Electrochemical Determination of 17-β-Estradiol Using a Glassy Carbon Electrode Modified with α-Fe2O3 Nanoparticles Supported on Carbon Nanotubes

In this study, a novel electrochemical assay for determining 17-β-estradiol (E2) was proposed. The approach involves modifying a glassy carbon electrode (GCE) with a nanocomposite consisting of α-Fe2O3 nanoparticles supported on carbon nanotubes (CNTs)—denoted as α-Fe2O3-CNT/GCE. The synthesis of the α-Fe2O3-CNT nanocomposite was achieved through a simple and cost-effective hydrothermal process. Morphological and chemical characterization were conducted using scanning electron microscopy (SEM), Raman spectroscopy, and energy-dispersive X-ray spectroscopy (EDX). The presence of the α-Fe2O3-CNT film on the GCE surface resulted in an enhanced electrochemical response to E2, preventing electrode surface fouling and mitigating the decrease in peak current intensity during E2 oxidation. These outcomes substantiate the rationale behind the GCE modification. After the optimization of experimental conditions, E2 was determined by the square wave voltammetry technique using 0.1 mol L−1 KCl solution (pH = 7.0) with 20% ethanol as a supporting electrolyte. A linear concentration range of 5.0–100.0 nmol L−1 and a low limit of detection of 4.4 nmol L−1 were obtained. The electroanalytical method using α-Fe2O3-CNT/GCE was applied for E2 determination in pharmaceutical, lake water, and synthetic urine samples. The obtained results were attested by recovery tests and by high-performance liquid chromatography as a comparative technique at a 95% confidence level. Thus, the developed electrochemical sensor is simple and fast to obtain, presents high accuracy, and is viable for determining E2 in routine analysis.


Introduction
In the last several years, the effect of endocrine-disrupting compounds in animal and human systems has drawn the attention of scientists, research communities, and the public.They do this by imitating the biological functions of natural hormones, occupying hormonal receptors, and disrupting the transportation and metabolic processes of these natural hormones [1,2].The 17-β-estradiol hormone (E2) is an estrogen with significant endocrine-disrupting capabilities.It is produced in the ovaries and is responsible for maintaining the menstrual cycle, reproductive system, and lipid metabolism.It also plays a role in the growth and development of the skin, brain, and sinus tissue.Additionally, E2 is present in the testicles and is important for maintaining bone structure and sperm production [3].Moreover, E2 is also used as a medication for hormone replacement therapy and contraceptives [4].
E2 is excreted daily by humans and released into aquatic environments and sewage effluents along with industrial waste, as water treatment systems do not completely remove this hormone and its derivatives.The bioaccumulation of E2 in fish, birds, and reptiles Molecules 2023, 28, 6372 2 of 15 can lead to immune system complications and reproductive anomalies.When present at abnormal levels in humans, E2 can cause health problems such as ovary cancer, cirrhosis, hyperthyroidism, early puberty, hypertension, and polycystic ovary syndrome [3,[5][6][7].
Hence, the detection of E2 using methods with high sensitivity and low detection limits is very important.In the literature, methods for determining E2 mainly employ gas [8] and liquid [9][10][11][12] chromatography.Most of these methods require tedious sample preparation and the analysis is very delayed.
Over the past few decades, electroanalytical methods have arisen as potent substitutes for conventional analysis techniques.The strength of these methods lies in their versatility and advantageous analytical capabilities, offering an array of benefits including costeffectiveness, heightened sensitivity, portability, and user-friendliness.As a result, many electroanalytical methods stand as viable alternatives for E2 determination, owing to their uncomplicated instrumentation, economical nature, and the absence of intricate sample pre-treatment procedures or the necessity for toxic organic reagents [2].
Therefore, in the present work, the primary objective is to propose a novel electroanalytical method utilizing a modified electrode with a nanocomposite obtained through a simple, less time-consuming, and cost-effective hydrothermal method.
Electrodes based on carbonaceous materials are widely used because they are inexpensive and readily available, have a broad potential range and low background current, and are chemically inert during electrochemical analysis [24,25].Additionally, the modification of these electrodes with different materials can improve the selectivity and high detection limits of the electroanalytical methods.As a result, the use of chemically modified electrodes has significantly increased, leading to the need for exploring new materials that provide high surface area, stability, and excellent conductivity [26][27][28][29].Among the most used materials were metallic oxides [30,31] and carbonaceous nanomaterials such as graphene nanosheets [32,33], and carbon nanotubes (CNTs) [34,35].
CNTs are cylindrical structures produced by the winding of graphene sheets composed of carbon atoms with sp 2 hybridization in a hexagonal structure.They find wide applications in electrochemistry, primarily due to their conjugated π-bonds, which give them a steric bulkiness like an alkene with electron deficiency.Additionally, this material consists of hexagonal nets with high aromaticity, which enhances conductivity on its surface [36].Therefore, CNTs are commonly employed in electrode modifications due to their large electrochemical window, excellent conductivity, expansive surface area, and stability, all amplifying electron transfer [17].
α-Fe 2 O 3 , a metal oxide, possesses great utility and finds applications in numerous fields like magnetic storage devices, supercapacitors, photocatalysts, and sensors [37,38].It possesses advantageous qualities such as non-toxicity, biocompatibility, and affordability.Additionally, the α phase of this semiconductor is formed during the final stage of iron oxide synthesis, contributing to its exceptional chemical stability and resistance to corrosion [39,40].In the literature, there are examples of the use of the α-Fe 2 O 3 and CNT composites in the modification of electrodes.They have been primarily used for the electroanalytical determination of drugs [41][42][43].
In this context, the purpose of the present work was a novel assay of electrochemical E2 determination using a glassy carbon electrode (GCE) modified with a nanocomposite of α-Fe 2 O 3 nanoparticles supported by CNTs, obtained by a simple and inexpensive hydrothermal synthesis.

