The nanosized ceramic samples of Ga_1−xZn_xFeO₃ (GZFO, x = 0, 0.01, 0.05 and 0.1) were prepared through a sol-gel method using nitrates as metal precursors. First, gallium nitrate [Ga(NO₃)₃×xH₂O], iron nitrate [Fe(NO₃)₃×9H₂O] and zinc nitrate [Zn(NO₃)₂×6H₂O] in stoichiometric proportions were dissolved in distilled water. Citric acid (C₆H₈O₇) in 1:1 molar ratio with respect to the metal nitrates was added to the solution as a complexant. The clear solution was dried at 120˚C to form a gel, and then the obtain gel was burned until the combustion process was completed. After that, the precursory powders were reground and sintered at 800˚C for 12 h.

The crystalline structure and the phase purity of the obtained samples were characterized with a typical X-ray diffraction (XRD), acquired by a Bruker D8 Advance X-ray diffractometer Cu K_α1 radiation. The XRD data were fit using the General Structure Analysis System (GSAS) Rietveld refinement program. Micrographs of the samples were analyzed using a field emission transmission electron microscope (TEM) operated at 120 kV. The temperature- and field-dependent magnetizations were measured with a Quantum Design superconducting quantum interference device (SQUID) magnetometer. For the magnetocapacitance (MC) measurements, the powders were pressed into the disk (5 mm in diameter and 0.5 mm in thickness) under a pressure of 1.5 GPa and then coated with 100 nm thick silver layers on both the top and the bottom sides of the disk as electrodes. A capacitance bridge (Agilent E4980A Precision LCR meter) hooked to a probe station with a closed-cycle low temperature system was used. Room temperature ferroelectric measurements were performed using a Sawyer-Tower circuit and a commercial FE test system (TF Analyzer, aixACCT Co.).

The RT XRD patterns of the Ga_1−xZn_xFeO₃ samples with x = 0, 0.01, 0.05, and 0.1 are given in Figure 1. Based on the standard reference (JCPDS Files No.26-0673), all the observed peaks of the GZFO samples can be indexed on the basis of an orthorhombic unit cell of space group, Pc2₁n, suggesting that all samples without any secondary phase. The GZFO samples of various Zn-content, namely, 0, 1, 5, and 10 at % designated as GZFO1, GZFO2, GZFO3, and GZFO4, respectively. The lattice parameters were determined by Rietveld refinement method and listed in Table 1. As an example, the profile fitting of GZFO2 sample using GSAS Rietveld refinement program is shown in Figure 2. The fact that the value of the unit cell volume and the lattice parameters a, b, and c of GZFO increase systematically with increasing Zn-content (x) implies that the Zn-ions do replace the Ga-ions. This could be attributed to the larger ionic radius of Zn²⁺ ion (~0.74 Å) than that of Ga³⁺ ion (~0.62 Å), which clearly confirms that the zinc is successfully doped in the lattice of GFO.

According to the calculation using a Debye-Scherrer equation [13] , the average grain size of the GZFO samples are about 30 nm. The grain sizes were also confirmed by TEM. As an example, the TEM image of the GZFO2 sample is shown in the inset of Figure 2. In order to measure the orthorhombic distortion [14] , we have calculated the values of c/a and 2 c/b of each GZFO samples (shown in Table 1). It is evident that the Zn substitution affects the structure of the pristine sample by increasing of the orthorhombic distortion of the perovskite lattice parameters of nanocrystalline GFO.

The temperature dependence of magnetization, M(T), was measured in a magnetic field of 1000 Oe under the condition of field-cooled (FC). Figure 3 displays the temperature-dependent FC magnetization for the GZFO samples with different Zn concentrations. The magnetic transition temperature (T_C) is obtained for the GZFO samples as the differential minimum of the FC curves. The graphical representation of dM/dT versus temperature curves gives the T_C for the GZFO1 and GZFO4 samples, as shown in the inset of Figure 3. As can be seen, a typical FM to paramagnetic phase transition is observed for the GZFO1 sample, and the T_C shifts to the higher temperatures and it equal to 320 K for the GZFO4 sample. The value of T_C is rather high compare to the previously reported T_C of

228 K in bulk GFO [4] . The results could be explained on the basis of variation of Ga/Fe occupancy due to the disorder effects, which in turn alters the bond parameters of Fe-O-Fe and hence the strength of the magnetic interactions [4] [7] .

