A highly sensitive non-enzymatic hydrogen peroxide and hydrazine electrochemical sensor based on 3D micro-snowflake architectures of α-Fe₂O₃

Abstract

In the present work, well crystalline 3D micro-snowflake structured α-Fe₂O₃ has been successfully synthesized on a large scale via a simple hydrothermal reaction by hydrolysis of a K₃Fe(CN)₆ precursor. The structure, composition, purity and morphology of the synthesized α-Fe₂O₃ samples are examined using powder X-ray diffraction (PXRD), field emission scanning electron microscopy (FESEM), transmission electron microscopy (TEM), energy dispersive X-ray spectroscopy (EDS), Fourier transform infrared (FTIR) spectroscopy and Mössbauer spectroscopy. The FESEM and TEM images reveal that the sample exhibits a micro-snowflake like shape having six-fold symmetry with symmetric branching along each arm consisting of a long central trunk and secondary branches. The 3D micro-snowflake structured α-Fe₂O₃ embedded ITO electrode exhibits high selectivity and sensitivity for electrochemical probing of hydrogen peroxide (H₂O₂) and hydrazine (N₂H₄) with a very low detection limit in a wide linear range. Amperometric measurements show a sensitivity of 7.16 μA mM⁻¹ cm⁻² in a wide linear range from 0.1 to 5.5 mM with the lowest detection limit of 0.01 mM (S/N = 3) towards H₂O₂ sensing. The sample also exhibits a sensitivity of 24.03 μA mM⁻¹ cm⁻² in the linear range between 50 μM and 1340 μM with the lowest detection limit of 5 μM towards hydrazine detection. The excellent electrochemical activity of the sample is rendered to the presence of a large number of catalytic sites in the sample due to its 3D micro-snowflake like architecture. Good reproducibility, stability and selectivity suggest its suitability for the fabrication of H₂O₂ and hydrazine sensors.

---

1. Introduction

Recently, iron based nanoparticles have emerged as an area of considerable interest owing to their striking nanoscale physicochemical properties which are markedly different from their bulk counterparts.^1–9 Equipped with the advantages of nanoscale materials (small size, high surface to volume ratio, size and shape dependent properties), iron based nanoparticles have been successfully used in diverse technological fields mainly due to their novel optical, electronic and magnetic properties. These materials are profitably used in magnetic data storage devices, electronic devices and magnetic resonance imaging, and also as optical sensors, gas sensors, magnetically recoverable catalysts and photo-catalysts.^1–5,10 In this context, it is pertinent to mention that iron oxide (Fe₃O₄ and α-Fe₂O₃) nanoparticles are biocompatible, non-toxic, chemically stable, low cost, abundant materials and have been successfully used in electrochemical sensing systems such as enzyme sensors, ethanol sensors, humidity sensors and DNA sensors.^11–15

Hydrogen peroxide (H₂O₂) is extensively used in clinical, biopharmaceutical, environmental and industrial fields.¹⁶ Oxidation of glucose produces H₂O₂.¹⁷ Thus, detection of H₂O₂ is of immense importance in the field of biological science because it helps in the diagnosis of diseases like diabetes.¹⁸ On the other hand, hydrazine (N₂H₄) and its derivatives are generally used in fuel cells, explosives, rocket propellants, insecticides and plant growth regulators.¹⁹ Although these two materials are applied in diverse technological sectors, it has been found that they have adverse effects towards health and the environment.¹⁹ H₂O₂ is a strong oxidizing agent which causes corrosion of eyes and skin and may cause irreversible cell damage. It can also trigger cell proliferation, which may culminate into various diseases like cancer, cardiovascular and neurodegenerative disorders.²⁰ Again, hydrazine is a carcinogenic, hepatotoxic and mutagenic material which may cause severe damage to the liver, kidneys and central nervous system.^19,20 Therefore, sensitive and accurate detection of a small quantity of H₂O₂ and N₂H₄ is of great scientific importance. Up to now, numerous techniques have been proposed for the fruitful detection of H₂O₂ and N₂H₄.^19,21,22 However, nonenzymatic amperometric sensing is the prime choice due to its high sensitivity, short response time and low cost of instrumentation. Furthermore, as nonenzymatic biosensors remove the impediments of enzyme-based sensors (like complex fabrication processes and low chemical/thermal stability), a lot of effort has been devoted to developing new enzyme free, stable and reproducible biosensors.²³

It may be noted that α-Fe₂O₃ nanostructures show size and shape dependent properties, which have triggered lots of research into the controlled synthesis of α-Fe₂O₃ nanostructures.^24–27 Up to now, a variety of α-Fe₂O₃ structures, such as 1D, 2D, and 3D have been developed. In 3D architectures α-Fe₂O₃ often exhibits more interesting electrical, optical, magnetic and catalytic properties, which are markedly superior compared to 2D or 1D architectures.^24–27 The α-Fe₂O₃ nanorod array displays excellent selectivity and stability for the electrochemical detection of H₂O₂,²⁸ whereas the α-Fe₂O₃ nanowire array exhibits excellent catalytic performance towards the oxidation of glucose.²³ The α-Fe₂O₃ nanorings have been found to be effective in sensing of H₂O₂.²⁹ Conversely, to the best of our knowledge there is no report on hydrazine sensing by α-Fe₂O₃ nano/micro structures. This issue has inspired us to synthesize a snowflake like 3D architecture of α-Fe₂O₃ and check its performance as a nonenzymatic H₂O₂ and N₂H₄ sensor.

In this paper, 3D micro-snowflake structured α-Fe₂O₃ has been synthesized using a facile one step solvothermal method. The sample has been thoroughly characterized by powder X-ray diffraction (PXRD), high resolution transmission electron microscopy (HRTEM), field emission scanning electron microscopy (FESEM), Fourier transform infrared (FTIR) spectroscopy and Mössbauer spectroscopy. The growth mechanism of the 3D micro-snowflake structured α-Fe₂O₃ has been followed by varying the reaction parameters. Further, detailed electrochemical analysis has been performed to study the electrocatalytic activity of the samples towards sensing of H₂O₂ and N₂H₄ in alkaline medium. It has been shown that the sample can be used as an amperometric probe for highly sensitive and selective detection of H₂O₂ and N₂H₄.