Chemical and Morphological Characterization of α-Fe 2 O 3 -CNT/GCE
Raman, SEM, and EDX were used to investigate the successful preparation of α-Fe 2 O 3 /CNT nanocomposite, and the results are shown in Figure 1.As can be seen from Figure 1A, in the Raman spectra of the CNTs the D and G band characteristics of this material for α-Fe 2 O 3 observed bands located at 227, 294, and 407 cm −1 can be well assigned to Eg modes of Fe 2 O 3 [44].The Raman spectra obtained for α-Fe 2 O 3 -CNT show only the bands corresponding to Fe-O bonds, characteristic of this iron oxide.It was not possible to observe characteristic bands of carbonaceous materials in the region from 1330 to 1600 cm −1 .This can be justified by the small region where the laser beam is incident on the materials, performing measurements in areas where no carbonaceous materials were present, and/or by the low concentration of these materials in the hydrothermal synthesis.
In this context, the purpose of the present work was a novel assay of electrochemical E2 determination using a glassy carbon electrode (GCE) modified with a nanocomposite of α-Fe2O3 nanoparticles supported by CNTs, obtained by a simple and inexpensive hydrothermal synthesis.

Chemical and Morphological Characterization of α-Fe2O3-CNT/GCE
Raman, SEM, and EDX were used to investigate the successful preparation of α-Fe2O3/CNT nanocomposite, and the results are shown in Figure 1.As can be seen from Figure 1A, in the Raman spectra of the CNTs the D and G band characteristics of this material for α-Fe2O3 observed bands located at 227, 294, and 407 cm −1 can be well assigned to Eg modes of Fe2O3 [44].The Raman spectra obtained for α-Fe2O3-CNT show only the bands corresponding to Fe-O bonds, characteristic of this iron oxide.It was not possible to observe characteristic bands of carbonaceous materials in the region from 1330 to 1600 cm −1 .This can be justified by the small region where the laser beam is incident on the materials, performing measurements in areas where no carbonaceous materials were present, and/or by the low concentration of these materials in the hydrothermal synthesis.In EDX spectra (in Figure 1B), the presence of essentially three elements in the materials: carbon, iron, and oxygen, was observed, as expected [45].This indicated that the synthesis of this material was satisfactory, and based on the intensities of the bands, a higher percentage of iron and oxygen compared to the carbonaceous material can be noted.This confirmed the initial proportions of the materials used, with a larger amount of the iron precursor when compared with CNTs.The SEM image (Figure 1C) obtained for α-Fe2O3-CNT showed a material in the form of rods with good homogeneity, and the CNTs were coated with α-Fe2O3.

