Journal of Materials Science: Materials in Electronics https://doi.org/10.1007/s10854-018-8824-7

Photoelectrochemical impedance spectroscopy of electrodeposited hematite α-Fe2O3 thin films: effect of cycle numbers Z. Landolsi1,2 · I. Ben Assaker1 · R. Chtourou1 · S. Ammar1,2,3 Received: 13 November 2017 / Accepted: 27 February 2018 © Springer Science+Business Media, LLC, part of Springer Nature 2018

Abstract In this work, hematite (α-Fe2O3) onto (FTO)-coated glass substrates have been prepared by electrodeposition technique using cyclic voltammetry process. Various samples of α-Fe2O3 with cycle numbers varied between 25 and 300 were prepared. XRD analysis shows that all films crystallize in a rhombohedral phase of hematite with a preferential orientation along (104) direction. The variation in intensity of the principal peaks shows that the crystallinity of thin films was influenced by the cycle numbers or thickness. Surface morphology studies by scanning electron microscopy (SEM) showed that an increase in the cycle numbers causes an increase in the grain size and homogenous of the surface. Moreover, the optical analysis reveals that the band gap energy varied between 2.1 and 1.9 eV in terms of cycle numbers increase. On the other hand, the photo-electrochemical impedance spectroscopy (PEIS) data have been modeled using an equivalent circuit approach. Finally, from Mott–Schottky plot, the flat-band potential and carrier density of α-Fe2O3 thin films have been determined. The results reveal that all the films showed n-type semiconductor character with a flat band potential and a carrier density changed with cycle numbers.

1 Introduction Hematite (α-Fe2O3) is a kind of typical n-type traditional semi-conductor, it possess attractive and unique properties such as high quantum yield, good thermodynamic stability at high temperatures, non-toxicity, and low cost. Moreover, it has emerged as a promising material for photoelectrochemical (PEC) water splitting thanks to its abundance, stability in aqueous environment, favorable position of the electronic valence band and optical band gap that lies in the 2.0–2.2 eV range, to absorb 45% of the incident solar radiation [1]. All these characteristics mean that hematite is used in many applications such as photocatalysis [2], pigment [3], and gas sensors [4]. * I. Ben Assaker [email protected] 1



Laboratoire de Nanomatériaux et Systèmes pour les Energies Renouvelables, Centre de Recherches et des Technologies l’Energie, Technopole Borj Cedria, Bp 95, Hammam Lif 2050, Tunisia

2



Faculté des Sciences de Bizerte, Université de Carthage, Carthage, Tunisia

3

Faculté des Sciences de Gabès, Université de Gabès, Gabès, Tunisia



So far, several kinds of techniques have been investigated to synthesis α-Fe2O3 films such as spray pyrolysis [5], chemical coprecipitation [6], casting method [7], hydrothermal synthesis [8] and electrochemical technique [9–17]. Among these, electrochemical deposition is a simple technique, versatile and very convenient for producing large area devices. On the other hand, the possibility to control the morphology, the formed layers and the film thickness by adjusting the electrical parameters as well as the electrolytic solution composition makes it even more attractive. Indeed, electrochemical preparation of α-Fe2O3 films has been carried out by several studies. For instance, Meng et al. [12] have electrochemically synthesized F ­ e2O3 films on conducting glass (ITO) from an acidic aqueous mixture solution containing ­FeCl2 and ­FeCl3. They focused on the effect of deposition time, current density, reaction temperature and electrolyte concentration in the morphology of α-Fe2O3 nanoparticles. While Tamboli et al. [17] have reported the effect of deposition time on α-Fe2O3 films grown on indium tin oxide (FTO) coated glass substrate they prove that the film thickness played a crucial role in photocurrent variation via thin film modifications. Also, Liu et al. [15] have studied the influence of Ni dopant concentration on Ni-Fe2O3 properties especially on the PEC performance.

13

Vol.:(0123456789)



However, to the best of our knowledge, there are few reports about synthesize of F ­ e2O3 using cyclic voltammetry process and have studies the effect of cycle number on the photo-electrical behavior of hematite [14, 18]. For this reason, the aim of this study is to show how the process of cycle number as well as thickness films affects in the structural, optical, morphological and electrical characteristics of the obtained thin films.

2 Experimental 2.1 Films preparation Electrodeposition of α-Fe2O3 thin films was performed using a potentiostat/galvanostat Autolab PGSTAT 30 (Eco Chemie B V) connected to a three-electrode cell. The working electrode was (FTO)-coated glass substrate, the reference electrode was an Ag/AgCl/3 M KCl and a platinum wire was used as counter electrode. Before using, the FTO substrate must be ultrasonically cleaned with acetone and isopropanol during 15 min then rinsed with deionized water. Inspired to the work of Liu et al. [15], Shwarsctein et al. [19] and Schrebler et al. [21], the electrolyte solution was prepared using iron(III) chloride ­(FeCl3, 5 × 10−3 M), sodium fluoride (NaF, 5 × 10−3 M), potassium chloride (KCl, 0.1 M) and hydrogen peroxide (­ H2O2, 1 M, 35% purity). Electrochemical deposition was conducted by a cyclic voltammetry process at a potential sweep rate of 0.2 V/s, from − 0.5 to 0 V versus Ag/AgCl. The number of cycles has been varied at room temperature from 25 to 300. The color of the films changed from light orange to dark orange with increasing cycle numbers from 25 to 300 (see Fig. 1). To improve the crystallinity of thin films, all as deposited films have been annealed in air atmosphere at 600 °C for 30 min.