The field-dependent magnetization M(H) curves of all the GZFO samples measured at 5 K are presented in Figure 4. And also, the M(H) loops for all the GZFO samples observed at RT are shown in the inset of Figure 4. On can observe from the M(H) data that all the samples exhibit pinched-like hysteresis loop behavior of two phase systems as indicated by the fact that the loops could be decomposed in two additive loops with different coercivities. A similar type of M-H loop is reported by Kim et al. [5] , in GFO and AlFeO₃, and explained it in terms of the existence of hard and soft magnetic phases. Because of the fact that the four cation sites have different magnetic anisotropy energies in GFO, namely, the octahedral sites (F1, F2, and G2) have strong anisotropy, whereas the tetrahedral site (Ga1) is weak anisotropic, and hence these act as hard and soft magnetic phases, respectively [15] . In addition, the magnetization has been enhanced due to doping of Zn²⁺ in GFO compared to that of the pristine sample. The cause of magnetic enhancement could be explained from the modification exchange interaction due to the substitution of larger size cations of Zn²⁺ in place of relatively smaller size cations of Ga³⁺. Therefore, it is expected that with the variation of Zn-content in GZFO samples, there is a substantial change in cation distribution and hence in the magnetic properties.

To investigate the presence of spontaneous polarization, the ferroelectric polarization-electric field (P-E) loops for the Zn-doped GFO samples were measured at RT are shown in Figure 5(a). Here, the P-E loops taken at an applied maximum electric field of 30 kV/cm for each samples. It is found that all three samples show clear hysteresis loop behaviors that confirms the existence of ferroelectricity. Figure 5(a) also shows that maximum polarization and the tendency

of saturation increase with the increase of Zn concentration. From the P-E loops, we see that the values of the maximum polarization of GZFO2, GZFO3 and GZFO4 samples are 0.08, 0.16 and 0.23 μC/cm², respectively. The enhancement of electric polarization may be ascribed to the increase in distortion of the crystal lattice due to the substitution of Zn²⁺-cations in place of Ga³⁺-cations in the doped samples.

In order to know the coupling between the magnetic and ferroelectric orderings, we have measured the dielectric constant of the GZFO samples in presence of an externally applied magnetic field. The change in dielectric constant with the applied magnetic field is one of the important ways to study the coupling effect [16] [17] . Figure 5(b) shows the variation of MC which is calculated by the expression [17] , MC = [ε¢(H, T) - ε¢(0, T)]/ε¢(0, T). Here, ε¢(H, T) and ε¢(0, T) represents the dielectric constants in presence and absence of external magnetic field (H), respectively. In such GZFO sample, the dielectric constant is strongly perturbed by the external magnetic field. It is confirmed the presence of FM and ferroelectric coupling, that is, the multiferroic state in all the GZFO samples at RT. It is to be noted that the value of MC increases in the doped samples for different applied magnetic fields. And also, the values of MC for nanosized GZFO ceramic samples are higher than that observed for bulk GFO [6] . To further probe the electrical leakage effect, the leakage current densities (J) were measured for all the GZFO samples at RT as shown in Figure 6. It shows that the Zn substitution in GFO allows decreasing the leakage current density by two orders of magnitude, as shown in the inset of Figure 6. We attribute the decrease of the leakage currents to the substitution of Fe²⁺ with Zn²⁺, decreasing the hopping possibilities between Fe²⁺ and Fe³⁺ within the structure; a similar behavior has been reported by Lefevre et al. [18] .