Electrochemical Sensing Fabricated with Ta2O5 Nanoparticle-Electrochemically Reduced Graphene Oxide Nanocomposite for the Detection of Oxytetracycline

A novel tantalum pentoxide nanoparticle-electrochemically reduced graphene oxide nanocomposite-modified glassy carbon electrode (Ta2O5-ErGO/GCE) was developed for the detection of oxytetracycline in milk. The composition, structure and morphology of GO, Ta2O5, and Ta2O5-ErGO were characterized by X-ray diffraction (XRD) and scanning electron microscopy (SEM). Oxytetracycline electrochemical behavior on the bare GCE, GO/GCE, ErGO/GCE, and Ta2O5-ErGO/GCE was studied by cyclic voltammetry. The voltammetric conditions (including scan rate, pH, deposition potential, and deposition time) were systematically optimized. With the spacious electrochemical active area, the Ta2O5-ErGO/GCE showed a great magnification of the oxidation signal of oxytetracycline, while that of the other electrodes (GCE, GO/GCE, ErGO/GCE) could not reach the same level. Under the optimum conditions, the currents were proportional to the oxytetracycline concentration in the range from 0.2 to 10 μM, and a low detection limit of 0.095 μM (S/N = 3) was detectable. Moreover, the proposed Ta2O5-ErGO/GCE performed practically with satisfactory results. The preparation of Ta2O5-ErGO/GCE in the current work provides a minor outlook of detecting trace oxytetracycline in milk.


Introduction
Antimicrobial residues in food products of animal origin have become a great global safety concern. Oxytetracycline (OTC) is one of the antibiotics in the tetracycline group. It is commonly used to treat and prevent bacterial infections in livestock, as well as growth promotion. However, the non-prudent use of OTC in animals leads to accumulation of the residues in animal products such as milk. Small doses of antibiotic exposure to humans through the food chain poses health threats including hypersensitivity [1,2], antibacterial drug resistance [3], and toxic effects [4]. In protecting the public from exposure to residues, different regulatory organizations (Food and Agriculture Organization, European Union, and Food and Drug Authority) have set a maximum residue limit serving as a counter electrode, and saturated calomel electrode (SCE) serving as a reference electrode). Oxytetracycline of 10μMwas prepared in 1.0 M PBS (pH 6.5). With the deposition time (90 s) and potential (0 V), the cyclic voltammetry was recorded on the electrodes. The calibration curve was established by plotting the relationship between the measured current signal and analyte concentration. The amount of oxytetracycline in milk sample solutions was obtained by the standard addition method. Scheme 1. The preparation of a tantalum pentoxide nanoparticle-electrochemically reduced graphene oxide nanocomposite-modified glassy carbon electrode (Ta2O5-rGO/GCE) for the determination of tryptophan.

Materials Characterization
The morphologies of the different synthesized materials were first explored by SEM. As Figure  1A shows, profuse plication and a crumple-like surface structure were displayed, demonstrating the successful preparation of GO. Figure 1B shows the typical image of Ta2O5 nanoparticles. Each homogeneous particle is clearly observed. The estimated size is 750 nm, and the large particles are formed by the small particles with the aggregation. Though these small particles are smaller than 100nm, the true size could not be measured with the low resolution of the scanning electron Scheme 1. The preparation of a tantalum pentoxide nanoparticle-electrochemically reduced graphene oxide nanocomposite-modified glassy carbon electrode (Ta 2 O 5 -rGO/GCE) for the determination of tryptophan.