2.2 Characterizations The samples were studied and characterized using several techniques. The structural properties were determined by X-ray diffraction technique using an automated Bruker D8 advance X-Ray Diffractometer with Cu Kα (λ = 1.54 Å) in 2θ ranging from 20° to 60°. A Raman scattering study was recorded on a DXR-SMRT Raman (thermo scientific) using 785 nm irradiation. The Surface morphology of α-Fe2O3 was

Journal of Materials Science: Materials in Electronics

observed using a scanning electron microscopy (SEM, JEOL JSM-6700) operating at 15 KV. Absorption and transmission measurements were performed by Perkin Elmer Lambda 950 spectrophotometer with the range of 300–800 nm. Electrical properties of thin films obtained at different numbers of cycle were measured using photo-electrochemical impedance spectroscopy (PEIS) method and Mott–schottky (MS), which were carried out employing the same electrochemical device described above. A solution of 1 M NaOH was used as inert electrolyte. PEIS of the films was performed at open circuit potential (Voc) with the frequency range adjusted between 100 kHz to 0.01 Hz at an amplitude frequency 10 mV in dark and under light irradiation. For illumination, a 300-W Xe lamp was used with a universal AM 1.5. The incident intensity of the focused beam was adjusted to 200  mW/cm2. The fitting of the obtained results in PEIS measurements and the equivalent circuit were carried out using Autolab software command. Mott–Schottky plots were carried out in the range from − 1.2 to 0.2 V versus Ag/AgCl under a frequency of 1 kHz.

3 Results and discussion 3.1 Electrochemical study Generally, cyclic voltammetry (CV) is an effective electroanalytical tool for finding redox couples and confirming the electrode potential during the deposition process. However, this technique can be used also for the deposition of thin films. In this work, we have inspired by the publication of Liu et al. [15], Shwarsctein et al. [19] whose have used this technique to electrodeposited hematite thin films. In comparison to the cathodic electrodeposition synthesis methods, the mechanism of the cyclic voltammetry deposition is based on the redox reaction at the semiconductorliquid junction (SCLJ). The nucleation and growth of the asdeposited films are directly governed by the adsorption and desorption processes, which can be systematically controlled by tunning the electrodeposition synthesis parameters [20]. Generally, the film obtained by cyclic voltammetry deposition is adherent and uniform. Figure 2 illustrate typical two cyclic voltammograms performed at 20 mV/s (Fig. 2a) and at 200 mV/s (Fig. 2b) respectively onto α-Fe2O3 electrode using the electrolytic

Fig. 1  photos of different Hematite thin films electrodeposited onto FTO substrate with cycle numbers varying from 25 to 300

13

Journal of Materials Science: Materials in Electronics

(a)

(b)

0.0

0,0

-3

-2.0x10

-3

-4.0x10

currant (A)

current (A)

-2,0x10 -3

C1 -3

-6.0x10

-3

-6,0x10

-3

-8.0x10

-2

-1.0x10

-0.6

-3

-4,0x10

-0.5

-0.4

-0.3

-0.2

-0.1

0.0

Potential (V vs.Ag/AgCl)

-3

-8,0x10

-0,6

-0,5

-0,4

-0,3

-0,2

-0,1

0,0

potential (V vs.Ag/AgCl)

Fig. 2  Cyclic Voltammogram of α-Fe2O3 in 0.1 M KCl + H2O2 1 M and F ­ eCl3 5 × 10−3 M electrolyte at room temperature at a 1 cycle and b 200 cycles

solution described in the Sect. 2. The voltammetric scan was initiated at 0 V toward the negative direction of up to − 0.6 V versus Ag/AgCl. From (Fig. 2a), we can observe that the cathodic threshold potential for the reduction ­H2O2 to hydroxide (OH–) ions in the solution was about to − 0.15 V versus Ag/AgCl. As a result, the pH level near the working electrodes increases while the solubility of ­Fe3+ ions in the electrolyte decreases causing the precipitation of ­Fe3+ ions. The presence of NaF in electrolyte solution involves the complexation of F ­ e3+ ions onto F ­ eF2+ 2+ according to Eq. 1. In fact, the complexing ­FeFe reacts with OH- ions to form FeOOH films (Eq. 3). Thus, the overall reduction reaction is presented in (Eq. 4). The mechanism of the deposition of the iron oxide is described by many works in the literature and presented by the following equations [20, 21]:

F− + Fe3+ → FeF2+

(1)

H2 O2 + 2 e− → 2OH−

(2)

FeF2+ + 3OH− → FeOOH + 2F− + H2 O

(3)

With an overall reaction:

(4) In order to obtain, thick thin films which can be analyzed by several techniques, the number of cycle has been varied from 25 to 300. An example of cyclic voltammogram performed at 200 mV/s with cycle number equal to 200 presented in Fig. 2b. In this figure, we can see that the cathodic current increase progressively with the voltammetric scan varied from 0 to − 0.5 V. In this case, any reduction peak at about − 0.15 V was observed. Thus, the reasonable assumption is that the scan rate is quick to detect the mechanism of the reduction of ­H2O2.

3H2 O2 + 2FeF2+ + 6e− → 2 FeOOH + 2F− + 2H2 O

However the continuous increase of the cathodic current in the voltammetric scan varied from 0 to − 0.5 V, prove clearly the deposition of thin films by reduction mechanism. Whatever the cycle numbers varied between 25 and 300, the look of the cyclic voltammogram is similar to that presented in Fig. 2b. This behavior indicate that the electrodeposition of ­Fe2O3 follows by the same mechanism describe previously. To the best of our knowledge, except the work of Bouhjar et al. [22] and Schrebler et al. [21] no other reports have presented the cyclic voltammogram of ­Fe2O3 obtained on the electrolyte solution at varied cycle numbers. In order to study the effect of cycle number on the properties of hematite thin films, this semiconductor have been synthesized by cyclic voltammetry process at scan rate fixed to 0.2 V/s with number of cycle varied between 25 and 300.

3.2 Structural analysis X-ray diffraction (XRD) is a reliable technique to determine the crystal structures and phases of the as-prepared films. Figure 3 shows XRD patterns recorded for α-Fe2O3 thin films grown on FTO-coated glass substrates under different deposition cycle numbers ranging from 25 to 300. We can clearly see that the diffraction peaks of all the samples well indexed to FTO substrate and hematite α-Fe2O3. According to Joint Committee on Power Diffraction Standards (JCPDS) database with card number (01-084-0307), F ­ e2O3 crystallize following the rhombohedral phase with space group R3c (167). This result is consistent with data provided by Liu et al. [23], who have synthesized α-Fe2O3 by electrodeposition technique using the cyclic voltammetry process. It can be noticed that the intensity of the hematite peaks (104) and (110) increase with the number of cycles, at the same time

13



Journal of Materials Science: Materials in Electronics 25 cycles

FTO

FTO (00-003-1114)

FTO (00-003-1114) Fe * 2O3 (01-084-0307)