2. Materials and methods

2.1. Materials

Potassium ferricyanide (K₃[Fe(CN)₆]) was purchased from Loba Chemie and used without further purification for synthesis purposes. Other relevant reagents (potassium ferrocyanide (K₄[Fe(CN)₆]), potassium chloride (KCl), sodium hydroxide (NaOH), potassium hydroxide (KOH), hydrogen peroxide (H₂O₂), hydrazine (N₂H₄), glucose, ascorbic acid (AA), dopamine (DA), uric acid (UA), NaBr and NaCl) used in the experiments were also purchased from Loba Chemie and Mark Germany. Doubly distilled water was used as the solvent for all the experiments.

2.2. Synthesis of 3D micro-snowflake structured α-Fe₂O₃

0.2640 g of K₃[Fe(CN)₆] was dissolved in 80 mL deionized water. The aqueous solution was then transferred into a 100 mL Teflon-lined autoclave and the hydrothermal reaction was conducted at 180 °C for 24 (S1), 36 (S2), 48 (S3) and 96 (S4) hours. Then the autoclave was allowed to cool down to room temperature and the precipitates were washed several times in deionized water and ethanol by vigorous centrifugation. Afterward, the collected samples were dried at 70 °C for 10 hours in a vacuum oven. It may be noted that we have reported the H₂O₂ sensing property of the sample S1 in our earlier study.³⁰ Herein, the sensing property of sample S4 has been reported and the other samples have been synthesized only to study the morphological changes taking place with the increase of reaction time.

2.3. Characterization of micro-snowflake structured α-Fe₂O₃

The powder X-ray diffraction (PXRD) patterns of the samples were recorded using a Bruker D8 Advanced Diffractometer using Cu Kα (λ = 1.54184 Å) radiation at ambient temperature (21 °C). The in depth structural and microstructural studies have been performed by using high resolution transmission electron microscopic (JEOL, 2100) and field emission scanning electron microscopic (FEI, INSPECT F50) techniques. The presence of constituent elements in sample S4 was probed using a BRUKER EDS system attached to the FESEM equipment. The FTIR spectrum of S4 was recorded using a Perkin Elmer spectrometer (Spectrum Two) equipped with an attenuated total reflectance (ATR) attachment in the wave number range 500–4000 cm⁻¹. The room temperature Mössbauer spectrum of S4 was recorded in transmission geometry using constant acceleration drive (CMTE-250) with a 15 mCi ⁵⁷Co source in a Rh matrix. A natural iron sample was used for calibration. The Mössbauer spectrum of the sample was fitted using the Recoil program.³¹ The solid circles represent experimental data points, whereas the solid lines represent the simulated pattern.

2.4. Fabrication of working electrodes and electrochemical experiments

Electrochemical characterizations (cyclic voltammetry and amperometry) were performed to evaluate the electro-catalytic ability of S4 towards sensing of H₂O₂ and N₂H₄. Electrochemical impedance spectroscopy (EIS) was performed to evaluate the charge transfer properties of the working electrode at the electrode/electrolyte interface. All electrochemical characterizations were performed in a standard three electrode configuration, with α-Fe₂O₃ coated ITO (Indium Tin Oxide coated glass) as the working electrode, Pt wire as the counter electrode and Ag/AgCl as the reference electrode. To prepare the working electrode, firstly the ITO sheets (1 × 1 cm²) were cleaned by ultrasonication in distilled water, ethanol and acetone for 10 minutes each and subsequently dried in air. 10 mg of S4 was dispersed in chitosan solution (1 wt% in acetic acid) and ultrasonicated for 1 hour to achieve a homogeneous suspension. 10 μL of the suspension was dropped on to the ITO coated glass and dried overnight for further measurements. The cyclic voltammetry (CV) and amperometry studies were performed in 0.1 M NaOH aqueous solution. EIS measurements were carried out in an electrolyte comprising of 5 mM of K₄Fe(CN)₆/K₃Fe(CN)₆ (1 : 1 mixture) as the redox probe with 0.1 M KCl as the supporting electrolyte.

3. Results and discussion

3.1. Structural study

In the powder X-ray diffraction (PXRD) patterns of all the samples (S1, S2, S3 and S4) only the characteristic diffraction peaks for the α-Fe₂O₃ phase (JCPDS no. 33-0664) have been observed and no additional peaks due to possible impurity phases have been noticed. This confirms that S1, S2, S3 and S4 are single phase α-Fe₂O₃. It is well known that the Rietveld refinement of powder X-ray patterns is a trustworthy tool for structural and microstructural characterization of nanometric oxides.^1,5,6 It may be noted that α-Fe₂O₃ has a rhombohedrally centered hexagonal structure with a close-packed oxygen lattice in which two thirds of the octahedral sites are occupied by Fe³⁺ ions.³² To determine the crystal structure of S4 precisely we have fitted the PXRD pattern of the sample using the Rietveld refinement method using the GSAS program with a EXPGUI interface.^33,34 The results are presented in Fig. 1. The detail of the methodology is reported in our earlier works.^1,5,6 The final Rietveld refinement converged with an excellent agreement between the experimental and simulated patterns. The values of R_P, R_WP and χ² are 0.02, 0.025 and 0.72, respectively. The results indicate that the sample is well crystalline α-Fe₂O₃ and it crystallizes in the hexagonal R3̄c space group. The miller indices and the space group obtained for the sample are in good agreement with the JCPDS database (JCPDS no. 33-0664). The values of the structural and microstructural parameters, fractional coordinates and occupancy of different ions are listed in Tables 1 and 2.

Fig. 1 (a) The final Rietveld plot (output file of GSAS) of the α-Fe₂O₃ sample, showing the difference (blue color line) between the experimental (black color line) and the simulated pattern (red color line). (b) Unit cell of the sample.