Equipment and Apparatus
Voltammograms were obtained through a cyclic voltammetry (CV) electrochemical technique using a CHI 660E electrochemical workstation from the company of Shanghai Chenhua Instruments, China. The experiment conducted involved the use of GCE as the working electrode, which was fabricated with a Ta 2 O 5 NPs-ErGO, platinum electrode and saturated calomel electrode as the counter and reference electrodes respectively. The pH analysis was conducted ona pH-3c exact digitalpH meter (Leichi Instrument Factory, Shanghai, China). Scanning electron microscopy was performed and images were obtained at an acceleration voltage of 2.0 kV by a scanning electron microscope (EVO10, ZEISS, Jena, Germany). The crystalline structure analysis of Ta 2 O 5-was operated with Cu Kα radiation (0.1542 nm) ona powder X-ray diffractometer (PANalytical, Amsterdam, Holland).

Ta 2 O 5 Nanoparticle Preparation
The hydrothermal method was applied in Ta 2 O 5 preparation [34]. Then, 0.01 M NaOH 100 mL) was poured rapidly into 300 mL of 0.05 M TaCl 5 containing 0.1ml of diethanolamine, which was added as a stabilizer. The solution was stirred for 1 h at room temperature and transferred into a hydrothermal autoclave, and then heated at 80 • C for 48 h for crystallization to occur. Upon cooling to room temperature, the powder was washed using distilled water until no precipitate remained between the filtrate and AgNO 3 . To remove impurities, the powder was washed with ethanol three times and dried at room temperature using a vacuum drier. To obtain Ta 2 O 5 nanoparticles, the dried powder was calcinated at 700 • C for 3 h in a muffle furnace.

Preparation of Ta 2 O 5 -GO
Commercial GO was used in preparing the dispersion. 20 mg of GO powder was mixed with water (20 mL) to obtain a solution of GO with the concentration of 1 mg/mL. Then, 1 mg Ta 2 O 5 nanoparticles was mixed in 10 mL of the prepared GO solution. Subsequently, the mixed solution was ultrasonicated for 2 h to form a Ta 2 O 5 -GO dispersion.

Electrode Fabrication
A glassy carbon electrode (GCE) surface (3 mm in diameter) was polished with different fine sizes (1.0, 0.3, and 0.05 µM) of α-Al 2 O 3 powder and then ultrasonicated in distilled water, absolute ethanol, and then distilled water for one minute in each and left to dry under an infrared lamp. The Ta 2 O 5 -GO dispersion (5µL) was drop-casted on the clean GCE surface and left to dry on air. The Ta 2 O 5 -ErGO/GCE was prepared by dipping the Ta 2 O 5 -GO/GCE in 0.1 M phosphate buffer solution (PBS, pH 6.5), and the reduction process was run at a constant potential of −1.2 V for 120 s. For the purpose of comparing the electrodes, a similar method was used to prepare other fabricated electrodes such as a GO/GCE, ErGO/GCE, and Ta 2 O 5 -GO/GCE.

Real Sample Pretreatment
Whole milk samples were bought from local supermarkets. For the detection of OTC, the milk samples were pretreated as follows: Weighing 3 g of trichloroacetic acid, and then dissolving it in 100 mL of PBS pH 6.5 solution. Next, 15 mL of milk was put in a 50 mL centrifuge tube, and the addition of 15 mL of the dissolved trichloroacetic acid was then followed by gently shaking. After that, centrifugation was done at 4000 rpm for 30 min. In order to obtain a very clear solution, the supernatant was filtered using the 0.22 µM pore filter. For practical testing of the developed sensor, milk was spiked with different concentrations of OTC.

Electrochemical Measurements
All the processes for material synthesis and electrochemical measurements are shown in Scheme 1. All electrochemical experiments were carried out by cyclic voltammetry (CV) with the three electrode systems (bare or modified GCE serving as a working electrode, platinum wire electrode serving as a counter electrode, and saturated calomel electrode (SCE) serving as a reference electrode). Oxytetracycline of 10 µM was prepared in 1.0 M PBS (pH 6.5). With the deposition time (90 s) and potential (0 V), the cyclic voltammetry was recorded on the electrodes. The calibration curve was established by plotting the relationship between the measured current signal and analyte concentration. The amount of oxytetracycline in milk sample solutions was obtained by the standard addition method.