Intensity (a,u)

Intensity (a,u)

(110)

(211) (200) (101)

(104) (110)

*

*

(211) 20

30

40

50

20

60

30

40

Bragg angle 2θ (degree)

FTO (00-003-1114) * Fe2O3(01-084-0307)

(110)

*

20

*

(104)

*

(113)

(012)

(104)

(110)

*

*

*

*

40

*

50

60

20

30

40

60

FTO (00-003-1114)

* (110) *

30

50

*Fe2O3(01-084-0307)

(024)

40 Bragg angle 2θ (degree)

*

50

(116) * (122)

Intensity (a,u)

Intensity (a,u)

20

*

300 cycles

(104)

*

(116) * (122)

Bragg angle 2θ (degree)

FTO (00-003-1114) * Fe2O3(01-084-0307)

(113)

*

*

200 cycles

*

(024)

*

Bragg angle 2θ (degree)

(012)

(113)

(024) (116) (122)

*

30

60

100 cycles

Intensity (a,u)

Intensity (a,u)

50 cycles FTO (00-003-1114) * Fe2O3(01-084-0307)

(012)

50

Bragg angle 2θ (degree)

(012)

(104)

* (110) *

*

(113)

*

60

20

30

*

(024)

40

50

*

(116) (122)

*

*

60

Bragg angle 2θ (degree)

Fig. 3  XRD pattern of electrodeposited α-Fe2O3 onto (FTO)-coated glass substrates at various deposition cycle numbers varying from 25 to 300

13

Journal of Materials Science: Materials in Electronics

the intensity of (211) plane of the substrate (FTO) decreased. This is in agreement with the increase of the layer thickness [16, 24]. In fact, Khedmi et al. [25], fabricated the ­CuIn5S8 crystals by vacuum evaporation on glass substrates. They found that the structural properties using X-ray diffraction analysis reveals that the crystallinity of ­CuIn5S8 increases by increasing the film thicknesses and only the C ­ uIn5S8 phase is present. Making a comparison between the XRD patterns of the samples, the number and the intensities of the characteristic peaks increased with the numbers of cycles. In fact, if the number of cycles are fixed to 25, only two small peaks at 2θ = 33.23° and 35.70° attributed respectively to (104) and (110) reticular planes of α-Fe2O3 were appearing. When the number of cycles increased to 200 cycles, all characteristic peaks of hematite, referred to the JCPDS card number (01084-0307), were presented and indexed in the spectra. Also for the samples synthesized with the number of cycles ranging from 25 to 100, the peak which appears at 2θ = 35.70° is the most intense and sharp, indicating that the growth of α-Fe2O3 was done according a preferential orientation along (110) plane. For the sample electrodeposited with 200 number of cycles, a transition from (110) to (104) plane privilege direction was observed. This change of orientation was also reported by our groups who have studied the effect of time deposition on the structural properties of others type of semicondutcors [16, 24]. They explained the behavior in the structural changes might occur by the surface diffusion and the migration of ad-atoms leading to increased grain growth during the coalescence. To elucidate the effect of cycle numbers or thickness films, some parameters such us crystallite size, microstrain, dislocation density and thickness were calculated. The crystallite size (D) of the hematite α-Fe2O3 through (104) orientation has been calculated from the Debye–Scherrer equation Eq. (5) [26]:

D=

0.9 λ β cos θ

(5)

where D is the crystallite size, k is the shape factor (0.9), λ is the wavelength of incident X-ray from X-ray diffraction (λ = 1.5406 Å), β is the observed angular width at half Table 1  Evolution of crystallite size (D), disslocation density (δ), microstrain (ε) and thickness according to cycle numbers

maximum intensity (FWHM) of the (104) peak and θ is the Bragg angle of diffraction peak. Also, the microstrain (ε) which is an interesting structural parameter of α-Fe2O3 electrodeposited with various cycle numbers can be calculated from the following relation Eq. (6) [27]:

𝜀=

β 4 tan 𝜃

(6)

The dislocation density (δ) is associated with defaults in thin films. It represents the amount of defects in a crystal. The dislocation density of thin films is given by the Williamson and Smallman’s relation Eq. (7) [28]:

𝛿=

1 D2

(7)

Finally, the thickness of the films is estimated by double weight method according to the following relation Eq. (8) [29]:

e=

m 𝜌S

(8)

where m is the mass of thin film, S the area of the deposited film and ρ the density of the α-Fe2O3 (ρ = 5.27 g/cm3 [30]). These results are listed in Table 1. It can be seen clearly an increasing in crystallite size (D) with number of cycles, which revealing a real enhancement of crystallinity and a decreasing in the grain boundary discontinuities. Moreover, the residual stress of α-Fe2O3 thin films corresponding to (104) plane has been found to decrease with number of the cycle’s as well as the dislocation density. It is found that the increase of thin films thickness ensures fragmentation of crystallites and leads indeed to a decrease of the stress parameter. The same results are reported previously by our groups whose have electrodeposited ­CuIn5S8 and ­I n 2S 3 thin films onto an ITO substrate at different deposition times [16, 24]. The sample deposited with an optimum value of cycles = 200, shows lower value of ε = 2.31 × 10−3 and δ = 3.60 × 10−13 m−2. This is attributed to the decreased of defect levels and grain boundaries due to the increase of crystallite size.

Number of cycles

Bragg angle 2θ (°)

FWHM (°)

Crystallite dislocations size D (nm) density δ ­(m−2) × 10−13

Strain ε × 10−3

Thickness (µm)

25 50 100 200 300

– – 33.18 33.07 32.80

– – 0.24 0.16 0.16

– – 35.09 52.65 52.61

– – 3.45 2.31 2.33

0.75 0.94 1.89 2.46 2.75

– – 8.12 3.60 3.61

13



Journal of Materials Science: Materials in Electronics

Intensity (cps)

4000

3000

25 cycles (292) * 50 cycles 100 cycles 200 cycles 300 cycles (224)

*

(408)

*

2000 (608)

1000

(244)

(493)

* 0

200

*

*

300

400

500

600

-1 Raman shift (cm )

Fig. 4  Raman spectra of α-Fe2O3 with different number of cycles varying from 25 to 300. (Color figure online)

In order to confirm the crystal structure and studied the chemical composition of samples, Raman spectroscopy at room temperature was performed. Figure 4 shows the evolution of the Raman spectra of α-Fe2O3 thin films electrodeposited with various number of cycles ranging from 25 to 300. The Raman spectrum of Rhombohedral phase of α-Fe2O3 was identified by Maiti et al. [31] and Iervolino et al. [32] at 224, 244, 293, 413, 501 and 608 cm−1. For the sample obtained with 25 cycles, the Raman spectrum presented any peaks, confirms that the film is amorphous. By increasing the number of cycles, the Raman bands become sharper and intense, thus proving an increase in the size of crystallites which calculated by XRD result. When the number of cycles exceeds 200, the intensities of the peaks relatively decreased indicating a loss of crystallinity.