Table 1 Structural parameters of the sample obtained using the Rietveld refinement method

Crystal system	Space group	Lattice parameter (Å)	Volume V [Å^3]	Crystalline size (nm)	Microstrain
Hexagonal	R3̄c	a = b = 5.03564(4), c = 13.7589(2)	302.151(0.01)	18.2(0.1)	3 × 10^−3 (2 × 10^−5)

---

Table 2 Fractional coordinates and occupancy of different ions obtained using the GSAS program

Ions	x	y	z	Occupancy (±0.003)
Fe (1)	0	0	0.14451	0.025
Fe (2)	0	0	0.35530	0.021
O (1)	0.3051	0	0.25	0.025
O (2)	−0.3061	−0.3061	0.25	0.024

---

3.2. FESEM, EDS and FTIR study on S4

The microstructural property and surface morphology of S4 has been investigated using field emission scanning electron microscopy (FESEM). The FESEM images of S4 both at high and low magnification are depicted in Fig. 2. The FESEM images clearly establish the formation of 3D micro-snowflake structures with dendritic arms. The individual dendritic arms consist of a long central trunk with a length of ∼5–6 μm with secondary branches with a length of ∼1–2 μm. The branches are uniformly distributed on the both sides of the central trunk.

Fig. 2 FESEM images of the 3D micro-snowflake structured α-Fe₂O₃ sample.

The presence of the constituent elements in the sample has been verified using an energy dispersive X-ray study (EDS). The EDS spectrum, illustrating the well resolved peaks originating from the constituent atoms in the energy range from 0 to 8 keV, is shown in Fig. S1.† The Au Lα peak in the EDS spectrum can be attributed to the gold coating of the sample prior to the EDS study. The EDS spectrum clearly indicates the presence of all the constituent elements (Fe, O) of α-Fe₂O₃.

The FTIR spectrum of the sample is shown in Fig. S2.† The strong peak at about 664 cm⁻¹ appears due to the stretching vibrational mode of the Fe–O bond.¹ The broad absorption bands centered around 3400 and 1633 cm⁻¹ are assigned to the stretching vibration of O–H bonds of the free or absorbed water molecules in the sample.³⁵

3.3. Mössbauer spectroscopic study

Magnetic ordering and hyperfine properties of iron containing materials can be easily investigated with the help of ⁵⁷Fe Mössbauer spectroscopic studies.^1,2,5,6,36 The room temperature Mössbauer spectrum (Fig. S3†) of S4 has been fitted using the Lorentzian site analysis method of the Recoil program. The Mössbauer spectrum at 300 K is best fitted by a single sextet with the value of the isomer shift (IS) and quadrupole splitting (QS) at 0.46 and 0.004 mm s⁻¹, respectively. The value of the hyperfine field obtained from the fitting is 58.5 T. The value of the IS suggests that only Fe³⁺ ions are present in the sample.^5,6 It may be noted that the hyperfine field of the present sample is substantially higher compared to its counterparts reported earlier.^24,25 These provide clear evidence in favour of the presence of only a α-Fe₂O₃ phase in S4 and no traces of γ-Fe₂O₃ or Fe₃O₄ have been found in the Mössbauer spectrum of the sample.

3.4. Growth mechanism of 3D micro-snowflake structured α-Fe₂O₃

It is well known that the synthesis route, reaction parameters and use of growth controlling agents (surfactants) play a major role in determining the shape, size, porosity and specific surface area of the synthesized anisotropic complex 3-dimensional architectures of α-Fe₂O₃.²⁴ To develop a standard, simple, reliable and high yielding reaction mechanism for the synthesis of complex architectures of α-Fe₂O₃ is a big challenge. Up to now, many types of complex architectures of α-Fe₂O₃ such as nanoparticles, nanocages, cantaloupe like structures, hollow spheres, dendrites and snowflakes have been synthesized by different chemical routes.^12,23–27

Wang et al. have described a one pot hydrothermal route for the synthesis of a single crystalline dendritic micropine structure of α-Fe₂O₃.²⁵ Recently, Bharathi et al. have shown that a hydrothermal reaction without surfactant leads to the formation of 3-D dendritic α-Fe₂O₃ structures, whereas the use of CTAB and PEG as surfactants during the reaction results in single and double layered snowflake structures of six-fold symmetry, respectively.²⁴ According to Zhen et al., NP-9 surfactant converts dendritic micropines of α-Fe₂O₃ to hexagonal micro-snowflakes.³⁷ Liu et al. have shown how the morphology of α-Fe₂O₃ changes from snowflakes to starfish then flowers and finally to hexagonal plates upon increasing the concentration of octylamine.³⁸ Sun et al. prescribed a hydrothermal reaction in the presence of benzoic acid for precise tailoring of α-Fe₂O₃ dendritic structures.³⁹ The influence of reaction time and pH of the hydrothermal chamber on the growth of different α-Fe₂O₃ structures have been studied by Zhang et al.⁴⁰ They have shown that in a basic pH (∼12), the dendritic α-Fe₂O₃ structures convert into snowflake like microstructures with enhanced reaction time. Further increase in the reaction time promotes more crystal growth and forms paired microplates and dumbbell-like or spindle-like microstructures.⁴⁰

In the present case we have synthesized a 3D architecture of micro-snowflake structured α-Fe₂O₃ from K₃[Fe(CN)₆] using a hydrothermal technique in normal pH without using any surfactant. The reaction time has been varied to study the evolution of the structure and morphology of the sample. We have carried out the hydrothermal reaction for different time periods (24 (S1), 36 (S2), 48 (S3) and 96 (S4) hours) by maintaining the reaction temperature at 180 °C. The FESEM micrographs of the samples are given in Fig. 3. The FESEM micrograph of S1 suggests that this sample possesses a 3D dendritic micropine like structure with a single symmetric branched arm. This structure consists of a single central trunk with secondary branches on both sides of the central trunk. The morphology of S2 is similar to that of S1. However, S2 has more fractal branches than S1. This suggests that fractal branch formation increases with the increase of reaction time (from 24 to 36 hours). The FESEM images of S2, S3 and S4 suggest that with the increase of reaction time the micropine like structures get attached to form micro-snowflakes of α-Fe₂O₃ (sample S3) with symmetric branching along each arm. This micro-snowflake structure has six-fold symmetry.

Fig. 3 Schematic diagram of the formation of 3D micro-snowflake structured α-Fe₂O₃ (above) and FESEM images of the synthesized samples S1, S2, S3 and S4 (below).