Materials Characterization
The morphologies of the different synthesized materials were first explored by SEM. As Figure 1A shows, profuse plication and a crumple-like surface structure were displayed, demonstrating the successful preparation of GO. Figure 1B shows the typical image of Ta 2 O 5 nanoparticles. Each homogeneous particle is clearly observed. The estimated size is 750 nm, and the large particles are formed by the small particles with the aggregation. Though these small particles are smaller than 100 nm, the true size could not be measured with the low resolution of the scanning electron microscope images. The morphology of Ta 2 O 5 -GO is clearly shown in Figure 1C. The majority of nanoparticles are Biomolecules 2020, 10, 110 5 of 13 uniformly dispersed on the surface of GO, illustrating that the Ta 2 O 5 composited with GO nanosheets shows the excellent dispersity in contrast to the pure Ta 2 O 5 . Compared to the pure GO, Ta 2 O 5 -GO with an enormous specific surface area provides extra active sites that would absorb sufficient analytes.
Biomolecules 2020, 10, x 5 of 13 microscope images. The morphology of Ta2O5-GO is clearly shown in Figure 1C. The majority of nanoparticles are uniformly dispersed on the surface of GO, illustrating that the Ta2O5composited with GO nanosheets shows the excellent dispersity in contrast to the pure Ta2O5. Compared to the pure GO, Ta2O5-GO with an enormous specific surface area provides extra active sites that would absorb sufficient analytes. The crystalline structure of the GO, pure Ta2O5 nanoparticles, and Ta2O5-GO composite nanoparticles were further investigated by X-ray diffraction (XRD). As  [37], demonstrating that the synthetic Ta2O5nanoparticles were successfully prepared. Most importantly, not only the plane of GO, but also the those of Ta2O5 are observed in the XRD pattern of pure GO and Ta2O5-GO composites, indicating the successful synthesis of Ta2O5-GO composites.  Biomolecules 2020, 10, x 5 of 13 microscope images. The morphology of Ta2O5-GO is clearly shown in Figure 1C. The majority of nanoparticles are uniformly dispersed on the surface of GO, illustrating that the Ta2O5composited with GO nanosheets shows the excellent dispersity in contrast to the pure Ta2O5. Compared to the pure GO, Ta2O5-GO with an enormous specific surface area provides extra active sites that would absorb sufficient analytes.  [37], demonstrating that the synthetic Ta2O5nanoparticles were successfully prepared. Most importantly, not only the plane of GO, but also the those of Ta2O5 are observed in the XRD pattern of pure GO and Ta2O5-GO composites, indicating the successful synthesis of Ta2O5-GO composites.

Electrochemical Characterization of the Modified Electrodes
The cyclic voltammograms of bare GCE, GO/GCE, ErGO/GCE, and Ta 2 O 5 -ErGO/GCE were obtained in K 3 [Fe(CN) 6 ] containing 0.1 M KCl, as indicated in Figure 3. As i pa /i pc ≈ 1, the redox peaks appearing in each electrode tested were a pair of quasi-reversible ones. At the bare GCE, the potential separation (∆E p ) of the redox peaks was 0.079 V with the weak current of 60.52 (i pa ) µA and 63.87 (i pc ) µA, while with the presence of the poor conductive material GO, the redox peak currents of GO/GCE decreased precipitously with i pa = 23.96 µA and i pc = 24.18 µA. Moreover, after electrochemical reduction of GO to ErGO, the current on ErGO/GCE was higher than that of GO/GCE, implying a better conductivity of ErGO than that of GO. The intensity of the anodic and cathodic peak current of ErGO was 51.13 and 55.71 µA, respectively. Above all, the largest redox peak current (i pa = 135.7 µA, i pc = 134.5 µA) occurred to Ta 2 O 5 -ErGO/GCE, which approximately increased by twofold. The result shows that Ta 2 O 5 -ErGO/GCE dramatically improved electrochemical performances owing to the large specific surface area. For further investigation, the effective electroactive areas of various electrodes could be figured out by the Randles-Sevcik equation [21,22]: where the reduction peak current of K 3 [Fe(CN) 6 ] is indicated as i pc ; the number of electrons transferred during the redox process is indicated as n; the electrochemical active area (cm 2 ) is indicated as A; is indicated as C; and the scanning rate (V·s −1 ) is indicated as v. Thus, the electrochemical active areas of bare GCE, GO/GCE, ErGO/GCE, and Ta 2 O 5 -ErGO/GCE can be worked out as 0.0545, 0.0206, 0.0475, and 0.1147 cm 2 , respectively. The Ta 2 O 5 -ErGO/GCE electrochemical active area is almost twofold that of bare GCE, elucidating that the Ta 2 O 5 -ErGO composite modification enables the effective electroactive surface area to expand, promoting the adsorption of oxytetracycline on the electrode surface, and thus facilitating the redox process of oxytetracycline.