Electrochemical Behavior of E2 at α-Fe2O3-CNT/GCE
Cyclic voltammetry (CV) technique assays were performed in the presence of 0.1 µmol L −1 of E2 in a 0.1 mol L −1 KCl solution with 10% v/v ethanol (to prevent the precipitation of E2).Initially, we observed a comparison of the bare GCE and GCE modified with α-Fe2O3-CNT, and the concentration of α-Fe2O3-CNT in the suspension also was evaluated, with concentrations of 1.0 and 2.0 mg mL −1 .As shown in Figure 2, the hormone E2 exhibits a single oxidation peak around 0.8 V, indicating an irreversible electrochemical process.This behavior is consistent with other reports found in the literature regarding E2 determination [6,16,[46][47][48].In EDX spectra (in Figure 1B), the presence of essentially three elements in the materials: carbon, iron, and oxygen, was observed, as expected [45].This indicated that the synthesis of this material was satisfactory, and based on the intensities of the bands, a higher percentage of iron and oxygen compared to the carbonaceous material can be noted.This confirmed the initial proportions of the materials used, with a larger amount of the iron precursor when compared with CNTs.The SEM image (Figure 1C) obtained for α-Fe 2 O 3 -CNT showed a material in the form of rods with good homogeneity, and the CNTs were coated with α-Fe 2 O 3 .

