Ferroelectric and photoelectrochemical studies of lead-free Ba_0.925Bi_0.05 □_0.025(Ti_0.65Zr_0.30Sn_0.05)O₃ ceramic and its application to Rhodamine B oxidation under solar light

2.1 Materials preparation

Pure powders of BaCO₃, Bi₂O₃, TiO₂, ZrO₂, and SnO₂ (all Sigma Aldrich) were used as starting materials; they were previously dried at 200 °C (3 h) to eliminate the adsorbed water. The ceramic of composition Ba_0.925Bi_0.05□_0.025(Ti_0.65Zr_0.30Sn_0.05)O₃ have been prepared by solid-state reaction, as described below: 0.925 BaCO₃ + 0.025 Bi₂O₃ + 0.65 TiO₂ + 0.30 ZrO₂ + 0.05 SnO₂ →

2.2 Characterization

However, the identification of the phase, the related symmetry and the unit-cell parameters were achieved using a D8 Advance X-ray diffractometer (Vantack detector). The data were collected at room temperature using Cu_Kα1+2 radiation in the 2θ range (10–80°). Scanning electron microscopy (SEM) of the ceramic was obtained with a Philips XL30 microscope. The Raman spectrum was plotted with an Yvon Jobin T64000 spectrometer using Ar excitation laser line (514.5 nm).

The thermal dependence of the dielectric properties was performed in the temperature (100–450 K) and frequencies (10²–0⁶ Hz) ranges thanks to a Solartron Impedance Analyzer SI 1200.

The photo electrochemistry was studied in a conventional cell with Pt as auxiliary electrode and SCE as reference in neutral medium (Na₂SO₄, 0.1 M); the intensity-potential J(E) were plotted at a scan rate of 10 mV s⁻1 while the capacitance was measured against the potential at a constant frequency of 10 kHz. The photocatalytic experiments were carried out in a batch mode in an open Pyrex reactor, exposed to direct sunlight. The solubility of Rh B in water is relatively high (∼15 g/L). The powder catalyst (50 mg) was suspended in 100 mL of Rh B solution at a concentration of 50 mg/L at pH 7, with a dark adsorption duration of 30 min. The powder was centrifuged (3000 rpm, 15 min.) after each test to separate the powder from the electrolyte for the chemical analysis. The greatest sensitivity detection limit improves both the selectivity and resolution between the investigated compounds and degradation products. UV–Visible spectrophotometry (λ_max = 554 nm) was used to measure the remaining Rh B concentration (C_t).

3.1 Microstructural study

Fig. 1 (a Inset) shows the microstructure of BBiTZS. The grains appear in square shape arranged with each other leading to few vacant spaces, giving rise to a high microstructural density. The histogram of grain distribution reveals small sizes within this ceramic (0.3–0.8 μm). Tang et al. (2004) reported the grain size influence on the Ba (Ti_1−yZr_y)O₃ system. They showed that the coarse-grained sample undergoes a structural phase transition, instead to a relaxor-like behavior for the fine-grained sample. As BBiTZS is distinguished by small grain sizes, we expected that it would behave as a relaxor ferroelectric as confirmed by the dielectric measurements (see below).

Close

Fig. 1

Rietveld refinement and microstructure (Inset a) and Raman spectrum (b) for BBiTZS room temperature.

Export to PPT

3.2 Room-temperature X-ray diffraction

The single phase formation of the BBiTZS sample was confirmed by XRD analysis at room temperature. The Rietveld refinement and calculation reliability factors were determined in the cubic symmetry (Pm $\overline{3}$ m, JCPDS N°75-0461) using the Fullprof software (Boultif and Loueer, 1991) integrated in the Win PLOTR program (Rodríguez-Carvajal, 2002) where all the reflection peaks are indexed. A perfect agreement is observed between the observed XRD patterns and fitted data (Table 1). Moreover, the agreement between the observed and calculated inter-planer spacing suggests a cubic symmetry at room temperature.