Electrochemical Characterization of the Modified Electrodes
The cyclic voltammograms of bare GCE, GO/GCE, ErGO/GCE, and Ta2O5-ErGO/GCE were obtained in K3[Fe(CN)6] containing 0.1 M KCl, as indicated in Figure 3. As ipa/ipc ≈ 1, the redox peaks appearing in each electrode tested were a pair of quasi-reversible ones. At the bare GCE, the potential separation (∆Ep) of the redox peaks was 0.079 V with the weak current of 60.52 (ipa) μA and 63.87 (ipc) μA, while with the presence of the poor conductive material GO, the redox peak currents of GO/GCE decreased precipitously with ipa = 23.96 μA and ipc = 24.18 μA. Moreover, after electrochemical reduction of GO to ErGO, the current on ErGO/GCE was higher than that of GO/GCE, implying a better conductivity of ErGO than that of GO. The intensity of the anodic and cathodic peak current of ErGO was 51.13 and 55.71 μA, respectively. Above all, the largest redox peak current (ipa = 135.7 μA, ipc = 134.5 μA) occurred to Ta2O5-ErGO/GCE, which approximately increased by twofold. The result shows that Ta2O5-ErGO/GCE dramatically improved electrochemical performances owing to the large specific surface area. For further investigation, the effective electroactive areas of various electrodes could be figured out by the Randles-Sevcik equation [21,22]: where    Figure 4, there are no cathodic peaks at all electrodes, while a redox peak is presented, illustrating that the oxidation of OTC is an irreversible process. Relatively, there is a weak anodic peak observed clearly at the bare GCE, which   Figure 4, there are no cathodic peaks at all electrodes, while a redox peak is presented, illustrating that the oxidation of OTC is an irreversible process. Relatively, there is a weak anodic peak observed clearly at the bare GCE, which is similar to that of the GO/GCE. Most importantly, after the electrochemical reduction, the signal of the peak current of ErGO is magnified, but it is still less than that of Ta 2 O 5 -ErGO. This study shows that Ta 2 O 5 -ErGO/GCE has an excellent detection performance for OTC, which was unmatched by the other electrodes. The result may be owed to the synergistic effect of ErGO and Ta 2 O 5 , further enhancing the current of the peak. Thus, in this work, Ta 2 O 5 -ErGO/GCE is the optimal choice for the determination of oxytetracycline.

Electrochemical Behavior of Oxytetracyclineon Various Electrodes
Biomolecules 2020, 10, x 7 of 13 is similar to that of the GO/GCE. Most importantly, after the electrochemical reduction, the signal of the peak current of ErGO is magnified, but it is still less than that of Ta2O5-ErGO. This study shows that Ta2O5-ErGO/GCE has an excellent detection performance for OTC, which was unmatched by the other electrodes. The result may be owed to the synergistic effect of ErGO and Ta2O5, further enhancing the current of the peak. Thus, in this work, Ta2O5-ErGO/GCE is the optimal choice for the determination of oxytetracycline.

