Structural, electrical and magnetic features of Kagomé YBaCo₄O₇ system

XRD pattern of the YBaCo₄O₇ sample recorded at ambient temperature shown in Fig. 1 was refined with the Rietveld refinement method using the fullproof program [22]. The refinement confirmed the single phase hexagonal structure with P6₃mc space group and no evidence for impurities was found within the detection limit of the X-ray diffractometer. The unit cell parameters a and c of prepared YBaCo₄O₇ are 6.2873 Å and 10.2533 Å respectively, that is very close to the reported values [23] which together with the other parameters calculated from this pattern are summarised in Table 1.

Table 1

YBaCo₄O₇ different XRD parameters; space group - P6₃mc, a = 6.2874 Å, c = 10.2533 Å, volume = 660.9953 ų, χ² = 1.351, R_wp = 19.5, R_exp = 16.7.

Scanning electron microscopy with EDAX accessory has been used to examine the microstructure and composition of the pellet at the centre which provides a better insight and accuracy of the grain development than at the top and bottom surfaces. The grain morphology is depicted in Fig. 2a. From the image it is clear that the hexagonal grains having the sizes of the order of 5 µm are well defined and distributed homogeneously across the structure. The fine network distribution among the grains is due to smooth mixing of different constituents in a fine network aggregate form and is a feature responsible for electronic and magnetic properties.

The compositional analysis confirmed by EDAX, displayed in Fig. 2b, shows the elemental composition for Y, Ba, and Co close to the stoichiometric value as shown in Table 2. The EDX patterns clearly give an indication of the structural phase purity in the prepared high quality sample that supports the XRD result also.

Table 2

Elemental composition details in wt.% and at.% of YBaCo₄O₇ obtained from EDX pattern.

Fig. 3 shows the Raman spectra of YBaCo₄O₇ (hexagonal space group P63mc) recorded at room temperature in the 1600 cm⁻¹ to 100 cm⁻¹ region. The group theory predicts the following irreducible representation as [24]: Γoptical=4A1+A2+4B1+1B2+5E1+5E2 $${\Gamma _{{\text{optical}}}} = 4 A_1+A_2+4B_1+1 B_ 2 + 5 E_ 1 + 5 E_ 2 $$ (1)

where A₁, E₁ and E₂ are Raman active (14 modes) and A₁ and E₁ are infrared-active (9 modes), while A₂, B₁ and B₂ are optically inactive.

The most intense mode at 542 cm⁻¹ could be the symmetric stretching mode of the CoO₄ tetrahedron and the lowest frequency mode at 255 cm⁻¹ arises from its bending vibrations. The strong broad overlapped Raman peak of O–Y–O stretching and bending vibrations of YO₆ of octahedral symmetry are obtained as included components at 636, cm⁻¹ 436 cm⁻¹, respectively. The only report available on Raman spectra of the bulk YBCo₄O₇ by Izquierdo et al. [25] and our Raman spectra are in good agreement with the broad band centred at 542 cm⁻¹[26].

The electrical response is depicted in Fig. 4 where we have shown the resistivity of YBaCo₄O₇ compound measured at magnetic fields of 0 and 8 Tesla (T). We have measured the temperature dependence of the resistance of YBaCo₄O₇ pellets upon cooling and warming cycles. There is no change observed between these two processes. The semiconducting behavior of these compounds, which is not affected by 8 T field, is clearly seen in the measured temperature range [27].

The suppression of any kind of transition is probably due to the presence of extra oxygen in the structure which tends to increase locally the coordination of the cobalt cations with formal oxidation state Co^2.25+ (δ = 0). This, in turn, creates locally a more disordered crystallographic structure for 114 compounds YBaCo₄O₇ which seems not to be compatible with the cobalt magnetic ordering [28]. Indeed, by increasing the hole concentration in the O_7+δ phase, the extra oxygen atoms would hinder the triangular arrangement for the magnetic moments of the tetrahedrally coordinated CoO4 since new polyhedra are created. This suppresses the transition. On the other hand, it is probable that a structural transition takes place at a local scale. However, if a structural transition had happened at a local scale, it would have been very difficult to detect by normal XRD or transport measurements.