Electrochemical Behavior of E2 at α-Fe 2 O 3 -CNT/GCE
Cyclic voltammetry (CV) technique assays were performed in the presence of 0.1 µmol L −1 of E2 in a 0.1 mol L −1 KCl solution with 10% v/v ethanol (to prevent the precipitation of E2).Initially, we observed a comparison of the bare GCE and GCE modified with α-Fe 2 O 3 -CNT, and the concentration of α-Fe 2 O 3 -CNT in the suspension also was evaluated, with concentrations of 1.0 and 2.0 mg mL −1 .As shown in Figure 2, the hormone E2 exhibits a single oxidation peak around 0.8 V, indicating an irreversible electrochemical process.This behavior is consistent with other reports found in the literature regarding E2 determination [6,16,[46][47][48].
The voltammograms obtained with the bare GCE for E2 showed a peak with good current intensity.However, after the modification with α-Fe 2 O 3 -CNT, the peak current increased, indicating an improvement in the interaction with the electrode surface and an effective enhancement in the conductivity of the working electrode.This can be attributed to a reduction in the electron transfer resistance and an increase in the electrode surface area.Furthermore, two concentrations were evaluated, and no significant difference in the voltammetric response was observed.Thus, the α-Fe 2 O 3 -CNT suspension at a concentration of 1.0 mg mL −1 was adopted for further experiments.The voltammograms obtained with the bare GCE for E2 showed a peak with good current intensity.However, after the modification with α-Fe2O3-CNT, the peak current increased, indicating an improvement in the interaction with the electrode surface and an effective enhancement in the conductivity of the working electrode.This can be attributed to a reduction in the electron transfer resistance and an increase in the electrode surface area.Furthermore, two concentrations were evaluated, and no significant difference in the voltammetric response was observed.Thus, the α-Fe2O3-CNT suspension at a concentration of 1.0 mg mL −1 was adopted for further experiments.
The composition and pH of supporting electrolytes play an important role in the electrochemical response.Therefore, KCl solution (0.1 mol L −1 ), phosphate buffer solution (0.01 mol L −1 ), and Britton-Robinson buffer solution (0.04 mol L −1 ), all at pH 7.0, were evaluated (results not shown).It was observed that the KCl solution provided a peak with higher intensity and a lower oxidation potential when compared to the others.Thus, KCl 0.1 mol L −1 was chosen as the supporting electrolyte for the next experiments.
The influence of the pH of the KCl solution was also investigated (see Figure SM-1 in Supplementary Material).For pH values above 10.0, the E2 oxidation process becomes kinetically unfavorable as it leads to the passivation of the electrode surface.High peak current intensities were obtained at pH 6.0, 7.0, and 8.0.However, the electrochemical response at pH 7.0 exhibited better peak definition and repeatability (RSD of 0.65%; n = 3).According to Mustafa et al. (2004) [49], α-Fe2O3 nanoparticles have a point of zero charge (pH pcz) of 6.50.Conversely, the pKa of the E2 molecule is 10.7.Therefore, at pH 7.0, the E2 molecule is protonated, while the α-Fe2O3 nanoparticles are negatively charged, resulting in an electrostatic interaction between α-Fe2O3-CNT/GCE and the E2 molecules.
A linear relationship between peak potential (Ep) and pH (Ep(V) = 1.00193-0.05060pH; R 2 = 0.9934) was obtained in the pH range of 2.0 to 6.0, where the Ep shifts to less positive values as the pH increases.This behavior was described by Vega et al. (2007) [50] as characteristic behavior for the oxidation of phenolic compounds.The slope of 0.0506 V pH −1 is close to the theoretical Nernstian slope of 0.0592 V pH −1 , which indicates that the E2 oxidation process involves the same number of protons and electrons.This result is in accordance with the oxidation mechanism of E2 proposed by Ngundi et al. (2003) [51], in which the oxidation reaction involves transferring two protons and two electrons.
The electrochemical behavior of α-Fe2O3-CNT/GCE was studied by CV at different scan rates (from 10 to 500 mV s −1 ) in the presence of E2, as can be seen in Figure 3.It was noted that the overpotential was shifted positively (Figure 3A), which is characteristic of an irreversible process.Furthermore, the relation of Ip versus ν 1/2 shown in Figure 3B reveals a linear behavior, and the slope of 0.6 obtained in the relation of log Ip versus log ν The composition and pH of supporting electrolytes play an important role in the electrochemical response.Therefore, KCl solution (0.1 mol L −1 ), phosphate buffer solution (0.01 mol L −1 ), and Britton-Robinson buffer solution (0.04 mol L −1 ), all at pH 7.0, were evaluated (results not shown).It was observed that the KCl solution provided a peak with higher intensity and a lower oxidation potential when compared to the others.Thus, KCl 0.1 mol L −1 was chosen as the supporting electrolyte for the next experiments.
The influence of the pH of the KCl solution was also investigated (see Figure S1 in Supplementary Material).For pH values above 10.0, the E2 oxidation process becomes kinetically unfavorable as it leads to the passivation of the electrode surface.High peak current intensities were obtained at pH 6.0, 7.0, and 8.0.However, the electrochemical response at pH 7.0 exhibited better peak definition and repeatability (RSD of 0.65%; n = 3).According to Mustafa et al. (2004) [49], α-Fe 2 O 3 nanoparticles have a point of zero charge (pH pcz) of 6.50.Conversely, the pK a of the E2 molecule is 10.7.Therefore, at pH 7.0, the E2 molecule is protonated, while the α-Fe 2 O 3 nanoparticles are negatively charged, resulting in an electrostatic interaction between α-Fe 2 O 3 -CNT/GCE and the E2 molecules.
A linear relationship between peak potential (E p ) and pH (E p (V) = 1.00193-0.05060pH; R 2 = 0.9934) was obtained in the pH range of 2.0 to 6.0, where the E p shifts to less positive values as the pH increases.This behavior was described by Vega et al. (2007) [50] as characteristic behavior for the oxidation of phenolic compounds.The slope of 0.0506 V pH −1 is close to the theoretical Nernstian slope of 0.0592 V pH −1 , which indicates that the E2 oxidation process involves the same number of protons and electrons.This result is in accordance with the oxidation mechanism of E2 proposed by Ngundi et al. (2003) [51], in which the oxidation reaction involves transferring two protons and two electrons.
The electrochemical behavior of α-Fe 2 O 3 -CNT/GCE was studied by CV at different scan rates (from 10 to 500 mV s −1 ) in the presence of E2, as can be seen in Figure 3.It was noted that the overpotential was shifted positively (Figure 3A), which is characteristic of an irreversible process.Furthermore, the relation of I p versus ν 1/2 shown in Figure 3B reveals a linear behavior, and the slope of 0.6 obtained in the relation of log I p versus log ν (inset Figure 3B) is close to the theoretical value of 0.5 for a diffusion-controlled process; these results confirm a diffusion-controlled process of species from solution to the electrode surface [52].
(inset Figure 3B) is close to the theoretical value of 0.5 for a diffusion-controlled process; these results confirm a diffusion-controlled process of species from solution to the electrode surface [52].where Ep is the anodic peak potential, Ep/2 is the potential associated with the peak current half-height, α is the charge-transfer coefficient (0.50), which is pre-determined for organic molecules, and n is the number of electrons consumed in the reaction [52,53].
Considering the mean value of Ep = 0.769 V and Ep/2 = 0.709 V obtained with CV assays in different scan rates, the value of n calculated was 1.6.This result is in accordance with the oxidation mechanism reported by Ngundi et al. (2003) [51], where the oxidation reaction involves the transfer of two electrons and two protons, as shown in Figure 4.