Electrochemical Kinetics of Oxytetracycline on Ta2O5-ErGO/GCE
For a deep exploration of the reaction mechanism, the cyclic voltammograms of oxytetracycline (1.0 × 10 −5 M) were plotted under different scan rates (30-300 mV/s) on Ta2O5-ErGO/GCE. The result is presented in Figure 5A. Consequently, with increasing scan rate, the ipa of oxytetracycline rose as did the background current. Probably, high scanning rates are able to enhance the charging current of the double layer. Additionally, the linear relationship between the anodic peak currents of oxytetracycline (ipa) and the square root of the scan rate (v) is shown in Figure 5B. The linear equation of ipa(OTC) (μA) = 52.315v 1/2 (V/s) + 2.150 with a correlation coefficient (R 2 ) of 0.98, identifying that the electrooxidation of oxytetracycline belongs to an adsorption-limited process. Furthermore, for an adsorption-controlled and totally irreversible electrode process, the relationship between the peak potential and the scanning rate is based on the Lavrion equation [41]: where Ep is defined as the peak potential, E 0 is defined as the formal potential (V), T is defined as temperature (298.15 K), α is defined as the electron transfer coefficient, n is defined as the electron transfer number, k 0 is defined as the rate constant, F is defined as the Ferrari constant (F = 96485 C/mol), R = 8.314 J/(K·mol), and v is defined as the scan rate (s). The relationship between the oxidation peak potentials (Ep) and Napierian logarithm of scanning rates (lnv) is shown in Figure 5C.

Electrochemical Kinetics of Oxytetracycline on Ta 2 O 5 -ErGO/GCE
For a deep exploration of the reaction mechanism, the cyclic voltammograms of oxytetracycline (1.0 × 10 −5 M) were plotted under different scan rates (30-300 mV/s) on Ta 2 O 5 -ErGO/GCE. The result is presented in Figure 5A. Consequently, with increasing scan rate, the i pa of oxytetracycline rose as did the background current. Probably, high scanning rates are able to enhance the charging current of the double layer. Additionally, the linear relationship between the anodic peak currents of oxytetracycline (i pa ) and the square root of the scan rate (v) is shown in Figure 5B. The linear equation of i pa(OTC) (µA) = 52.315v 1/2 (V/s) + 2.150 with a correlation coefficient (R 2 ) of 0.98, identifying that the electrooxidation of oxytetracycline belongs to an adsorption-limited process. Furthermore, for an adsorption-controlled and totally irreversible electrode process, the relationship between the peak potential and the scanning rate is based on the Lavrion equation [41]: where E p is defined as the peak potential, E 0 is defined as the formal potential (V), T is defined as temperature (298.15 K), α is defined as the electron transfer coefficient, n is defined as the electron transfer number, k 0 is defined as the rate constant, F is defined as the Ferrari constant (F = 96485 C/mol), R = 8.314 J/(K·mol), and v is defined as the scan rate (s). The relationship between the oxidation peak potentials (E p ) and Napierian logarithm of scanning rates (lnv) is shown in Figure 5C. Obviously, the oxidation peaks directly moved toward a positive potential in different degrees at various scanning rates. The oxidation potentials (E p ) are linearly proportional to the Napierian logarithm of the scan rate (lnv) with the linear equation: E p = 0.040 lnv + 0.822 (R 2 = 0.9931). According to Equation (2)

Effect of pH
For the investigation of the effect of solution pH, the CVs of 10 μΜ oxytetracycline were recorded in different pH of PBS, as shown in Figure 6A. With the enhancement of pH, the oxidation peaks directly shift toward negative potential, and the peak potentials (Ep) decrease linearly with pH, as depicted in Figure 5B. The linear equation of peak potential and pH is Ep = −0.049pH + 1.0313 (R 2 = 0.9818), illustrating that protons (H + ) participated in the electrochemical reaction of oxytetracycline. As Figure 6C depicts, with the increase in pH, the ipa increases constantly until the pH is 6.5. Subsequently, the pH keeps aggrandizing, while the oxidation current of oxytetracycline decreases. Thus, pH = 6.5 is the optimal pH for detecting the oxytetracycline in this study.