Faster crystal growth rate along one of the six crystallographic directions compared to the other five crystallographic directions results in the formation of the dendritic micropine like structure of α-Fe₂O₃. The TEM micrograph of S4 (Fig. 4) shows a clear and well defined dendritic fractal structure with a prolonged trunk with ordered branches on both sides of it. The length of the main trunk is 5.37 μm, whereas the length of the branches range from 1.25–1.6 μm. Actually, the formation of desired microstructures of α-Fe₂O₃ depends upon the dissociation rate of [Fe(CN)₆]³⁻ ions in the hydrothermal reaction. S. Bharati et al. have shown that the use of metal salts in the hydrothermal reaction, which readily dissociate in aqueous solution (FeCl₃, Fe(NO₃)₃), results in irregular shaped nanoparticles.²⁴ On the contrary, the [Fe(CN)₆]³⁻ ion is highly stable at room temperature. Hydrothermal treatment of the aqueous solution of K₃[Fe(CN)₆] at a high temperature (180 °C) for an extended time results in the dissociation of [Fe(CN)₆]³⁻ ions into Fe³⁺ ions, which upon further hydrolysis get converted into FeOOH/Fe(OH)₃. Due to the high temperature hydrothermal reaction the FeOOH/Fe(OH)₃ further dissociates into α-Fe₂O₃.^24,25

Fig. 4 (a) TEM micrograph of each symmetric branch of micro-snowflake structured α-Fe₂O₃, (b) TEM micrograph of micro-snowflake structured α-Fe₂O₃, (c) HRTEM showing atomic planes with 0.25 nm spacing, and (d) SAED pattern of the sample.

α-Fe₂O₃ crystallizes in a hexagonal closed pack structure. The Fe³⁺ and O²⁻ ions arrange alternately in a regular fashion, parallel to the c direction, i.e., [0001] direction. It may be noted that each layer is terminated with Fe³⁺ and O²⁻ ions, thus developing an ionic charge on the surface. The c plane, being a polar surface, neutralizes the ionic charge by adsorbing molecules present in the hydrothermal chamber and the growth along the [0001] direction gets terminated. On the other hand, growth proceeds along the other six crystallographically equivalent directions ±[101̄0], ±[11̄00], and ±[01̄10] which give rise to the formation of hexagonally shaped particles.^24,25 The stacking of α-Fe₂O₃ nanoparticles along the [11̄00] direction leads to the formation of the main trunk, while the formation of branches can be assigned to the slower crystallization rate along the equivalent [101̄0] and [01̄10] directions. By extending the reaction time to 48 hours, the complete snowflake like structure of α-Fe₂O₃ appears.^24,25 The prolonged reaction time promotes attachment to form the micro-snowflake structure from micropines and this gives rise to a single-layered snowflake like structure. The HRTEM image (Fig. 4(c)) taken at the periphery of the branch shows clear lattice fringes, which indicate that the sample is crystalline in nature. The lattice spacing of 2.5 Å corresponds to the {112̄0} crystal planes. The selected area electron diffraction (SAED) pattern of S4 (Fig. 4(d)) clearly indicates that the sample is single crystalline in nature. The central spot of the SAED pattern corresponds to [0001] plane. The SAED pattern suggests that the branches are formed along the six crystallographically equivalent directions ±[101̄0], ±[11̄00], and ±[01̄10].

4. Electrochemical sensing study

4.1. Electrochemical activity and EIS study

The electrocatalytic activity of S4 was evaluated using both cyclic voltammetry and EIS techniques. It may be noted that the intensity of the oxidation and reduction peak markedly increases (the current density at the oxidation potential for ITO and S4 coated ITO are 100 and 200 μA cm⁻², respectively) and the difference between the oxidation and reduction potential decreases (the redox potential for ITO and S4 coated ITO are 0.8 and 0.075 V, respectively) upon coating the ITO substrate with S4 (Fig. 5). This was further confirmed by examining the interfacial charge transfer resistance at the electrode-electrolyte interface using EIS. Fig. 6 displays the electrochemical impedance spectra of bare ITO and S4 coated ITO in the frequency range between 0.1 Hz and 100 kHz. The plot consists of a semicircular arc in the high frequency region followed by a linear part. The diameter of the semicircular region is a measure of the charge transfer resistance at the electrode/electrolyte interface and the linear part corresponds to a diffusion limited process. It is evident from this figure that the diameter of the semicircular arc in the impedance spectrum of the S4 coated ITO working electrode is smaller than that of bare ITO and this in turn indicates that the interfacial resistance at the electrode-electrolyte interfaces for the S4 coated ITO working electrode is less than that of bare ITO. Thus, the S4 coated ITO working electrode provides a lower interfacial resistance in the path of charge transport compared to that of bare ITO and it favors the movement of charges through it. It seems that S4 coated ITO may show good electrocatalytic activity.

Fig. 5 Cyclic voltammetry in 5 mM K₃Fe(CN)₆/K₄Fe(CN)₆ redox probe (1 : 1) with 0.1 M KCl as supporting electrolyte.

Fig. 6 Electrochemical impedance spectra (EIS) for both electrodes.

4.2. Electrochemical sensing of H₂O₂

Fig. 7 depicts the CV plot for the reduction of H₂O₂ at the S4 coated ITO working electrode conducted in 0.1 M NaOH solution as the electrolyte between the potential range of 0 and −1 V at a scan rate of 0.1 V s⁻¹. According to an early report, a sharp reduction peak appears at −0.82 V in the CV plot of a α-Fe₂O₃ nanorod coated stainless steel (SS) substrate with 1 M KOH as the electrolyte due to the reduction of α-Fe₂O₃.⁴¹ Thus, in the CV plot of the bare S4 embedded ITO electrode the peak at −0.8 V can be attributed to the reduction of α-Fe₂O₃. It is noteworthy that with increasing molar concentrations of H₂O₂ from 0.1 to 0.2 mM, the current density at the reduction potential of −0.8 V increases markedly compared to that of the bare S4 coated ITO electrode indicating the enhancement of the electrochemical response of S4 coated ITO upon addition of H₂O₂ in the electrolyte. Whenever H₂O₂ is added in an alkaline solution it is converted to either HO₂⁻ or O₂²⁻. In the present case the working electrode is actually detecting the presence of H₂O₂ in the electrolyte (NaOH) by sensing the subsequent reduction of these anionic byproducts. The α-Fe₂O₃ coating on the working electrode may therefore be identified as an electrocatalyst responsible for the indirect detection of H₂O₂. The effect of scan rate on the reduction of 0.1 mM H₂O₂ is shown in Fig. S4.† The inset of Fig. S4† shows that the current density at the reduction potential increases linearly with scan rate, thereby indicating a surface confined process.⁴²

Fig. 7 Cyclic voltammetry (CV) study with different molar concentrations of H₂O₂.