Determination of E2 at α-Fe2O3-CNT/GCE
As reported in the literature, it is common for the E2 electrochemical oxidation process to involve the adsorption of the E2 molecule or its oxidation products on the electrode surface.This can result in a decrease in Ipa (peak current) and reduced precision of the analytical method [6].Therefore, this phenomenon was evaluated using the square wave voltammetry (SWV) technique with the bare GCE, CNT/GCE, and α-Fe2O3-CNT/GCE.A total of 15 consecutive measurements were conducted in the presence of 0.1 µmol L −1 E2 in KCl solution (10% v/v ethanol) at pH 7.0, as shown in Figure 5.The results obtained in the scan rate study were also used to determine the number of electrons involved in the oxidation process of the E2 molecule through Equation ( 1): where E p is the anodic peak potential, E p/2 is the potential associated with the peak current half-height, α is the charge-transfer coefficient (0.50), which is pre-determined for organic molecules, and n is the number of electrons consumed in the reaction [52,53].
Considering the mean value of E p = 0.769 V and E p/2 = 0.709 V obtained with CV assays in different scan rates, the value of n calculated was 1.6.This result is in accordance with the oxidation mechanism reported by Ngundi et al. (2003) [51], where the oxidation reaction involves the transfer of two electrons and two protons, as shown in Figure 4.
(inset Figure 3B) is close to the theoretical value of 0.5 for a diffusion-controlled process; these results confirm a diffusion-controlled process of species from solution to the electrode surface [52].where Ep is the anodic peak potential, Ep/2 is the potential associated with the peak current half-height, α is the charge-transfer coefficient (0.50), which is pre-determined for organic molecules, and n is the number of electrons consumed in the reaction [52,53].
Considering the mean value of Ep = 0.769 V and Ep/2 = 0.709 V obtained with CV assays in different scan rates, the value of n calculated was 1.6.This result is in accordance with the oxidation mechanism reported by Ngundi et al. (2003) [51], where the oxidation reaction involves the transfer of two electrons and two protons, as shown in Figure 4.

Determination of E2 at α-Fe2O3-CNT/GCE
As reported in the literature, it is common for the E2 electrochemical oxidation process to involve the adsorption of the E2 molecule or its oxidation products on the electrode surface.This can result in a decrease in Ipa (peak current) and reduced precision of the analytical method [6].Therefore, this phenomenon was evaluated using the square wave voltammetry (SWV) technique with the bare GCE, CNT/GCE, and α-Fe2O3-CNT/GCE.A total of 15 consecutive measurements were conducted in the presence of 0.1 µmol L −1 E2 in KCl solution (10% v/v ethanol) at pH 7.0, as shown in Figure 5.