Effect of Deposition Parameters
The influence of deposition potential for the current of the anodic oxytetracycline was investigated in the voltage range of −0.2-1.0 V. Oxytetracycline was deposited on the surface of Ta2O5-ErGO/GCE at numerous potentials for 90 s, and the ipa of them were then carried out by CV. As Figure  7A shows, the peak current increases with the positive potential. When the deposition potential is over 0 V, the value of the peak current reduces, which is far less than that at 0 V. Thus, 0 V was suggested as the best deposition potential in this work. Furthermore, the effect of deposition time was studied at a fixed deposition potential of 0 V as well. In Figure 7B, it is evident to observe that the ipa of oxytetracycline enlarges when the deposition time prolongs from 0 to 180 s. The ipa gradually increases until 90 s first. However, after 90 s, the currents are unfluctuating and roughly the same as they are at 90 s. This result demonstrates that the deposition time can effectively improve the

Effect of pH
For the investigation of the effect of solution pH, the CVs of 10 µM oxytetracycline were recorded in different pH of PBS, as shown in Figure 6A. With the enhancement of pH, the oxidation peaks directly shift toward negative potential, and the peak potentials (E p ) decrease linearly with pH, as depicted in Figure 5B. The linear equation of peak potential and pH is E p = −0.049pH + 1.0313 (R 2 = 0.9818), illustrating that protons (H + ) participated in the electrochemical reaction of oxytetracycline. As Figure 6C depicts, with the increase in pH, the i pa increases constantly until the pH is 6.5. Subsequently, the pH keeps aggrandizing, while the oxidation current of oxytetracycline decreases. Thus, pH = 6.5 is the optimal pH for detecting the oxytetracycline in this study.

Effect of pH
For the investigation of the effect of solution pH, the CVs of 10 μΜ oxytetracycline were recorded in different pH of PBS, as shown in Figure 6A. With the enhancement of pH, the oxidation peaks directly shift toward negative potential, and the peak potentials (Ep) decrease linearly with pH, as depicted in Figure 5B. The linear equation of peak potential and pH is Ep = −0.049pH + 1.0313 (R 2 = 0.9818), illustrating that protons (H + ) participated in the electrochemical reaction of oxytetracycline. As Figure 6C depicts, with the increase in pH, the ipa increases constantly until the pH is 6.5. Subsequently, the pH keeps aggrandizing, while the oxidation current of oxytetracycline decreases. Thus, pH = 6.5 is the optimal pH for detecting the oxytetracycline in this study.

Effect of Deposition Parameters
The influence of deposition potential for the current of the anodic oxytetracycline was investigated in the voltage range of −0.2-1.0 V. Oxytetracycline was deposited on the surface of Ta2O5-ErGO/GCE at numerous potentials for 90 s, and the ipa of them were then carried out by CV. As Figure  7A shows, the peak current increases with the positive potential. When the deposition potential is over 0 V, the value of the peak current reduces, which is far less than that at 0 V. Thus, 0 V was suggested as the best deposition potential in this work. Furthermore, the effect of deposition time was studied at a fixed deposition potential of 0 V as well. In Figure 7B, it is evident to observe that the ipa of oxytetracycline enlarges when the deposition time prolongs from 0 to 180 s. The ipa gradually increases until 90 s first. However, after 90 s, the currents are unfluctuating and roughly the same as they are at 90 s. This result demonstrates that the deposition time can effectively improve the