Table 2 gathers the lattice parameters and refinement reliability factors obtained by the Rietveld refinement in comparison with those of BT. The fitting parameters (R_B, R_F, Rwp and χ²) indicate good agreement between refined and observed XRD and confirm the cubic symmetry.

Table 3 assembles the refined atomic positions, occupancy and isotropic atomic shift parameters of BBiZTS composition. As it can be seen, Ba²⁺ and Bi³⁺ are located in the same A-site with a fractional percentage. In addition, Ti⁴⁺/Zr⁴⁺/Sn⁴⁺cations located in the B-sites are connected with the oxygen atoms to form (Ti⁴⁺/Zr⁴⁺/Sn⁴⁺)O₆ octahedra sharing corners. In addition, relatively good reliability factors (compared to standard values) are in the range of disordered perovskites with a tolerance factor below unity (Cross, 1994).

3.3 Raman analysis

The Raman spectrum of BBiTZS (Fig. 1b) confirms the observed active modes in the cubic symmetry, in agreement with the relaxor ferroelectric compositions. The assignment of the Raman modes is carried out by analogy with the Ba_1−xBi _xTi_1−xYb_x/2Fe_x/2O₃ and BaZr_xTi_1-xO₃ (Schileo et al., 2013; Dobal et al., 2001). The spectrum consists of six distinguished modes 110, 186, 256, 301, 518 and 720 cm⁻¹ according respectively to the following modes: E(TO₁), A₁(TO₁), A₁(TO₂), E(TO₂), A₁(TO₃) and A₁(LO₃) + E(LO). The A₁TO₂ mode tends to disappear while the modes A₁(LO₃) + E(LO) become wider. These observations are related to the fact that BBiZTS is in its paraelectric (cubic) phase at RT. We recall that the cubic -paraelectric phase allows 12 optical modes (3F1u + F2u) that are not Raman active However, for BBiTZS, the activation modes are caused by a considerable structural disorder characterizing the ferroelectric relaxor (Dobal et al., 2001). The RT Raman study about the influence on the structural transition has been reported in detail in our previous work (Haddadou et al., 2018b; Smail, 2020).

3.4 Dielectric studies

3.4.1

3.4.1 Thermal variation of dielectric permittivity

The thermal variation relative permittivity (ε′_r) and dielectric losses (tg δ) of BBiTZS at various frequencies exhibit a single broad peak with a frequency dispersion compatible with a relaxor ferroelectric behavior (Fig. 2). Indeed, ε′_r increases gradually to a maximum value ε′_rmax with increasing temperature and then decreases smoothly indicating a phase transition. ε′_rmax and the corresponding temperature maximum (T_m) depend upon the measurement frequency. Indeed, BBiTZS exhibits a broad peak with a maximum ε′_rmax that decreases with increasing frequency. In addition, the temperature related to this maximum (T_m) shifts toward higher values as the frequency increases. All these dielectric characteristics agree with the ferroelectric relaxor behavior. Generally, the broadening of the dielectric peaks is attributed to some structural disorder generated by the compositional fluctuation in the crystalline network as can be considered for the present sample.

Close

Fig. 2

Thermal evolution of the dielectric constant and dielectric losses at various frequencies for BBiTZS.

Export to PPT

3.4.3

3.4.3 Deviation from Curie-Weiss law in BBiTZS

The dielectric constant of a normal ferroelectric obeys the Curie Weiss (C-W) law in the paraelectric phase. The following relation expresses this:

(3)

$$\frac{1}{{\epsilon }_r^{\prime }}=\frac{(T-T_o)}{C}(T>T_c)$$ where C is the Curie constant and T_o the C-W temperature. The curve 1/ε′_r = f (T) (Fig. 3) allowed the determination of the corresponding constant C and T_o for BBiTZS. These values as well as the parameter ΔT_m (eq. (4)) are reported in Table 4:

(4)

$$\mathrm{\Delta }{\text{T}}_{\text{m}}={\text{T}}_{\text{dev}}-{\text{T}}_{\text{m}}$$ T_dev is the temperature at which the dielectric permittivity begins to deviate from the C-W equation and T_m is the temperature corresponding toat which the maximum permittivity (Tang et al., 2005).