Amperometric current density (J) versus time (t) (J–t) measurements were performed to evaluate the H₂O₂ sensing performance of S4 coated ITO at the reduction potential of −0.8 V (Fig. 8). A H₂O₂ stock solution of different molar concentrations was added successively to a homogeneously stirred NaOH solution at regular intervals of 50 s. In this experiment stable current steps were obtained as shown in Fig. 8. The corresponding calibration curve shown in the inset of Fig. 8 can be fitted with the equation J = 7.16 μA mM⁻¹ cm⁻²C_H2O2 − 4.8968 with a correlation coefficient R² = 0.99434. The value of sensitivity calculated from the slope of the calibration plot is obtained as 7.16 μA mM⁻¹ cm⁻². Further, the lowest detection limit was estimated to be 0.01 mM at S/N = 3.

Fig. 8 Amperometric J–t curve of the sample for the H₂O₂ sensing study, linear response of H₂O₂ concentration with the current values (inset).

The analytical performance (sensitivity, detection limit and linear response range) of the 3D micro-snowflake structured α-Fe₂O₃ is compared with previously reported H₂O₂ sensors and the results are shown in Table 3. From Table 3 it is evident that the sensitivity of the proposed H₂O₂ sensor is the second best.^{17,18,20,43–59} Moreover, according to our previous work³⁰ the sensitivity of sample S1 is 0.8 μA mM⁻¹ cm⁻² for H₂O₂ sensing. But in these 3D micro-snowflake structured α-Fe₂O₃ architectures the higher sensitivity value of the system can be attributed to the additional catalytic sites available due to the three dimensional morphology of the structure resulting in favorable dissociation of H₂O₂ leading to higher current density and improved sensitivity.

Table 3 Comparison of the analytical performance of the proposed H₂O₂ biosensor with other H₂O₂ biosensors reported previously

Electrode	Sensitivity	Linear range	Detection limit	Reference
FeS nanosheet/GC	—	0.5–150 μM	9.2 × 10^−8 M	43
NiFe2/OMC	4.29 μA mM^−1 cm^−2	6.2–42 710 μM	0.24 μM	17
Pt/CNT	1.4 μA mM^−1 cm^−2	5.0–25 000 μM	1.5 μM	44
Pt/OMC/Au	6.83 μA mM^−1 cm^−2	—	42 μM	45
OMC/Au	0.071 μA mM^−1 cm^−2	—	0.73 μM	45
Mesoporous Pt	2.8 μA mM^−1 cm^−2	20–40 000 μM	4.5 μM	46
Ag@CuNW	—	1–10 mM	3 μM	47
MnO2 microspheres	—	10–150 mM	2.0 μM	48
Mn2O3 hollow spheres	—	0.1–126.5 μM	0.07 μM	49
Ag NP/MB biocomposite	0.357 μA μM^−1 cm^−2	1 μM to 3 mM	0.088 μM	50
Ag/SnO2/GCE	—	0.01 to 35 mM	5 μM	51
Ag/RGO/ITO	—	0.1 to 100 mM	1.6 μM	52
PB-graphene/GCE	0.53 μA μM^−1 cm^−2	10–1440 μM	3 μM	53
PANI–PS–sulphonate-carboxylated graphene modified graphite	201.4 nA μM^−1 cm^−2	25–350 μM	1 μM	54
Fe2O3/GC	2.5 μA μM^−1 cm^−2	10–150 μM	7 μM	55
Au NPs–N–GQDs	182.22 μA mM^−1 cm^−2	0.25–13 327 μM	0.12 μM	56
Ta/Fe–Phen–CFS	—	0.1–100 μM	68 nM	20
Fe3C/NG	9.44 (μA L^−1) mM^−1	50 μM L^−1 to 15 mM L^−1	0.035 Mm L^−1	18
MnO2/CF2	0.054 μA μM^−1 cm^−2	2.5–2055 μM	0.12 μM	57
Co3O4–NWs/CF	230 nA μM^−1 cm^−2	0.01–1.4 mM	1.4 μM	58
Cu2O/GNs	—	0.3–7.8 mM	20.8 μM	59
3D micro-snowflake structured α-Fe2O3	7.16 μA mM^−1 cm^−2	0.1–5.5 mM	0.01 mM	This work

---

4.3. Electrochemical hydrazine sensing

Electrochemical hydrazine (N₂H₄) sensing studies were conducted in 0.1 M NaOH electrolyte. Fig. 9 shows the comparative CV in the absence and presence of 10 mM N₂H₄. Electro-oxidation of N₂H₄ was observed at 0.1 V in the background NaOH electrolyte. The cyclic voltammograms of 10 mM N₂H₄ at different scan rates using S4 coated ITO as the working electrode are depicted in Fig. S5.†

Fig. 9 Comparative CV in the presence and absence of 10 mM hydrazine.

The amperometric studies for N₂H₄ were performed at the operating potential of 0.1 V corresponding to the oxidation peak current in the CV and the results are presented in Fig. 10. In this amperometric J–t curve stable steps are observed upon successive addition of 50 μM N₂H₄ (Fig. 10). The current density versus concentration curve (calibration plot) is linear in nature and fitted by the regression equation J = 0.02403C_hydrazine − 1.19176 with correlation coefficient R² = 0.9843. The value of the sensitivity obtained from the calibration curve is 24.03 μA mM⁻¹ cm⁻² and the lowest detection limit is estimated to be 5 μM at S/N = 3. The limit of detection (LOD) value is quite impressive in comparison with other noble metal based N₂H₄ sensing platforms. The amperometric measurements remain stable despite the evolution of N₂ indicating a satisfactory response for hydrazine detection. The analytical performance (sensitivity, detection limit and linear response range) of the 3D micro-snowflake structured α-Fe₂O₃ was compared with previously reported hydrazine sensors and is presented in Table 4. From Table 4 it is clear that the proposed N₂H₄ sensor shows excellent electrochemical activity compared to previously reported hydrazine sensors.^60–68

Fig. 10 Amperometric J–t curve of the sample for the hydrazine sensing study, and linear response of hydrazine concentration with the current values.