Effect of Deposition Parameters
The influence of deposition potential for the current of the anodic oxytetracycline was investigated in the voltage range of −0.2-1.0 V. Oxytetracycline was deposited on the surface of Ta 2 O 5 -ErGO/GCE at numerous potentials for 90 s, and the i pa of them were then carried out by CV. As Figure 7A shows, the peak current increases with the positive potential. When the deposition potential is over 0 V, the value of the peak current reduces, which is far less than that at 0 V. Thus, 0 V was suggested as the best deposition potential in this work. Furthermore, the effect of deposition time was studied at a fixed deposition potential of 0 V as well. In Figure 7B, it is evident to observe that the i pa of oxytetracycline enlarges when the deposition time prolongs from 0 to 180 s. The i pa gradually increases until 90 s first. However, after 90 s, the currents are unfluctuating and roughly the same as they are at 90 s. This result demonstrates that the deposition time can effectively improve the sensitivity of detecting oxytetracycline, recommending 90 s for the optimal choice. Therefore, the deposition was carried out at 0 V for 90 s in the following performances.
Biomolecules 2020, 10, x 9 of 13 sensitivity of detecting oxytetracycline, recommending 90 s for the optimal choice. Therefore, the deposition was carried out at 0 V for 90 s in the following performances.

Selectivity Reproducibility and Stability Investigation
At the optimum analytical conditions, the selectivity performance was studied by measuring the response peak currents in the presence of potential interfering substances containing some metal ions and other antibiotics. The experimental results showed the 100-fold of Zn 2+ , Al 3+ , Ca 2+ , Mg 2+ , Cu 2+ , Na+, and K+, and the 20-fold of amoxicillin and tetracycline, which have no apparent influences on the response peak current of 10 μM oxytetracycline (signal change ≤ ±5%). The experimental results illustrate that the proposed Ta2O5-ErGO/GCE has high selectivity for the OTC determination.
For the evaluation of reproducibility, five different electrodes modified with Ta2O5-ErGO were measured in 10 μM oxytetracycline in 0.1 M PBS with a pH of 6.5. The relative standard deviation (RSD) for oxytetracycline was 2.34%, implying the good reproducibility. Additionally, the modified electrode stability was tested through running the cyclic voltammograms of 10 μM oxytetracycline, which was recorded approximately 25 times for 5 days. After the test, a drop of 13% was observed on the current signal of oxytetracycline. This shows that the Ta2O5-ErGO/GCE prepared have excellent stability.

Calibration Curve of Oxytetracycline
The calibration curve for oxytetracycline at Ta2O5-ErGO/GCE was characterized by cyclic voltammetry under the optimal experimental conditions. The typical voltammograms are depicted in Figure 8A. In the range of 0.2-100 μM, the oxidation peak current is reasonably linear to oxytetracycline concentration ( Figure 8B). The linear regression equation can be described as ipa (μA) = 0.332cOTC (μM) + 0.328 (R 2 = 0.9893). According to the equations, LOD = 3S/N (S: Blank solution standard deviation; N: The slope of calibration plots), and the detection limit was identified as 0.095 μM. It mainly ascribes the excellent sensing performance to the synergistic effect of Ta2O5 and ErGO mentioned above.
In comparison to other electrochemical methods developed, as summarized in Table 1, this study provides a wide linear range and detection limit, which is relatively comparable to the fabricated electrodes reported for OTC determination. The minimum limit of detection indicates a good performance of the developed sensor. Thus, the modified electrode Ta2O5-ErGO/GCE can be used effectively for practical applications.