Close

Fig. 3

Thermal variation of the inverse of the real part of the relative permittivity (ε′⁻¹) at 1 kHz for BBiTZS ceramic.

Export to PPT

Table 4 Dielectric characteristics of the BBiTZS composition at 1 kHz.

parameters	Tm (°C)	T0(°C)	Tdev (°C)	C.10−5(°C)	ΔTm (°C)	ΔTmf (°C)	Δε′r/ε′r	${\epsilon }_{\mathit{rmax}}^{\prime }$	tg (δ)	γ
BBiTZS	−88	−261	83	1.09	171	16	0.045	345	0.016	1.80

It should be noted the low values of T_m and ε′_rmax are certainly related to the dual association of Zr⁴⁺ and Sn⁴⁺ in the B site of BBiTZS. Indeed, it is well-known that the addition of Zr to BaTiO₃ leads to a relaxor effect and the temperature T_m decrease as the content of Zr increases (Ciomaga et al., 2007). The same behavior occurs when Sn substitutes Ti in BaTiO₃ (Jiwei et al., 2005; Xiaoyong et al., 2003). As the amount of Sn increases, the temperature of the ferroelectric-paraelectric transition decreases and the phase transition becomes more diffuse over a broad temperature range. Consequently, the combination of Zr and Sn to replace partially Ti leads to a synergistic phenomenon causing a large decrease in T_m. We observe for BBiTZS where the high quantity of Zr (0.30) gives rise to a very low temperature (T_m = − 88 °C), far from room temperature.

3.4.4

3.4.4 Diffuse phase transition parameters of BBiTZS

Likewise, in order to quantify the diffuse phase transition nature within BBiTZS, we determined the diffusivity coefficient (γ) from the following relation (Uchino and Nomura, 1982):

(5)

$$\frac{1}{{\epsilon }_r^{\prime }}-\frac{1}{{\epsilon }_{rmax}^{\prime }}=\frac{{(T-T_m)}^{\gamma }}{C}\left(T\text{>}{T}_m\right)$$ with 1 ≤ γ ≤ 2 ; γ = 1 and 2 are the limit values corresponding respectively to a normal ferroelectric behavior and to a completely disordered (i.e. relaxor system). This expression corresponds to the modified C-W law that allows the description of the diffuse character in the paraelectric phase. The curves of ln (1/ε_r − 1/ε_rmax) versus ln (T − T_max) at 1 kHz is shown in Fig. 4. A linear variation is obtained for this composition where the coefficient γ was computed by fitting the line; a γ value of 1.80 (Table 4) corroborates the diffuse nature of the phase transition for this composition.

Close

Fig. 4

Curves ln(1/ε_r − 1/ε_max) versus ln(T-T_max) at 1 kHz for BBiTZS composition.

Export to PPT

3.4.6

3.4.6 The Vogel-Fülcher (V-F) analysis

The strong frequency dependence of T_m is supported by the curve ln f vs. T_m (Fig. 5). The nonlinear behavior indicates that the data cannot be adjusted with the Debye equation. To analyze the principal relaxation features of this composition, the curve was fitted by the Vogel–Fülcher (V-F) equation (Viehland et al., 1991)

(8)

$$f=f_{0}exp\frac{E_a}{k_B(T_m-T_g)}$$ where f_o is the attempt frequency, E_a the average activation energy and T_g are the freezing temperature.

Close

Fig. 5

Plot of ln(f) as a function of Tm for BBiTZS composition (the symbols: experimental data; the solid curve: fitting to the Vogel– Fulcher relation).