Table 4 Comparison of the analytical performance of the proposed hydrazine biosensor with other hydrazine biosensors reported previously

Electrode	Sensitivity	Linear range (μM)	Detection limit (μM)	Reference
Ag nanocubes/GCE	—	5–460	1.1	60
Pd–TiO2	0.55 mA mM^−1 cm^−2	1000–20 000	23	61
CuS/RGO	7.96 μA mM^−1	1–1000	0.3	62
Graphene-Bi/GCE	1.49 μA mM^−1 cm^−2	0.02–280	0.005	63
Pd/carbon black/GCE	—	0–500	13.4	64
Au NPs/graphite	—	25–1000	3.07	65
Au NPs/ZnO-MWCNT/GCE	0.0428 μA μM^−1 cm^−2	0.5–1800	0.15	66
Au NPs/poly(BCP)/CNT/GCE	—	0.5–1000	0.1	67
Ag dendritics/ITO	20.81 μA mM^−1 cm^−2	100–1700	0.5	68
3D micro-snowflake structured α-Fe2O3	24.03 μA mM^−1 cm^−2	50–1340	5	This work

---

4.4. Interference study

The most important consideration for any sensing platform is its selectivity in the presence of other coexisting electroactive species. The same was tested for H₂O₂ reduction and N₂H₄ oxidation in the present case and the results are shown in Fig. 11(a) and (b), respectively. The H₂O₂ reduction selectivity was investigated in the presence of ascorbic acid (AA), dopamine (DA), uric acid (UA) and glucose solutions. The amperometric J–t plot indicates that the α-Fe₂O₃ coated ITO electrode shows a high response towards H₂O₂ and almost negligible response with the addition of AA, DA, UA and glucose. The selectivity of the α-Fe₂O₃ coated ITO electrode towards H₂O₂ sensing may be due to the fact that when AA, DA, UA and glucose are added in the NaOH electrolyte either no anionic species are evolved in the solution, as in the case of H₂O₂, or there is no reduction peak for AA, DA, UA and glucose at about the working potential of −0.8 V. The selectivity of N₂H₄ oxidation was evaluated in the presence of AA, DA, UA, glucose, Br⁻, Cl⁻ and K⁺. From the amperometric J–t plot, it can be inferred that the response of the α-Fe₂O₃ coated ITO electrode towards N₂H₄ oxidation is highly selective with almost no change in the current density upon the addition of glucose, Br⁻, Cl⁻ and K⁺. However, a slight increase in current has been observed for the addition of AA which may be ascribed to the tendency of AA to oxidize at low potentials. The experimental results suggest that the tolerance level of α-Fe₂O₃ microstructures for Br⁻, Cl⁻ and K⁺ ions is high.⁵⁹ It may be noted that the oxidation of DA, UA and glucose usually takes place at higher potential.⁵⁹ As we have carried out the measurements at a low working potential (0.1 V) the oxidation of DA, UA and glucose does not interfere with amperometric response of N₂H₄. The cause behind the selectivity of the α-Fe₂O₃ coated electrode for the oxidation of N₂H₄ can also be due to the reason that the oxidation of N₂H₄ in alkaline medium is a kinetic-controlled reaction while the oxidation of the other interferences used is a diffusion-controlled electrochemical process.⁵⁹ If these are facts, then the better performance of N₂H₄ oxidation can be attributed to the presence of a large number of active catalytic sites due to the large surface area of the 3D snowflake architecture.

Fig. 11 Selectivity in the presence of (a) H₂O₂ and (b) hydrazine with other co-existing analytes.

4.5. Reproducibility and long term stability

The reproducibility of the sensing performance for both H₂O₂ and N₂H₄ was evaluated by recording the amperometric response of 5 identically prepared working electrodes. Fig. 12(a) and 13(a) show the plots of the sensitivity values for the electrodes towards H₂O₂ and N₂H₄ and the relative standard deviation for H₂O₂ and N₂H₄ are 1.62% and 1.81%, respectively. The long term stability was studied over a period of four weeks by performing the electrochemical tests after each week and the results for H₂O₂ and N₂H₄ are shown in Fig. 12(b) and 13(b), respectively. The study on H₂O₂ has revealed that the current density is 80% of its initial value after the first round of usage and drops to 60% of its initial value after four weeks. The N₂H₄ oxidation study also shows good long term stability with the oxidation peak current retaining 60% of its initial value at the end of four weeks. Hence, the S4 coated working electrode shows acceptable reproducibility and good long term stability towards the reduction of H₂O₂ and oxidation of N₂H_4.

Fig. 12 (a) The reproducibility of the sensing performance for H₂O₂, (b) reproducibility and good long term stability towards H₂O₂.

Fig. 13 (a) The reproducibility of the sensing performance for hydrazine, (b) reproducibility and good long term stability towards hydrazine.

5. Conclusion

Herein, we have successfully synthesized 3D micro-snowflake structured α-Fe₂O₃ by a simple hydrothermal reaction without using any surfactant and followed the growth mechanism with the variation of reaction time. We have employed PXRD, FESEM, TEM, FTIR spectroscopy and Mössbauer spectroscopy to characterize the sample. The crystal structure of S4 has been analyzed by Rietveld refinement of the PXRD pattern of the sample. FESEM and TEM images show that S4 exhibits a micro-snowflake like shape having six-fold symmetry with symmetric branching along each arm consisting of a long central trunk and secondary branches. The results of the PXRD, FTIR and Mössbauer spectroscopic studies together confirmed that this micro-snowflake like structure is formed by pure, crystalline and single phase α-Fe₂O₃. The sample displayed excellent electrochemical sensing performance towards H₂O₂ reduction (sensitivity 7.16 μA mM⁻¹ cm⁻² in a wide range from 0.1 mM to 5.5 mM) and N₂H₄ oxidation (sensitivity 24.03 μA mM⁻¹ cm⁻² in the linear range between 50 μM and 1340 μM).