3.3 Morphological analysis SEM pictures of uncovered of FTO substrate and covered by α-Fe2O3 electrodeposited with various numbers of cycles are exhibited in Fig. 5a–f, respectively. As can be seen, the surface morphology and grain size of hematite samples were affected by various numbers of cycles or thickness. In fact, for the sample synthesized with 25 and 50 cycles, the surface morphology shows minor grain size with spherical shapes heterogeneously distributed. In these cases, the surface of FTO substrate is not well covered by the films. Thus, the polydispersed nature of the grain size could be contributed to an incomplete nucleation and particle growth rate [16]. As the number of cycles is increased to 200 cycles, the sphere like microstructure becomes more uniform, compact and improvement of grain sizes (in order of 500 nm). This observation might be attributed to fewer grains boundaries so to enhancement in crystal quality which is a further proof

13

of XRD findings indicated the preferred orientation of the (104) peak. In contrast, at higher number of cycles (300), deposition film is not uniform, film surface presents rough and porous organization with an agglomeration of the crystallites placed side by side, indicating irregular growth rate of the grains. For this study, we can conclude that the 200 is the optimum number of cycles to synthesize a homogeneous film with a good crystallinity. The surface morphology of α-Fe 2O3 obtained in this study should be compared to the others presented in the literature. In fact, Cai et al. [33], have successfully fabricated hollow nanostructured hematite films by template-assisted electrodeposition and heat-treatment process. These authors have demonstrated that the hematite spheres have an average diameter about 300 nm and the film morphology was affected by the potential sweep rate and post heating temperature. Also, Liu et al. [15] have proved that using cyclic voltammetry process for a whole of 100 cycles, the prepared films reveals mesoporous spheres which were distributed regularly over the surface with diameter about 500 nm.

3.4 Optical properties The optical properties of α-Fe2O3 thin films electrodeposited onto FTO substrates using various cycle numbers (ranging from 25 to 300) were determined from absorbance measurement in the range of 200–800 nm (Fig. 6). Comparing to FTO substrate, all thin films presented an absorption edge between 500 and 600 nm, which corresponds to the excitonic band gap of F ­ e2O3 [34, 35]. When we compared the five curves, we can clearly note that the intensity of optical absorbance in the visible region increased with the increase of the number of cycles from 25 to 300. Indeed, the increasing of absorbance (inversely, the reduction of transmittance) at a higher number of cycles (or thickness) may be attributed to the increased scattering of photons by an increase in roughness of the surface morphology. This behavior is also obtained by Gannouni et al. [24]. From Fig. 6, one can clearly see that the absorbance is found to be maximum, in the range from 400 to 550 nm, for the film electrodeposited with 200 cycles. This may be due to the stable phase and homogeneous film obtained for this sample. On the other hand, an increase in the number of cycles (or thickness) has induced a remarkable shift in the visible light region (red shift) of the optical absorption values. Such effects may be attributed to the ­Fe2O3 grain growth and broadened distribution associated with the number of cycles [36]. This result is in good concordance with those obtained by XRD and SEM analysis. Generally, the fundamental absorption edge of the films corresponds to transitions of electrons from the valence band to the conduction band edge and this can be used to calculate

Journal of Materials Science: Materials in Electronics Fig. 5  SEM images of samples: a FTO, b α-Fe2O3-25 cycles, c α-Fe2O3-50 cycles, d α-Fe2O3-100 cycles, e α-Fe2O3-200 cycles and f α-Fe2O3-300 cycles

the difference in the optical band gap of the films. In the case of a direct transition, the absorption coefficient and optical band gap are related by the Tauc’s relation which corresponds to the direct band gap [37]:

(𝛼h𝜈) = A (h𝜈 − Eg)2 where A is a constant, hυ is the photon energy and Eg is the optical band gap. Figure 7a shows the plot of [α (hν)]2 versus the photon energy hν which yields in the sharp absorption edge for the high quality films by a linear fit. The calculated values of optical band gap Eg for the different samples are depicted in Fig. 7b. The optical energy band gap of thin films

was found to decrease from 2.1 to 1.90 eV with increasing the cycle numbers from 25 to 200. These relative high shifts may be due to the influence of several factors such as thickness, grain size, structural parameters, lattice strain, carrier concentration, presence of impurities. For these samples, the reason for the decreasing should be mainly due to the high crystallinity and an increase of grain crystal size of α-Fe2O3 as confirmed by XRD and SEM analysis. The slight increase of Eg for the sample synthesized with 300 cycles is related to the partial degradation of the microstructure. Similar scientific argument has been observed with different materials by many researchers. For example, Akaltun

13



Journal of Materials Science: Materials in Electronics

Absorbance (%)

FTO 25cycles 50cycles 100cycles 200cycles 300cycles

300

400

500

600

700

800

Wavelength(nm)

Fig. 6  Optical absorption spectrum of α-Fe2O3 thin films with different number of cycles. (Color figure online)

et al. [38] have deposited CdSe on glass substrates using successive ionic layer adsorption and reaction (SILAR) method. They found that band gap values of their material decreased with increasing the film thickness. They reported this behavior to the improvement in the crystals, in the morphological changes of the films, in the changes of atomic distances and the grain size and the structural defects in the films. Also, Khedmi et al. [39], have demonstrated that the film thickness of ­SnSb2S4 can be used to modify the structural and the optical properties of thin films.