Determination of E2 at α-Fe 2 O 3 -CNT/GCE
As reported in the literature, it is common for the E2 electrochemical oxidation process to involve the adsorption of the E2 molecule or its oxidation products on the electrode surface.This can result in a decrease in I pa (peak current) and reduced precision of the analytical method [6].Therefore, this phenomenon was evaluated using the square wave voltammetry (SWV) technique with the bare GCE, CNT/GCE, and α-Fe 2 O 3 -CNT/GCE.A total of 15 consecutive measurements were conducted in the presence of 0.1 µmol L −1 E2 in KCl solution (10% v/v ethanol) at pH 7.0, as shown in Figure 5.As can be seen for the bare GCE, a significant decrease in peak current intensity was observed, with a relative standard deviation (RSD) of 71%.Furthermore, the peak potential shifted toward more positive values as the measurements were taken, indicating passivation of the GCE surface and a slower interaction between the electrode surface and the E2 molecule.When the GCE electrode was modified with CNT, it was observed that the variation was lower compared to the bare GCE, with an RSD of 26%.On the other hand, when the α-Fe2O3-CNT/GCE was used, the RSD was 4%, indicating that the modification of the GCE surface significantly improved the electrode performance by preventing fouling and enhancing the analytical signal.
After this previous study, the analytical curve was obtained, and SWV voltammograms were carried out with successive additions at concentrations of E2 ranging from 5.0 to 100.0 nmol L −1 , as shown in Figure 6A.The peak current obtained for E2 exhibits a linear dependence on the respective concentration (see Figure 6B) and is described by the equation: Ip (µA) = 0.0072 µA + 0.4083 µA µmol −1 L [E2] (µmol L −1 ) (R 2 = 0.998).As can be seen for the bare GCE, a significant decrease in peak current intensity was observed, with a relative standard deviation (RSD) of 71%.Furthermore, the peak potential shifted toward more positive values as the measurements were taken, indicating passivation of the GCE surface and a slower interaction between the electrode surface and the E2 molecule.When the GCE electrode was modified with CNT, it was observed that the variation was lower compared to the bare GCE, with an RSD of 26%.On the other hand, when the α-Fe 2 O 3 -CNT/GCE was used, the RSD was 4%, indicating that the modification of the GCE surface significantly improved the electrode performance by preventing fouling and enhancing the analytical signal.
After this previous study, the analytical curve was obtained, and SWV voltammograms were carried out with successive additions at concentrations of E2 ranging from 5.0 to 100.0 nmol L −1 , as shown in Figure 6A.The peak current obtained for E2 exhibits a linear dependence on the respective concentration (see Figure 6B  The limit of detection (LOD) was calculated as 3 std/m, where std represents the standard deviation of 10 determinations of the blank and m is the slope of the analytical curve.The obtained LOD value was found to be 4.4 nmol L −1 .Next, the values of intraand inter-day repeatability were obtained for two concentrations of E2 (10.0 and 50.0 nmol L −1 ).For the intra-day repeatability experiments, ten successive measurements (n = 10) were obtained on the same day, while for the inter-day repeatability measurements, they were obtained over five successive days (n = 3).The obtained RSD values ranged from 3.8% to 8.6%, indicating good precision of the proposed method.
The analytical parameters obtained in this proposed method were compared with other electroanalytical methods for the determination of E2 using different electrochemical sensors reported in the literature.As shown in Table 1, the method proposed here presents similar or smaller LOD values compared to most works found in the literature.The significant advantage of the proposed method lies in its relatively low cost and simple synthesis of the material.The limit of detection (LOD) was calculated as 3 std/m, where std represents the standard deviation of 10 determinations of the blank and m is the slope of the analytical curve.The obtained LOD value was found to be 4.4 nmol L −1 .Next, the values of intra-and inter-day repeatability were obtained for two concentrations of E2 (10.0 and 50.0 nmol L −1 ).For the intra-day repeatability experiments, ten successive measurements (n = 10) were obtained on the same day, while for the inter-day repeatability measurements, they were obtained over five successive days (n = 3).The obtained RSD values ranged from 3.8% to 8.6%, indicating good precision of the proposed method.
The analytical parameters obtained in this proposed method were compared with other electroanalytical methods for the determination of E2 using different electrochemical sensors reported in the literature.As shown in Table 1, the method proposed here presents similar or smaller LOD values compared to most works found in the literature.The significant advantage of the proposed method lies in its relatively low cost and simple synthesis of the material.
To evaluate the effect of possible interfering molecules found in the pharmaceutical and urine samples, binary solutions were prepared containing E2 with uric acid, lactose, urea, or magnesium stearate (at a ratio of 1:10 (m/m) E2:interfering molecule).Additionally, solutions containing E2 with a mixture of interfering molecules (uric acid + urea or lactose + magnesium stearate) were studied at the same ratio of 1:10 (m/m) E2:interfering molecule mixture.As can be seen in Figure 7, the current intensity obtained for E2 in the presence of the interfering molecules was compared to the current obtained for E2 alone under the same experimental conditions.The relative errors ranged from 1.50% to 3.22%, indicating that the evaluated molecules did not interfere with the determination of E2 under the studied conditions.
It is important to emphasize that other estrogens such as estradiol, estrone, and estriol may also be present in biological and environmental samples [57].These hormones oxidize at the same potential as E2, making them potential interference species.Therefore, in the presence of other estrogens, it is only possible to determine the total estrogen concentration in the samples.
To evaluate the matrix effect of lake water and synthetic urine samples on E2 determination using α-Fe 2 O 3 -CNT/GCE, addition-recovery studies were conducted.For this purpose, the samples were spiked with 50 nmol L −1 of E2.The recovery values are presented in Table 2, and they were found to be 105% for the lake water sample and 100% for the synthetic urine sample.These results suggest that the proposed method is applicable for determining E2 in real urine samples.Moreover, it can also be employed for water analysis using samples containing a higher E2 concentration, or for water samples subjected to E2 preconcentration treatment, as no matrix effects were observed.Finally, the proposed electroanalytical method was used for the determination of E2 in commercial pharmaceutical samples.Two samples were evaluated, and the results obtained by the proposed method were statistically compared with the results obtained by the high-performance liquid chromatography (HPLC) method.The E2 concentrations found in the pharmaceutical samples for both methods are presented in Table 3.A low relative error (%) was observed for both samples, demonstrating the accuracy of the method.Additionally, the Student's t test was used to compare the results, yielding experimental t values of 1.2 for sample 1 and 1.6 for sample 2, both of which were lower than the critical t value (tcritical = 2.8).This indicates that there is no significant difference between the proposed method and the HPLC comparative method, at a 95% confidence level.Finally, the proposed electroanalytical method was used for the determination of E2 in commercial pharmaceutical samples.Two samples were evaluated, and the results obtained by the proposed method were statistically compared with the results obtained by the highperformance liquid chromatography (HPLC) method.The E2 concentrations found in the pharmaceutical samples for both methods are presented in Table 3.A low relative error (%) was observed for both samples, demonstrating the accuracy of the method.Additionally, the Student's t test was used to compare the results, yielding experimental t values of 1.2 for sample 1 and 1.6 for sample 2, both of which were lower than the critical t value (t critical = 2.8).This indicates that there is no significant difference between the proposed method and the HPLC comparative method, at a 95% confidence level.
Solutions were prepared using ultrapure water (18.2MΩ cm) from an ELGA ® PURE-LAB Maxima (Woodridge, IL, USA) purification system.Pharmaceutical tablet samples were purchased from a local drugstore.E2 stock solution was prepared in ethanol at 0.1 mol L −1 concentration, then, proper dilutions in ultrapure water were performed.For the preparation of the BR buffer solution (0.01 mol L −1 ), appropriate amounts of boric acid, phosphoric acid, and acetic acid were mixed and dissolved in ultrapure water.The phosphate buffer solution was prepared by weighing adequate amounts of KH 2 PO 4 and Na 2 HPO 4 and dissolving them in ultrapure water.The 0.1 mol L −1 KCl solution was prepared by dissolving its salts in ultrapure water.The pH of phosphate and BR buffer and KCl solution was adjusted using 2.0 mol L −1 NaOH solution and 1.0 mol L −1 HCl solution.All the solutions were stored in a refrigerator until use.