Export to PPT

The V-F equation models the shift of T_m toward low values of frequencies. The close agreement of the data for BBiTZS with the V-F relationship suggests a relaxor behavior. The modeling parameters by the V-F relation confirm this assertion and attribute this effect to the presence of polar nano regions in the crystal lattice. These adjusted parameters are: E_a = 0.112 eV, T_g = 195 K and f_o = 1.26 × 10⁷ Hz.

It is helpful to outline that the compounds BBiTZS have a similar behavior to that of Ba(Ti_1-yZr_y)O₃ solid solutions (0.26 ≤ y ≤ 0.40) (Laulhé, 2007). For these latter, it has been evidenced that the ZrO₆ and TiO₆ octahedra were deformed with a particular impact on the direction of local dipole moments carried by Ti⁴⁺ cations. A random distribution of the cations leads to a local polarization that would be at the origin of the frequency dispersion observed experimentally. All the ferroelectric features mentioned above are promising for photocatalytic applications.

3.5 Impedance spectroscopy

The impedance spectroscopy is performed to study the relationship between the microstructure and electrical properties in the temperature range (RT − 100 °C). Fig. 6 presents the frequency dependent real and imaginary part of the impedance at various temperature (22–80 °C) for Ba_0.925Bi_0.05(Ti_0.65Zr_0.30Sn_0.05)O₃. From the figure, we can observe the occurrence of the relaxation frequency for BBiTZS between 10⁴ and 10⁵ Hz. In the view to extract information on the resistive properties of the ceramic, we plotted the Cole-Cole plot of the impedance formalism at different temperatures (Fig. 6). We can observe from the plot, well-resolved semi-circles for the studied temperatures. Besides, the radius of the semi-circles decreases as the temperature increases, which demonstrates the semiconducting nature of our sample (Wang et al., 2017). By the use of a model (Fig. 6c), we could fit the collected data. Our proposed model includes a serie resistance (R₁), in addition to the bulk contribution (R₂, C₂). Besides, a departure from the ideal Debye behaviour was observed and better fitting results were obtained by introducing constant phase element (CPE) in the circuit (Benyoussef et al., 2020) and Table 5 regroups the electrical parameters. We can observe that the bulk resistance decreases as the temperature increases, which demonstrates the Negative Temperature Coefficient of Resistivity of our studied ceramic. Fig. 6d shows the logarithmic scale of the dc conductivity versus 1000/T for our sample. The plot is observed to be linear and follows an Arrhenius equation $\left(\sigma ={\sigma }_{0}exp\left(-\frac{E_{\alpha }}{K_{B}T}\right)\right)$ (Lanfredi et al., 2000). The activation energy of the system was found to be 0.016 eV.

Close

Fig. 6

(a) Real impedance Zʹ(f) curves, (b) Imaginary impedance Z″(f) curves of BBiTZS system at different temperatures, (c) Fitted Cole-Cole plots at different temperatures. The inset presents the model used to fit the BBiTZS system, (d) The Arrhenius plots of BBiTZS system. The activation energy extracted from the dc conductivity (at 10 kHz).

Export to PPT

3.6 Electrochemical and photocatalytic properties

The optical characterization of BBiTZS is elucidated from the diffuse reflectance. The photon energy (hν,) absorption coefficient (α) and gap (E_g) are related by the equation (Rekhila et al., 2018):

The exponent m = 2 and 0.5 are respectively for indirect and direct transitions. The intersection of the line (αhv)^0.5 with the abscissa axis yields the E_g value of the perovskite (2.00 eV) (Fig. 7). The defect states and grains boundaries have no influence on the thermo-power (S) unlike the conductivity (SM 1); the negative value implies that the free carriers are electrons. The non-dependence of S with temperature indicates thermal activation of the electronic concentration with a constant mobility (=360 µV K⁻¹), the S-value is compatible with an order of magnitude of the electrons density (N_D) of ∼10¹⁹ cm³ (see below).