In summary, we have shown that the 3D micro-snowflake structured α-Fe₂O₃ synthesized by simple hydrothermal decomposition of K₃[Fe(CN)₆] without using any surfactant can be used for highly selective, sensitive and stable amperometric sensing of H₂O₂ and N₂H₄ in the presence of common coexisting electroactive interferences. We have examined how different morphological forms evolve for microstructured α-Fe₂O₃ synthesized by this hydrothermal route with the variation of reaction time by keeping the temperature and reactant concentration the same. We have shown that the morphology of α-Fe₂O₃ can be gradually tailored from micropines to micropines with fractal branches and finally to 3D micro-snowflake structures simply by changing the reaction time without using any growth controlling agent (surfactant). It may be noted that the electrochemical performance of the S4 modified ITO working electrode for H₂O₂ sensing is second best among all the H₂O₂ sensing electrodes reported to date. Moreover, to the best of our knowledge, the electrochemical N₂H₄ sensing activity of α-Fe₂O₃ is explored for the first time. As an ideal enzyme-less sensing material, the sample has good stability and selectivity against common coexisting interferences. Hence, α-Fe₂O₃ coated ITO working electrodes could be suitable platforms for enzyme-less H₂O₂ and N₂H₄ sensing.

Acknowledgements

Authors (S. Majumder and B. Saha) gratefully acknowledge UGC, Govt. of India for providing research fellowship. S. Dey and R. Mondal gratefully acknowledge CSIR, Govt. of India and DST, Govt. of India, respectively, for providing fellowship. The UPE (JU) and DSA (Physics Department, JU) program of UGC, the PURSE and FIST programs (Physics Department, JU) of DST, Government of India are also acknowledged. We are thankful to Dr Sourav Das, Associate Professor, Department of Chemistry, JU for helpful discussions.