3.5 Electrical properties In order to investigate the electronic properties of α-Fe2O3 thin films and to explore the charge transport dynamics in the electrode, the electrochemical impedance spectroscopy (EIS) are measured at an open circuit potential (~ − 0.28 vs. Ag/AgCl) of F ­ e2O3 obtained with different cycle numbers

5

-1 2 2 hv) (eV,cm )

4

(b)

25cycles 50cycles 100cycles 200cycles 300cycles

2.15

2.10 Direct band gap E g (eV)

(a)

in dark and under illumination at an AC frequency varying from 100 kHz to 0.01 Hz. In fact, the EIS is an excellent method and well-established technique to characterize electrode interface and to investigate the electron transfer across between the electrolyte and the surface of the electrode [40–42]. Figure 8 illustrates the typical impedance spectra (Nyquist plots) of the FTO/Fe2O3 electrode in the solution of 1 M NaOH before (a) and after (b) light illumination. Under dark condition (Fig. 8a), only one arc is observed in all measurement frequency for the FTO/Fe2O3 electrode obtained at different number of cycles ranging from 25 to 300. However, when the light is on (Fig. 8b), the first arc becomes smaller and a second arc appears at a low frequencies. This behavior improved a charge carrier separation under illumination [43]. In fact, if the electrode is excited by xenon lamp, the Nyquist diagrams (Fig. 8b) exhibit two depressed overlapping capacitive semi-circles in high and low frequencies. The first semicircle show at high frequencies could be probably associated to charge transfer resistance at the electrode/electrolyte interface and the second one at low frequencies due to the transfer at the double layer [8, 44, 45]. Reductions or increases in the impedance should be correlated with an increase or decrease, respectively, in the electron drift mobility through the electrode. In other words, smaller arc radius means smaller charge transfer resistance [23]. This decrease in the electron transfer resistance is may be attributed to a kinetic barrier for the electron transfer, which perturbed the interfacial electron transfer considerably at the electrode surface [46]. In order to better see the effect of cycle number in the electrical properties of hematite thin films, the EIS spectra, can be modeled by a Randles equivalent circuit simulation see insert of Fig. 8a and b. In these circuits, Rs is the electrolyte resistance, R1 and C1 represent the space charge resistance and capacitance respectively.C2 and R2 are the

3

2

2.05

2.00

1.95

1

1.90 0

1.6

1.8

2.0

Photon energy hv (eV)

2.2

2.4

0

50

100

150

200

250

300

Number of cycles

Fig. 7  a Plots of (αhν)2 versus (hν) for the electrodeposited α-Fe2O3 thin films onto FTO-(coated) glass substrates and b evolution of band gap with different cycle numbers. (Color figure online)

13

Journal of Materials Science: Materials in Electronics

(a)

600000

(b)

25 cycles 50 cycles 100 cycles 200 cycles 300 cycles

500000

40000 -Z''(Ohm)

-Z''(Ohm)

25 cycles 50 cycles 100 cycles 200 cycles 300 cycles

50000

400000

300000

200000

30000 20000

100000

0

60000

R1

R2

c1

c2

Rs

10000

0

50000

100000

150000

200000

0

250000

Z'(Ohm)

0

10000

20000

30000

40000

50000

Z'(Ohm)

Fig. 8  Nyquist diagram of α-Fe2O3 thin films and, in insert graph the equivalent circuit obtained in a the dark condition and b under illumination. (Color figure online)

Cycle numbers

25 50 100 200 300

In dark

Under illumination

RS (Ω)

R (k Ω)

C (µF)

RS (Ω)

R1 (KΩ)

C1 (µF)

R2 (k Ω)

C2 (µF)

550 503 490 420 450

808 853 747 301 683

16.7 15.1 17.1 21.6 20.9

77 118 94.5 98.2 63.7

212 215 217.5 172 215.16

49.8 100 186 223 200

89 72 40 24.4 31.6

80.9 86.4 180 320 191

electrical elements related to double layer capacitance and faradic charge transfer resistance. The values of all parameters obtained in the dark and under illumination from the fitting to the equivalent circuit are summarized in Table 2. In the dark condition, it can be seen that the capacitance C increased slightly and conversely, the apparent resistance R decreased remarkably from the increase of cycle numbers. Under illumination, all the results imply that some changes were formed on the electrode surface, which resulted from the light excitation. A comparison of values, shows that the sample deposited with 200 cycle present a lower values of charge transfer resistance, e.g. R2 = 24.4 kΩ, indicating a reduction in electron recombination and enhancement in the efficiency of electron transport. Moreover, this sample has a higher conductivity which is related to the capacitance C2 about 320 µF. consequently, hematite deposited with 200 cycles has the lowest resistance and the highest double layer capacitance in the dark and under illumination. Such effect could be explained by the rapid transfer charge in the ­Fe2O3 electrode and by the high accumulation of electron at the interface electrode/electrolyte after increasing the number of cycles to 200. From the representative Bode (phase angle vs. frequency) diagrams (Fig. 9), two phases angle peaks were observed,

80

25 cycles 50 cycles 100 cycles 200 cycles 300 cycles

70 60

-Phase angle/degree

Table 2  Equivalent circuit parameters of α-Fe2O3 in dark and under illumination in NaOH electrolyte

50 40 30 20 10 0 -2

10

-1

10

0

10

1

10

2

10

3

10

4

10

Frequency (Hz)

Fig. 9  The phase impedance diagrams of α-Fe2O3 with different cycle numbers from 25 to 300. (Color figure online)

which correspond to two semicircles in the Nyquist plot under illumination (Fig. 8b). One can see that the phase angle of F ­ e2O3-200 cycles at low frequency such as 1 to 100 Hz is higher than that the others, which suggest that thin

13



Journal of Materials Science: Materials in Electronics

film prepared with 200 cycles can be more capacitive than the other films [47]. The higher capacitance obtained with hematite deposited at 200 cycle indicates an easy transfer of charge at this sample. This behavior is consistent with XRD and morphological results which confirm the good crystallinity and an homogeneous surface for the sample obtained with 200 cycle. The nature of the photoelectrode material, depletion layer width and flat band potential (­ Vfb) are useful parameters to examine the charge transfer process across the interface of semiconductor-electrolyte junction which are estimated from the Mott–Schottky relation [48]. [ ] ( ) KT 2 1 = V − V − n_type (9) Fb e c2 𝜀𝜀0 S2 Nd and