Apparatus
Voltammetric measurements were carried out using a potentiostat/galvanostat Palm-Sens 2.0 (PalmSens BV, Houten, Netherlands) driven by PStrace 5.3 software using a conventional three-electrode electrochemical cell containing the bare GCE or the α-Fe 2 O 3 -CNT/GCE as working electrodes, an Ag/AgCl (3.0 mol L −1 KCl) as reference electrode, and a platinum wire as auxiliary electrode.For the morphological and chemical characterization a scanning electronic microscope (model FEI Quanta 200) and a DeltaNu spectrometer were used.The determination of E2 by a comparative HPLC method was carried out using a high-performance liquid chromatograph (model Shimadzu LC-20AD) equipped with a UV-Vis detection system using a detector of diode array (purchased in Tokyo, Japan), a C18 column (model Phenomenex) at 250 mm × 4.5 mm in dimension and particle size of 5 µm.
The quantification of E2 was performed according to the procedure described by Yilmaz and Kadioglu (2013) [58] with some modifications: a mobile phase in isocratic mode consisting of methanol and water (70:30 v/v), a flow rate of 1.0 mL min −1 , injection volume of 20.0 µL, UV-Vis detection at 220 nm.Before injection in the chromatographic system, the samples and standards were filtered through a 0.25 µm PTFE membrane filter.The pH control in the preparation of solutions was performed using a Bench Top Water pH meter (model AZ86505).

Synthesis of α-Fe 2 O 3 -CNT and Preparation of Electrochemical Sensors
The synthesis of α-Fe 2 O 3 nanoparticles supported by carbon nanotubes was achieved by mixing 3.0 mg mL −1 of CNT, 3.0 g of Fe 2 O 3 , and subsequently adding 6.0 g of NaNO 3 , 0.45 g of PVP, and 80.0 mL of ultrapure water.The mixture was then stirred for 30 min.Subsequently, it was transferred to a Teflon autoclave, and the hydrothermal process was carried out at 160 • C for 12 h.After the reaction, the solid product was filtered, washed with ethanol and ultrapure water, dried in an oven at 80 • C for 12 h, and finally calcined using a muffle furnace at 600 • C for 3 h, resulting in the formation of the metallic oxide [59].
The α-Fe 2 O 3 -CNT suspensions were prepared at concentrations of 1.0 mg mL −1 and 2.0 mg mL −1 .In both suspensions, the polyelectrolyte dihexadecyl hydrogen phosphate (DHP) was added at a concentration of 1.0 mg mL −1 , and the mixture was dispersed in ultrapure water.The CNT suspension was prepared using 1.0 mg mL −1 of CNT and 1.0 mg mL −1 of DHP, also in ultrapure water.To achieve homogeneity, the suspensions were subjected to 2 h of ultrasonic treatment and 1 h of magnetic stirring.The polyelectrolyte DHP acts as a surfactant and forms a stable film on the electrode surface when dispersed in water under ultrasonic stirring [60].
Before the modification procedure, the surface of the GCE was polished using 3.0 µm alumina suspension.After washing, a second polishing was performed using fine-grit sandpaper and water.Subsequently, the electrode was subjected to a 3-min ultrasonic bath in 2-propanol, followed by another 3-min ultrasonic bath in ultrapure water.The cleaning and conditioning process was completed by carrying out an electrochemical treatment in a sulfuric acid solution with a concentration of 0.01 mol L −1 using the CV technique.The treatment involved 50 cycles in a potential range of −0.5 to 1.2 V and a scan rate of 100 mV s −1 .The GCE was modified by drop-casting 6 µL of the aqueous suspensions of CNT or α-Fe 2 O 3 -CNT (at concentrations of 1.0 or 2.0 mg mL −1 ) onto its surface.Subsequently, the modified electrode was left to dry at room temperature for a minimum of 2 h.