Selectivity Reproducibility and Stability Investigation
At the optimum analytical conditions, the selectivity performance was studied by measuring the response peak currents in the presence of potential interfering substances containing some metal ions and other antibiotics. The experimental results showed the 100-fold of Zn 2+ , Al 3+ , Ca 2+ , Mg 2+ , Cu 2+ , Na+, and K+, and the 20-fold of amoxicillin and tetracycline, which have no apparent influences on the response peak current of 10 µM oxytetracycline (signal change ≤ ±5%). The experimental results illustrate that the proposed Ta 2 O 5 -ErGO/GCE has high selectivity for the OTC determination.
For the evaluation of reproducibility, five different electrodes modified with Ta 2 O 5 -ErGO were measured in 10 µM oxytetracycline in 0.1 M PBS with a pH of 6.5. The relative standard deviation (RSD) for oxytetracycline was 2.34%, implying the good reproducibility. Additionally, the modified electrode stability was tested through running the cyclic voltammograms of 10 µM oxytetracycline, which was recorded approximately 25 times for 5 days. After the test, a drop of 13% was observed on the current signal of oxytetracycline. This shows that the Ta 2 O 5 -ErGO/GCE prepared have excellent stability.

Calibration Curve of Oxytetracycline
The calibration curve for oxytetracycline at Ta 2 O 5 -ErGO/GCE was characterized by cyclic voltammetry under the optimal experimental conditions. The typical voltammograms are depicted in Figure 8A. In the range of 0.2-100 µM, the oxidation peak current is reasonably linear to oxytetracycline concentration ( Figure 8B). The linear regression equation can be described as i pa (µA) = 0.332c OTC (µM) + 0.328 (R 2 = 0.9893). According to the equations, LOD = 3S/N (S: Blank solution standard deviation; N: The slope of calibration plots), and the detection limit was identified as 0.095 µM. It mainly ascribes the excellent sensing performance to the synergistic effect of Ta 2 O 5 and ErGO mentioned above.
In comparison to other electrochemical methods developed, as summarized in Table 1, this study provides a wide linear range and detection limit, which is relatively comparable to the fabricated electrodes reported for OTC determination. The minimum limit of detection indicates a good performance of the developed sensor. Thus, the modified electrode Ta 2 O 5 -ErGO/GCE can be used effectively for practical applications.

Detection of Oxytetracycline in Real Sample
CV was employed for quantitative analysis. The experimental results are listed in Table2. There is no oxytetracycline in the milk supernatant sample. Besides, the good recoveries (100.1-120.9%) of the proposed Ta2O5-ErGO/GCE can be applied well to the OTC detection in real samples.

Conclusions
In this study, the Ta2O5 was synthesized and then composited with GO to obtain a Ta2O5-GO composition. Ta2O5-GO was dispersed in pure water and subsequently drop-coated on the surface of GCE. Subsequently, Ta2O5-ErGO/GCE was fabricated by the electrochemical reduction method. In comparing the bare GCE, GO/GCE, ErGO, and Ta2O5-ErGO/GCE, the oxidation signals of oxytetracycline were significantly enlarged on the nanocomposite-modified GCE (Ta2O5-ErGO/GCE). In addition, this modified electrode has successfully been used to eliminate the interferences of oxytetracycline. The linear range for the determination of oxytetracycline was 0.2-100 μM on Ta2O5-

Detection of Oxytetracycline in Real Sample
CV was employed for quantitative analysis. The experimental results are listed in Table 2. There is no oxytetracycline in the milk supernatant sample. Besides, the good recoveries (100.1-120.9%) of the proposed Ta 2 O 5 -ErGO/GCE can be applied well to the OTC detection in real samples.

Conclusions
In this study, the Ta 2 O 5 was synthesized and then composited with GO to obtain a Ta 2 O 5 -GO composition. Ta 2 O 5 -GO was dispersed in pure water and subsequently drop-coated on the surface of GCE. Subsequently, Ta 2 O 5 -ErGO/GCE was fabricated by the electrochemical reduction method. In comparing the bare GCE, GO/GCE, ErGO, and Ta 2 O 5 -ErGO/GCE, the oxidation signals of oxytetracycline were significantly enlarged on the nanocomposite-modified GCE (Ta 2 O 5 -ErGO/GCE). In addition, this modified electrode has successfully been used to eliminate the interferences of oxytetracycline. The linear range for the determination of oxytetracycline was 0.2-100 µM on