] [ ) KT −2 ( 1 = V − V − Fb e c2 𝜀𝜀0 S2 Na

(10)

p_type

Respectively, where N ­ d/Na is the donor/acceptor concentration ­(cm−3), ε0 is the vacuum permittivity (8.85 × 10−14 F/ cm), ε is the relative dielectric constant of the corresponding passive film (ε = 80 for α-Fe2O3) [49]. K is the boltzmann’s constant (1.38 1­ 0−23 J/K) and T is the Kelvin temperature. V is the applied potential, ­VFb is the flat-band potential, e is the positive elementary charge (e = 1.6 × 10−19 C), S is the surface area of the electrode (1 cm2). The Mott–Schottky plots for thin films of ­Fe2O3 obtained with different number of cycles are shown in Fig. 10. The results were obtained in the measurements performed in 1 M NaOH at the frequency of 1 kHz. For all samples, a linear relationship of 1/C2 versus V can be observed. The positive slopes indicate that thin films of F ­ e2O3 are n-type semiconductor. This result is in good concordance with others

25 cycles 50 cycles 100 cycles 200 cycles 300 cycles

10

2.5x10

10

4 2 2 1/C (Cm /F )

2.0x10

published in the literature [32, 50, 51]. Extrapolating the linear part of the plot towards the potential axis abscissa gives the value of ­Vfb at the intercept on the potential axis. The values of ­Vfb deduced from the curves are presented in Table 3. By increasing the number of cycles from 25 to 300, the flat band potential (Vfb) shifts significantly from − 0.75 to − 0.98 V. versus Ag/AgCl. This can be related to the change in the morphology and crystal structure when increasing the film thickness [47]. The donor densities ­Nd can be calculated from the slopes of the Mott–Schottky analysis, the obtained values are summarized in Table 3. The increment implies that the increase of number of cycle or thickness improved the carrier concentration at the interface slightly. Clearly, the sample electrodeposited with 200 cycles possesses the highest donor density. Such results indicate that the charge transport in this sample will be faster than that in the other films. Generally, our values are comparable to those observed by Pravin.S.Shinde et al. [49] who have obtained a typical value of Vfb = -0.5V vs. Ag/AgCl and carrier density approximately ­1016 ­cm−3 for α-Fe2O3 synthesized by pulse reverse electrodeposition technique.

4 Conclusion In summary, hematite (α-Fe2O3) films have been successfully electrodeposited onto FTO substrate using cyclic voltammetry process. The properties were discussed from the effects of various cycle numbers (or thicknesses) on the structure, morphology, optical and electrical properties of α-Fe2O3 thin films in details. From the XRD analysis, it was found that after annealing at 600 °C, all films crystallize in the Rhombohedral phase and a significant change in the preferred orientation from (110) to (104) was observed with the increase of the number of cycles beyond 200. Furthermore, it was proved that the sample obtained with 200 cycles has a better crystallinity and a homogenous surface with an average grain size about 52.65 nm. From the optical measurements, the band gap of these films is calculated. All films show high absorption

10

1.5x10

Table 3  Mott Shottky results for the electrodeposited α-Fe2O3 at different number of cycles

10

1.0x10

9

5.0x10

0.0 -1.2

-1.0

-0.8

-0.6

-0.4

-0.2

0.0

0.2

Potential (V vs.Ag/AgCl)

Fig. 10  Mott Shottky plots of α-Fe2O3 films on FTO substrate with different number of cycles. (Color figure online)

13

Number of cycles

VFb (V)

Slope ­(1010)

ND ­(1016) ­(cm−3)

25 50 100 200 300

− 0.75 − 0.74 − 0.76 − 0.81 − 0.98

2.95 2.93 2 1.7 2

0.59 0.60 0.88 1.03 0.88

Journal of Materials Science: Materials in Electronics

in the visible region and the band gap is found to decrease from 2.1 to 1.9 eV with increasing the cycle numbers from 25 to 200. Photo-electrochemical impedance spectroscopy was employed to analyze the capacitive nature of α-Fe2O3/ NaOH interface in dark and under illumination. The interface was modeled by an equivalent circuits. Following this electrochemical study, we show that the increasing of the thickness contributes to the reduction of resistance of the charge transfer inside the electrode and across the interface of electrode and electrolyte. Using the Mott–Schottky plot, all thin films are n-type conduction with a flat band potential in the range of − 0.75 to − 0.98 V which can be shifted negatively by increasing the cycle numbers. The high value of donor density Nd (1.03 × 1016 cm−3) was obtained for films deposited with 200 cycles. Future investigations focus on the photoelectrochemical responses and photocatalytic properties of pure α-Fe2O3 obtained at 200 cycles and doped by several transition metals. The work is currently in progress and the results will be reported subsequently.

References 1. B. Landolo, B. Wickman, I. Zoric, A. Hellman, The rise of hematite: origin and strategies to reduce the high onset potential for the oxygen evolution reaction. J. Mater. Chem. A 33, 16896 (2015) 2. X. Yan, Y. Wu, D. Li, C. Luo, Y. Wang, J. Hu, G. Li, P. Li, H. Jiang, W. Zhang, Facile synthesis of ring-like α-Fe2O3 assembly composed of small hematite particles for highly efficient potocatalysis., J. Mater. Sci.: Mater. Electron. https​://doi.org/10.1007/ s1085​4-017-8186-6 3. S.A. Jadhav, R. Bongiovanni, D.L. Marchisio, D. Fontana, C. Egger, Surface modification of iron oxide (­ Fe2O3) pigment particles with amino-functional polysiloxane for improved dispersion stability and hydrophobicity. Pigm. Resin. Technol. 43, 219–227 (2014) 4. M. Dai, L. Zhao, H. Gao, P. Sun, F. Liu, S. Zhang, K. Shimanoe, N. Yamazoe, G. Lu, Hierarchical assembly of α-Fe2O3 nanorods on multiwall carbon nanotubes as a high-performance sensing material for gas sensors. ACS Appl. Mater. Interfaces 9, 8919– 8928 (2017) 5. P.D. More, P.R. Jadhav, A.A. Ghanwat, I.A. Dhole, Y.H. Navale, V.B. Patil, Spray synthesized hydrophobic α-Fe2O3 thin film electrodes for supercapacitor application. J. Mater. Sci.: Mater. Electron. 28, 1–10 (2017) 6. A. Lassoued, B. Dkhil, A. Gadri, S. Ammar, Control of the shape and size of iron oxide (α-Fe2O3) nanoparticles synthesized through the chemical precipitation method. Results Phys. 7, 3007– 3015 (2017) 7. A. Badawi, E.M. Ahmed, N.Y. Mostafa, F. Abdel-Wahab, S.E. Alomairy, Enhancement of the optical and mechanical properties of chitosan using F ­ e2O3 nanoparticles. J. Mater. Sci.: Mater. Electron. 28, 10877–10884 (2017) 8. H. Katsuki, E.K. Choi, W.J. Lee, W.S. Cho, K.T. Hwang, W. Huang, S. Komarnenic, Controlled synthesis of hexagonal α-Fe2O3 crystals for ceramic colors by hydrothermal reaction of ­FeCl3 and NaOH solutions. Ceram. Int. 34, 14050–14056 (2017) 9. S.S. Shendage, U.B. Patil, J.M. Nagarkar, Electrochemical deposition of highly dispersed palladium nanoparticles on