References

1. S. Majumder, S. Dey, K. Bagani, S. K. Dey, S. Banerjee and S. Kumar, Dalton Trans., 2015, 44, 7190.
2. S. Dey, S. K. Dey, B. Ghosh, V. R. Reddy and S. Kumar, Mater. Chem. Phys., 2013, 138, 833.
3. A. Tiano, G. C. Papaefthymiou, C. S. Lewis, J. Han, C. Zhang, Q. Li, C. Shi, A. M. M. Abeykoon, S. J. L. Billinge, E. Stach, J. Thomas, K. Guerrero, P. Munayco, J. Munayco, R. B. Scorzelli, P. Burnham, A. J. Viescas and S. S. Wong, Chem. Mater., 2015, 27, 3572.
4. M. M. Shirolkar, R. Das, T. Maity, P. Poddar and S. K. Kulkarni, J. Phys. Chem. C, 2012, 116, 19503.
5. S. Dey, S. K. Dey, K. Bagani, S. Majumder, A. Roychowdhury, S. Banerjee, V. R. Reddy and S. Kumar, Appl. Phys. Lett., 2014, 105, 063110.
6. S. Dey, S. K. Dey, B. Ghosh, P. Dasgupta, A. Poddar, V. R. Reddy and S. Kumar, J. Appl. Phys., 2013, 114, 093901.
7. S. Majumder, S. K. Jana, K. Bagani, B. Satpati, S. Kumar and S. Banerjee, Opt. Mater., 2015, 40, 97.
8. V. Šepelák, A. Feldhoff, P. Heitjans, F. Krumeich, D. Menzel, F. J. Litterst, I. Bergmann and K. D. Becker, Chem. Mater., 2006, 18, 3057.
9. S. Bedanta and W. Kleeman, J. Phys. D: Appl. Phys., 2009, 42, 013001.
10. S. K. Jana, S. Majumder, B. Satpati, S. K. Mishra, R. K. Srivastava and S. Banerjee, RSC Adv., 2015, 5, 37729.
11. P. V. Adhyapak, U. P. Mulik, D. P. Amalnerkar and I. S. Mulla, J. Am. Ceram. Soc., 2013, 96, 731.
12. Y. Wang, J. L. Cao, M. G. Yu, G. Sun, X. D. Wang, H. Bala and Z. Y. Zhang, Mater. Lett., 2013, 100, 102.
13. J. Wang, H. Gao, F. Sun, Q. Hao and C. Xu, Biosens. Bioelectron., 2013, 42, 550.
14. B. Xu, D. Zheng, W. Qiu, F. Gao, S. Jiang and Q. Wang, Biosens. Bioelectron., 2015, 72, 175.
15. N. Li, X. Hao, B. H. Kang, Z. Xu, Y. Shi, N. B. Li and H. Q. Luo, Biosens. Bioelectron., 2016, 77, 525.
16. M. Y. Wang, T. Shen, M. Wang, D. E. Zhang, Z. W. Tong and J. Chen, Sens. Actuators, B, 2014, 190, 645.
17. D. Xiang, L. Yin, J. Ma, E. Guo, Q. Li, Z. Li and K. Liu, Analyst, 2015, 140, 644.
18. R. Zhang and W. Chen, Sci. Bull., 2015, 60, 522.
19. Y. You, Y. Yang and Z. Yang, J. Solid State Electrochem., 2013, 17, 701.
20. R. Zhang, S. He, C. Zhang and W. Chen, J. Mater. Chem. B, 2015, 3, 4146.
21. B. Saha, S. K. Jana, S. Majumder, B. Satpati and S. Banerjee, Electrochim. Acta, 2015, 174, 853.
22. K. C. Fang, C. P. Hsu, Y. W. Kang, J. Y. Fang, C. C. Huang, C. H. Hsu, Y. F. Huang, C. C. Chen, S. S. Li, J. A. Yeh, D. J. Yao and Y. L. Wang, Biosens. Bioelectron., 2014, 55, 294.
23. X. Cao and N. Wang, Analyst, 2011, 136, 4241.
24. S. Bharathi, D. Nataraj, M. Seetha, D. Mangalaraj, N. Ponpandian, Y. Masuda, K. Senthil and K. Yong, CrystEngComm, 2010, 12, 373.
25. M. Cao, T. Liu, S. Gao, G. Sun, X. Wu, C. Hu and Z. L. Wang, Angew. Chem., Int. Ed., 2005, 44, 4197.
26. Z. Liu, B. Lv, D. Wu, Y. Zhu and Y. Sun, CrystEngComm, 2012, 14, 4074.
27. Y. Jiao, Y. Liu, F. Qu and X. Wu, CrystEngComm, 2014, 16, 575.
28. X. Liu, J. Liu, Z. Chang, L. Luo, X. Lei and X. Sun, RSC Adv., 2013, 3, 8489.
29. X. Hu, J. C. Yu, J. Gong, Q. Li and G. Li, Adv. Mater., 2007, 19, 2324.
30. S. Majumder, B. Saha, S. Dey, K. Bagani, M. K. Roy, S. K. Jana, S. Kumar and S. Banerjee, AIP Conf. Proc., 2015, 1665, 050117.
31. K. Lagarec and D. G. Rancourt, Recoil-Mössbauer Spectral Analysis Software for Windows, University of Ottawa Press, Ottawa, 1998.
32. L. Machala, J. Tuček and R. Zbořil, Chem. Mater., 2011, 23, 3255.
33. A. C. Larson and R. B. Von Dreele, General structure analysis system (GSAS), Los Alamos National Laboratory, Report LAUR, 2000, pp. 86–748.
34. B. H. Toby, J. Appl. Crystallogr., 2001, 34, 210.
35. R. Suresh, R. Prabu, A. Vijayaraj, K. Giribabu, A. Stephen and V. Narayan, Mater. Chem. Phys., 2012, 134, 590.
36. S. Dey, R. Gomes, R. Mondal, S. K. Dey, P. Dasgupta, A. Poddar, V. R. Reddy, A. Bhaumik and S. Kumar, RSC Adv., 2015, 5, 78508.
37. K. He, C. Y. Xu, L. Zhen and W. Z. Shao, Mater. Lett., 2008, 62, 739.
38. Z. Liu, C. Y. Chiang, W. Li and W. Zhou, Chem. Commun., 2015, 51, 9350.
39. G. Sun, B. Dong, M. Cao, B. Wei and C. Hu, Chem. Mater., 2011, 23, 1587.
40. X. Zhang, C. Sui, J. Gong, Z. Su and L. Qu, J. Phys. Chem. C, 2007, 111, 9049.
41. D. Sarkar, A. Shukla and D. D. Sharma, ACS Energy Lett., 2016, 1, 82.
42. J. Wang, D. F. Thomas and A. Chen, Anal. Chem., 2008, 80, 997.
43. Z. Dai, S. Liu, J. Bao and H. Ju, Chem.–Eur. J., 2009, 15, 4321.
44. Z. Wen, S. Ci and J. Li, J. Phys. Chem. C, 2009, 113, 13482.
45. X. Jiang, Y. Wu, X. Mao, X. Cui and L. Zhu, Sens. Actuators, B, 2011, 153, 158.
46. S. A. G. Evans, J. M. Elliott, L. M. Andrews, P. N. Bartlett, P. J. Doyle and G. Denuault, Anal. Chem., 2002, 74, 1322.
47. J. S. Easow and T. Selvaraju, Electrochim. Acta, 2013, 112, 648.
48. L. Zhang, Z. fang, Y. Ni and G. Zhao, Int. J. Electrochem. Sci., 2009, 4, 407.
49. C. Cheng, Y. Huang, N. Wang, T. Jiang, S. Hu, B. Zheng, H. Yuan and D. Xiao, ACS Appl. Mater. Interfaces, 2015, 7, 9526.
50. S. Palanisamy, C. Karuppiah, S. M. Chen, R. Emmanuel, P. Muthukrishnan and P. Prakash, Sens. Actuators, B, 2014, 202, 177.
51. Y. E. Miao, S. He, Y. Zhong, Z. Yang, W. W. Tjiu and T. Liu, Electrochim. Acta, 2013, 99, 117.
52. A. M. Golsheikh, N. M. Huang, H. M. Lim, R. Zakaria and C. Y. Yin, Carbon, 2013, 62, 405.
53. Y. Jiang, X. Zhang, C. Shan, S. Hua, Q. Zhang, X. Bai, L. Dan and L. Niu, Talanta, 2011, 85, 76.
54. S. Prasannakumar, R. Manjunath, C. Nethravathi, G. S. Suresh, M. Rajamathi and T. V. Venkatesha, Port. Electrochim. Acta, 2012, 30, 371.
55. A. K. Dutta, S. K. Maji, D. N. Srivastava, A. Mondal, P. Biswas, P. Paul and B. Adhikary, J. Mol. Catal. A: Chem., 2012, 360, 71.
56. J. Ju and W. Chen, Anal. Chem., 2015, 87, 1903.
57. S. He, B. Zhang, M. Liu and W. Chen, RSC Adv., 2014, 4, 49315.
58. M. Liu, S. He and W. Chen, Nanoscale, 2014, 6, 11769.
59. M. Liu, R. Liu and W. Chen, Biosens. Bioelectron., 2013, 45, 206.
60. Y. Wang, X. Yang, J. Bai, X. Jiang and G. Fan, Biosens. Bioelectron., 2013, 43, 180.
61. Q. Yi, F. Niu and W. Yu, Thin Solid Films, 2011, 519, 3155.
62. Y. J. Yang, W. Li and X. Wu, Electrochim. Acta, 2014, 123, 260.
63. R. Devasenathipathy, V. Mani and S. M. Chen, Talanta, 2014, 124, 43.
64. J. Panchompoo, L. Aldous, C. Downing and A. Crossley, Electroanalysis, 2011, 23, 1568.
65. M. A. Aziz and A. N. Kawde, Talanta, 2013, 115, 214.
66. C. Zhang, G. Wang, Y. Ji, M. Liu, Y. Feng, Z. Zhang and B. Fang, Sens. Actuators, B, 2010, 150, 247.
67. S. Kocak and B. Aslısen, Sens. Actuators, B, 2014, 196, 610.
68. R. Sivasubramanian and M. V. Sangaranarayanam, Sens. Actuators, B, 2015, 213, 92.

---

Footnotes

† Electronic supplementary information (ESI) available. See DOI: 10.1039/c6ra10470c

‡ These authors (S. Majumder and B. Saha) contributed equally.

---