Preparation of Lake Water, Synthetic Urine, and Pharmaceutical Samples
Natural water samples were collected at Igapó Lake in Londrina, Paraná, Brazil, using a clean amber glass bottle.Then, the samples were filtered through qualitative filter paper and used in the preparation of a 0.1 mol L −1 KCl solution (pH = 7.0), which was used as a supporting electrolyte.Aliquots of E2 were added to the water sample and quantified by the addition and recovery test.
The synthetic urine sample was prepared according to the procedure described by Laube, Mohr and Hesse [61] as follows: 0.110 g of CaCl 2 •2H 2 O, 0.296 g of NaCl, 0.225 g of Na 2 SO 4 , 0.140 g of KH 2 PO 4 , 0.160 g of NH 4 Cl, and 2.5 g of urea were dissolved in 100 mL of ultrapure water.For analysis, 1.0 mL of synthetic urine sample was diluted in 19 mL of 0.1 mol L −1 KCl solution (pH = 7.0) with 10% ethanol.Aliquots of E2 were added to the urine sample and quantified by the addition and recovery test.
For the pharmaceutical sample preparation, ten tablets were weighed and ground in a mortar with a pestle; then, the amount of one tablet was weighed and diluted in 5.0 mL of ethanol, then for E2 determination was used an external calibration method.

Measurement Procedure
CV and SWV were employed to investigate the electrochemical behavior and the quantification of E2.The instrumental parameters for SWV were optimized, and the respective analytical curve was obtained by adding small volumes of concentrated standard solutions of the E2 to the supporting electrolyte solution.The limit of detection (LOD) value was calculated as three times the standard deviation for 10 measurements of the blank solution (s) divided by the slope of the respective analytical curve (b) (LOD = 3 s/b) [62].The repeatability of the electroanalytical methods was checked with intra-day (n = 10) and inter-day (n = 5) determinations for two different concentrations of E2, for which the respective relative standard deviations (RSDs) were calculated.
The selectivity of the proposed methods was evaluated by the addition of possible interferents present in pharmaceutical formulations and urine samples (uric acid, urea, lactose, starch, povidone) to a standard solution containing E2, in the concentration ratios (standard solution to interferent) of 10, 1, and 0.1.

Conclusions
This work describes the successful development of an electrochemical method based on a glassy carbon electrode modified with α-Fe 2 O 3 nanoparticles supported on CNTs.We explored the high electroactive surface area of CNTs and the high conductivity of α-Fe 2 O 3 nanoparticles.The synthesis of the α-Fe 2 O 3 -CNT nanocomposite was achieved through a simple and cost-effective hydrothermal process.The α-Fe 2 O 3 -CNT/GCE provided an increase in current intensity obtained for E2 electrochemical oxidation and prevented fouling on the electrode surface when compared with a bare GCE.After optimizing the experimental conditions and SWV technique parameters, E2 determination was performed within a linear concentration range, and a LOD in the nanomolar order was achieved.
The proposed α-Fe 2 O 3 -CNT/GCE proved to be a simple and fast method for E2 determination in lake water, synthetic urine, and pharmaceutical samples, with good precision, accuracy, sensitivity, and no matrix effect or interference from other molecules.It can be a less expensive alternative for routine determinations of these drugs.

Figure 7 .
Figure 7. Bar diagram of the interference effect study for E2 determination.

Figure 7 .
Figure 7. Bar diagram of the interference effect study for E2 determination.

Table 1 .
Comparison of results obtained for determination of E2 by the here-proposed method and by other electrochemical methods reported in the literature.

Table 1 .
Comparison of results obtained for determination of E2 by the here-proposed method and by other electrochemical methods reported in the literature.

Table 3 .
Results obtained for E2 determination in commercial pharmaceutical samples by the proposed SW voltammetric method and the high-performance liquid chromatography (HPLC) comparative method.

Table 3 .
Results obtained for E2 determination in commercial pharmaceutical samples by the proposed SW voltammetric method and the high-performance liquid chromatography (HPLC) comparative method.