Close

Fig. 7

The indirect (αhν)^0.5 and direct (αhν)² optical transitions of BBiTZS.

Export to PPT

The electrochemistry is of high importance for predicting the electrochemical reactions; the profile of the intensity-potential J(E) profile of BBiTZS (Fig. 8) shows a little hysteresis with a decomposition voltage greater than 3 V, characteristic of high over voltages. The reduction peak at −0.5 V on the reverse scan indicates an irreversible process. Below ∼−1.2 V, the current increases continuously with no plateau region, due to hydrogen liberation. The small exchange current J_o (0.2 μA cm⁻²) and the polarization resistance (=10.7 kΩ cm²) indicate a long-lived material (Fig. 8 Inset). For zero potential, the current density (J_o) is similar to an electron transfer rate constant.

Close

Fig. 8

The intensity–potential characteristic of BBiTZS in Na₂SO₄ electrolyte. Inset: the semi-logarithmic plot.

Export to PPT

The flat band potential (E_fb) is a key parameter in photoelectrochemistry, and it can be determined from the capacitance plot; its value (−0.56 V) is calculated by extrapolating the line (C_SC⁻² − E) to infinite capacitance (C_SC⁻² = 0). (Fig. 9). The energy diagram of the junction BBiTZS/solution, plotted from the photoelectrochemical characterization, predicts from a thermodynamic point of view the degradation of Rh B upon solar irradiation. The Rh B mineralization occurs by the reactive radicals OH^• and O₂^•− respectively in the valence band (VB = 2.04 V) and conduction band (CB = -0.49 V); both the levels OH^•/H₂O and O₂/O₂^•− are inside of the band gap (Roumila et al., 2016).

Close

Fig. 9

The Mott–Schottky plot of n-type BBiTZS in neutral solution.

Export to PPT

The EIS spectrum of the interface BBiTZS/Na₂SO₄ electrolyte was plotted once the free potential stabilized; it shows a semicircle at high frequencies ascribed to the charge transfer (SM 2). The impedance of the system is governed by the extension of the space charge region and the wide diameter is equal to the bulk resistance of the material R (27 kΩ cm²) and the center positioned at −9° below the real axis, suggests an electron hopping by overcoming a low potential barrier. CP) is inserted to account for the non-ideality of the capacity (Aroutiounian et al., 2007; Aroutiounian et al., 2006):

As noted in the introduction, Rh B is a hazardous dye and its elimination is strongly motivated. The perovskite BBiTZS is promising in such a case owing to its attractive features i.e electronic bands appropriate localized and a gap E_g sensitive to the UV_A of the solar irradiance. The following sequences can be used to involve Rh-B mineralization via a photo-catalytic process:

Rh B is attacked by oxidative species (O₂^•−, ^•OH), resulting in a partial mineralization to CO₂ and H₂O (Bagtache et al., 2016). Fig. 10a illustrates the degradation mechanism on BBiTZS, which does not produce intermediate products, indicating that their concentrations are below the detection threshold of the spectrophotometer. Fig. 10b gives the photocatalytic evolution of Rh B conversion (10 mg L⁻¹) at various catalyst quantities. The decay in absorbance at 554 nm (λ_max) over the illumination time indicates a first order kinetic with a half-life of 90 min and a conversion yield of 50%.

Close

Fig. 10

(a) The UV–Visible spectra of RhB solutions on BBiTZS over illumination time. (b) The kinetic plot ln C_o/C vs. time. Experimental conditions: catalyst dose: 1 g/L; [RhB]_o = 10 mg/L; V(solution) = 100 mL.

Export to PPT

A comparison of the dye degradation with other relevant studies is reported in SM 3. One can conclude that the photoactivity of acceptable, compared with some previous studies, and BBiTZS could be used in photocatalysis.