10. 11.

12. 13.

14. 15. 16.

17.

18.

19.

20.

21.

22.

23.

24.

25.

nafion-graphene film in presence of ferrous ions for ethanol electrooxidation. Fuel Cells 13, 364–370 (2013) L. Li, C. Chang, H. Wu, J. Shiu, P. Wu, E. Diau, Morphological control of platinum nanostructures for highly efficient dyesensitized solar cells. J. Mater. Chem. 22, 6267–6273 (2012) M. Vanags, A. Šutka, J. Kleperis, P. Shipkovs, Comparison of the electrochemical properties of hematite thin films prepared by spray pyrolysis and electrodeposition. J. Ceram. Inter. 41, 9024– 9029 (2015) Q. Meng, Z. Wang, X. Chai, Z. Weng, R. Ding, L. Dong, Fabrication of hematite (α-Fe2O3) nanoparticles using electrochemical deposition. J. Appl. Surf. Sci. 368, 303–308 (2016) G. Iervolino, I. Antis, L. Sygellou, V. Vaiano, D. Sannino, P. Lianos, Photocurrent increase by metal modification of ­Fe2O3 photoanodes and its effect on photoelectrocatalytic hydrogen production by degradation of organic substances. J. Appl. Surf. Sci. 400, 176–183 (2017) S.L. Kadam, P.M. Padwal, S.M. Mane, S.B. Kulkarni, Electrodeposition of F ­ e2O3 nanoparticles and its supercapacitive properties, AIP Conference Proceedings (2016) Y. Liu, Y. Xiang Yu, W.D. Zhang, Photoelectrochemical properties of Ni-doped ­Fe2O3 thin films prepared by electrodeposition. J. Electrochim. Acta 59, 121–127 (2012) Z. Braiek, M. Gannouni, I.B. Assaker, A. Bardaoui, A. Lamouchi, A. Brayek, R. Chtourou, Correlation between physical properties and growth mechanism of ­In2S3 thin films fabricated by electrodeposition technique with different deposition times. Eur. Phys. J. Appl. Phys. 72, 10302 (2015) S.H. Tamboli, G. Rahman, O.S. Joo, Influence of potential, deposition time and annealing temperature on photoelectrochemical properties of electrodeposited iron oxide thin films. J. Alloy. Compd. 520, 232–237 (2012) P.S. Shinde, A. Annamalai, J.H. Kim, S.H. Choi, J.S. Lee, J.S. Jang, Photoelectrochemical, impedance and optical data for self Sn-diffusion doped ­Fe2O3 photoanodes fabricated at high temperature by one and two-step annealing methods. J. Data Brief 5, 796–804 (2015) A.K. Shwarsctein, Y.S. Hu, A.J. Forman, G.D. Stucky, E.W. McFarland, Electrodeposition of α-Fe2O3 doped with Mo or Cr as photoanodes for photocatalytic water splitting. J. Phys. Chem. 112, 15900–15907 (2008) Y.W. Phuana, W.J. Ongc, M.N. Chonga, J.D. Ocone, Prospects of electrochemically synthesized hematite photoanodes for photoelectrochemical water splitting: a review. J. Photochem. Photobiol. C 33, 54–82 (2017) R. Schrebler, K. Bello, F. Vera, P. Cury, E. Muñoz, R.d.. Río, H.G. Meier, R.C. Órdova, E.A. Dalchiele, An electrochemical deposition route for obtaining α-Fe2O3 thin films. Electrochem. Solid-State Lett. 9, 110–113 (2006) F. Bouhjar, S. Ullah, M.L. Chourou, M. Mollar, B. Mar, B. Bessais, Electrochemical fabrication and characterization of p-CuSCN/ n-Fe2O3 heterojunction devices for hydrogen production. J. Electrochem. Soc. 164, 936–945 (2017) Y. Liu, Z. Xu, M. Yin, H. Fan, W. Cheng, L. Lu, Y. Song, J. Ma, X. Zhu, Enhanced photoelectrocatalytic performance of α-Fe2O3 thin films by surface plasmon resonance of Au nanoparticles coupled with surface passivation by atom layer deposition of ­Al2O3. Nanoscale Res. Lett. (2015). https​://doi.org/10.1186/ s1167​1-015-1077-y M. Gannouni, I.B. Assaker, R. Chtourou, Role of deposition time on structural, optical and electrical properties of In-rich Cu–In–S spinel films grown by electrodeposition technique. Superlattices Microstruct. 61, 22–32 (2013) N. Khedmi, M. Ben Rabeh, M. Kanzari, Thickness dependent structural and optical properties of vacuum evaporated C ­ uIn5S8 thin films. Energy Procedia 44, 61–68 (2014)

