of RF. Figure 10A shows the amperogram obtained for the addition of 1.6 × 10 −5 M RF in a stirred solution of 0.1 M NH 4 NO 3 . A step-wise increase in the cathodic current was observed for every addition of RF at an applied potential of −0.6 V. For repeated additions of RF, the electrode gave a sharp increase in current and attained a steady state condition within 6 s. The corresponding calibration plot of catalytic current versus the concentration of RF is shown in Figure 10B, and it is observed that the modified electrode exhibits a linear current response with correlation coefficient of 0.9910.
The result indicates that the modified electrode exhibited good sensitivity and has a stable amperometric response under dynamic conditions.  The effect of solution pH on the electrochemical behavior of TH@CdSe QDs-modified PIGE in the presence of 1.6 × 10 −5 M RF was investigated by CV at various pHs (4-9) in 0.1 M NH4NO3 and the current response was plotted against pH, as shown in Figure 8. It was observed that the modified electrode exhibits maximum and a stable peak current response at pH 7.0. Hence, the determination of RF is well favored at the modified electrode at a physiological pH, and it is suitable for real time applications.

Hydrodynamic and Chronoamperometric Studies
Hydrodynamic voltammetric studies were performed to optimize the working potential of RF determination in flow systems. The cathodic current was measured at constant intervals of applied potential, for 3.2 × 10 −5 M RF in the potential range of 0 to −0.8 V. The curves a and b of Figure 9 show the HDVs of the TH@CdSe QDs-modified electrode in the presence and absence of RF, respectively. As can be seen from the figure, the current response for the reduction in RF starts at a potential of −0.4 V and exhibits a maximum response at −0.55 V. Hence, a potential of −0.6 V was chosen as the working potential for the amperometric determination of RF. Figure 10A shows the amperogram obtained for the addition of 1.6 × 10 −5 M RF in a stirred solution of 0.1 M NH4NO3. A stepwise increase in the cathodic current was observed for every addition of RF at an applied potential of −0.6 V. For repeated additions of RF, the electrode gave a sharp increase in current and attained a steady state condition within 6 s. The corresponding calibration plot of catalytic current versus the concentration of RF is shown in Figure 10B, and it is observed that the modified electrode exhibits a linear current response with correlation

Interference Studies
The anti-interference ability of the TH@CdSe QDs-modified PIGE towards the determination of RF was tested by DPV in the presence of various common ions such as Fe 3+ , Mg 2+ , Na + , K + , Ca 2+ , NO 3− and SO4 2− and some physiological interferents such as AA and UA. No change in the DPV current response was observed for 10 mM RF in the presence of a 100-fold excess of these metal ions and 50-fold excess of AA and UA. The results showed that the present modified electrode is highly selective towards the determination of RF.

Stability of TH@CdSe QDs Modified PIGE towards RF
The working stability of the TH@CdSe QDs-PIGE-modified PIGE towards the determination of RF was investigated by cycling the modified electrode in the presence of 3.2 × 10 −5 M RF in 0.1 M NH4NO3 at a scan rate of 50 mV/s for 50 continuous cycles. The results showed only a 1.5% loss from the initial peak current value after 50 cycles. This indicates that the modified electrode has good stability and does not undergo surface

Interference Studies
The anti-interference ability of the TH@CdSe QDs-modified PIGE towards the determination of RF was tested by DPV in the presence of various common ions such as Fe 3+ , Mg 2+ , Na + , K + , Ca 2+ , NO 3− and SO 4 2− and some physiological interferents such as AA and UA. No change in the DPV current response was observed for 10 mM RF in the presence of a 100-fold excess of these metal ions and 50-fold excess of AA and UA. The results showed that the present modified electrode is highly selective towards the determination of RF.

Stability of TH@CdSe QDs Modified PIGE towards RF
The working stability of the TH@CdSe QDs-PIGE-modified PIGE towards the determination of RF was investigated by cycling the modified electrode in the presence of 3.2 × 10 −5 M RF in 0.1 M NH 4 NO 3 at a scan rate of 50 mV/s for 50 continuous cycles. The results showed only a 1.5% loss from the initial peak current value after 50 cycles. This indicates that the modified electrode has good stability and does not undergo surface fouling. The long-term stability of the modified electrode was also studied for a period of 45 days. The modified electrode was stored in an airtight container when not in use. The modified electrode was used to determine the same concentrations of RF, which showed 97% of initial current, indicating that the electrode had excellent long-term stability.

Determination of RF in Commercial Tablets
The practical application of the modified electrode was studied by the determination of RF in two different commercially available tablets. The tablets were finely ground, and they were dissolved with distilled water and made into a known volume. Then a known volume of the prepared samples was spiked, and the results obtained are shown in Table 2. The recovery results allowed us to use the proposed sensor in the determination of RF in pharmaceutical tablets.

Conclusions
In summary, we have achieved a non-enzymatic TH-coated CdSe QDs-modified electrode for the determination of RF in nM levels. The prepared TH-coated CdSe QDs were characterized and confirmed by UV-Vis spectroscopy, Raman spectroscopy and HRTEM studies. Cyclic voltammetry and chronoamperometry studies confirm the excellent electrochemical behaviour of the modified electrode. The incorporation of CdSe QDs in TH reveals the better electrocatalytic performance of the modified electrode with very high sensitivity and with a very low detection limit. The proposed sensor is a promising candidate for the determination of RF in pharmaceutical tablets with good recovery.
Author Contributions: Conceptualization, methodology, software, validation, and writing-original draft preparation, A.K., formal analysis, and writing-review and editing, R.S.B.; visualization and supervision, S.S.N. All authors have read and agreed to the published version of the manuscript.
Funding: This research received no external funding.