13

26. B.E. Warren, X-ray diffraction. (Dover, New York, 1990) 27. P.P. Sahay, R.K. Nath, Al-doped ZnO thin films as methanol sensors. Sens. Actuators B 2, 654–659 (2008) 28. S. Thanikaikarasan, T. Mahalingam, M. Raja, T. Kim, Characterization of electroplated FeSe thin films. J. Mater Sci. 20, 727–734 (2009) 29. H. Maghraoui-Meherzi, T. Ben Nasr, N. Kamoun, M. Dachraoui, Structural, morphology and optical properties of chemically deposited ­Sb2S3 thin films. Phys. B 405, 3101–3105 (2010) 30. S. Sakurai, S. Kuroki, H. Tokoro, K. Hashimoto, S. Ohkoshi, Synthesis, crystal structure, and magnetic properties of ε-InxFe2–xO3 nanorod-shaped magnets. Adv. Func. Mater. 17, 2278–2282 (2007) 31. D. Maiti, V. Aravindan, S. Madhavi, P. Sujatha, Devi, Electrochemical performance of hematite nanoparticles derived from spherical maghemite and elongated goethite particles. J. Power Sources 276, 291–298 (2015) 32. G. Iervolino, L. Tantis, L. Sygellou, V. Vaiano, D. Sannino, P. Lianos, Photocurrent increase by metal modification of ­Fe2O3 photoanodes and its effect on photoelectrocatalytic hydrogen production by degradation of organic substances. J. Appl. Surf. Sci. 400, 176–183 (2017) 33. J. Cai, S. li, Z. Li, J. Wang, Y. Ren, G. Qin, Electrodeposition of Sn-doped hollow α-Fe2O3 nanostructuresfor photoelectrochemical water splitting. J. Alloy. Compd. 574, 421–426 (2013) 34. L. Li, C. Liu, Y. Qiu, N. Mitsuzak, Z. Chen, The influence of the hydrothermal temperature and time on morphology and photoelectrochemical response of α-Fe2O3 photoanode. J. Alloy. Compd. 696, 980–987 (2017) 35. N.F. Habubi, S.S. Chiad, K.H. Abass, M.M. Abood, Annealing effects on the some optical properties of ­Fe2O3 thin films doped NiO. Int. J. Eng. Technol. 8, 44–50 (2016) 36. X.G. Wen, S.H. Wang, Y. Ding, Z.L. Wang, S.H. Yang, Controlled growth of large-area, uniform, vertically aligned arrays of a-Fe2O3 nanobelts and nanowires. J. Phys. Chem. B. 109, 215–220 (2005) 37. J.I. Pankoye, Optical processes in semiconductors. (Prentice-Hall, Englewood Cliffs, 1971) 38. Y. Akaltun, M.A. Yıldırım, A. Ateş, M. Yıldırım, The relationship between refractive index-energy gap and the film thickness effect on the characteristic parameters of CdSe thin films. J. Optics Commun. 284, 2307–2311 (2011) 39. N. Khedmi, M. Ben Rabeh, M. Kanzari, Structural morphological and optical properties of S ­ nSb2S4 thin films grown by vacuum evaporation method. J. Mater. Sci. Technol. 30, 1006–1011 (2014) 40. T.V. Nguyen, H.C. Lee, O.B. Yang, The effect of pre-thermal treatment of ­TiO2 nano-particles on the performances of dyesensitized solar cells. Sol. Energy Mater. Sol. Cells. 90, 967–981 (2006)

13

Journal of Materials Science: Materials in Electronics 41. J. Van de Lagemaat, N.G. Park, A.J. Frank, Influence of electrical potential distribution, charge transport, and recombination on the photopotential and photocurrent conversion efficiency of dyesensitized nanocrystalline ­TiO2 solar cells: a study by electrical impedance and optical modulation techniques. J. Phys. Chem. B. 104, 2044–2052 (2000) 42. J. Bisquert, G. Garcia-Belmonte, F. Fabregat-Santiago, N.S. Ferriols, P. Bogdanoff, E.C. Pereira, Theory of the impedance of electron diffusion and recombination in a thin layer. J. Phys. Chem. B 106, 325–333 (2002) 43. Q. Wu, J. Ouyang, K.P. Xie, L. Sun, M. Wang, C. Lin, Ultrasoundassisted synthesis and visible-light-driven photocatalytic activity of Fe-incorporated T ­ iO2 nanotube array photocatalysts. J. Hazard. Mater. 410, 199–200 (2012) 44. M. Gannouni, I.B. Assaker, R. Chtourou, Experimental investigation of the effect of indium content on the C ­ uIn5S8 electrodes using electrochemical impedance spectroscopy. Mater. Res. Bull. 61, 519–527 (2015) 45. N. Arbi, I.B. Assaker, M. Gannouni, A. Kriaa, R. Chtourou, Effect of manganese concentration on physical and electrochemical properties of M ­ n2+-doped ZnS thin films deposited onto ITOglass) substrates by electrodeposition technique. J. Mater. Sci.: Mater. Electron. 28, 4997–5005 (2017) 46. S.K. Mishra, D. Kumar, A.M. Biradar, Rajesh, Electrochemical impedance spectroscopy characterization of mercaptopropionic acid capped ZnS nanocrystal based bioelectrode for the detection of the cardiac biomarker: myoglobin. J. Bioelectrochem. 88, 118–126 (2012) 47. O. Messaoudi, I. Benassaker, M. Gannouni, A. Souissi, H. Makhlouf, A. Bardaoui, R. Chtourou, Structural, morphological and electrical characteristics of electrodeposited ­Cu2O: effect of deposition time. Appl. Surf. Sci. 366, 383–388 (2016) 48. S.R. Morrison, Electrochemistry at Semiconductor and Oxidized Metal Electrodes (Plenum Press, New York, 1980) 49. P.S. Shinde, G.H. Go, W.J. Lee, Facile growth of hierarchical hematite (α-Fe2O3) nanopetals on FTO by pulse reverse electrodeposition for photoelectrochemical water splitting. J. Mater. Chem. 22, 10469–10471 (2012) 50. C. Zeng, Y. Bai, Y. Ling, Z. Xin, H. Liang, X. Deng, Hydrogen interaction characteristic of nanoscale oxide films grown on iron nickel based stainless steel by selective thermal oxidation. Int. J. Hydrog. Energy 42, 1–12 (2017) 51. Q. Liu, C. Chen, G. Yuan, X. Huang, X. Lü, Y. Cao, Y. Li, A. Hu, X. Lu, P. Zhu,Morphology-controlled α-Fe2O3 nanostructures on FTO substrates for photoelectrochemical water oxidation. J. Alloy. Compd. 715, 230–236 (2017)