The electrical transport mechanism was tested using different models which depict the temperature dependence of the electrical resistivity. The conventional Arrehenius plots: ρ<!-- ? -->(T) = Ce(E/kBT) $$\rho {\text{(}}T{\text{) = }}C{e^{{\text{(}}E/{k_B}T{\text{)}}}}$$ (2)

suggests simple thermal activation transport [29]. While considering the small polaron hopping conduction mechanism, the behavior of electrical resistivity versus temperature should be expressed as ρ(T)=CTe(E/kBT)${\rho (\text T) = \text{CTe}^{{\text{(E/k}}_{\text{B}}{\text{T})}}}$ [28]. Another approach is the Mott’s VRH model in which the dependence of resistivity on the temperature is given by [30]: ρ(T)=ρ0exp⁡(T0/T)1/4 $$\rho ({T}) = \rho _0 \exp (T^0/T)^{{\text{1/4}}} $$(3)

Here, ρ₀, T⁰ and n are constants. The quantity 1/4, depends on the dimensionality D and, in some compounds, on the temperature range. In the study of our sample, the resistivity data measured on the YBaCo₄O₇ pellets seemed to obey the VRH model. In Fig. 5 the fitting is shown for variable range hopping (VRH) model under 0 T and 8 T field. By plotting the experimental data ρ(T) as ρ(T) = C(T)exp(E/k_BT) in the temperature range of 300 K to 20 K, it is readily recognized that neither simple thermal activation nor small polaron hopping conduction describe well the electronic transport in the reported pellets. In the present study of YBaCo₄O₇ compound, only the VRH model is well fitted in the whole measured temperature range 300 K to 20 K, therefore we have shown only the fitting data for VRH model. The VRH model is promptly verified. From Fig. 5 it can be observed that a linear fit is obtained in all the compounds of this series, indicating that the Mott’s 3D-VRH mechanism is the most prevalent conduction mechanism in these compounds for this temperature range. The Mott’s parameter (T⁰) and the density of states (N(E_F)) were calculated for the samples studied in about 0 T and 8 T fields. The characteristic temperature T⁰ is related to the parameter ξ through the expression: kBT0=24/πN(EF)ξ3 $$ k_B T^0 = 24/\pi N({E_F}) \xi^3 $$ (4)

where N(E_F) is the density of states at the Fermi level [31]. By taking ξ≈c = 10.254 Å, N(E_F) was estimated for these polycrystalline samples. The fact that the electrical resistivity may be described by VRH model, points out to the presence of disorder induced localization of charge carriers in the samples. VRH mechanism normally occurs only in the low-temperature region (below room temperature), wherein the energy is insufficient to excite the charge carriers across the Coulomb gap. Hence, conduction is accomplished by hopping within a small region (∽k_BT) in the vicinity of E_F. In this region, the density of states remains almost constant [13]. The values are closer to the expected values for transition metal compounds.

The Mott’s parameter (T⁰) and the density of states N(E_F) were calculated for the sample studied in about 0 T and 8 T fields. In Table 1 the calculated values for T⁰ and N(E_F) are also reported. The typical value of T⁰ and N(E_F) are about 3.4 × 10⁸ K and 2.4 × 10¹⁷ eV⁻¹ cm⁻³, respectively.

The DC magnetic susceptibility (M/H) measurements for YBaCo₄O₇ were carried out by using SQUID magnetometer as shown in Fig. 6. In this paper, we have shown the magnetization of YBaCo₄O₇ compound measured at 1 Tesla (T) magnetic field. From 80 K to 300 K, the data do not obey a Curie-Weiss relationship. We have collected our data for zero field cooled (ZFC), field cooled (FC) and for field cooled warming (FCW), however during the analysis of the results we have not found any difference between FCW and FC data, as shown in the Fig. 6. Enlargement of ZFC, FC and FCW shown in the inset of Fig. 6, shows clear lack of hysteresis at TC here. The FC curve is noticeably different than the ZFC data. The compound YBaCo₄O₇ shows ferromagnetic-like transition that appears at about 91 K with large irreversibility between ZFC and FC curves as shown in Fig. 6. This type of behavior has also been reported in Huq et al., [13]. The ZFC curve also shows another magnetic transition at TC = 40 K for this pure sample.

In the present literature the magnetization data has been reported but still it is not clear. It is obvious that the sample YBaCo₄O₇ exhibits a complicated inhomogeneous magnetic phase. Interestingly, we have also found ferromagnetic transition at T_c = 80 K. Super-exchange couplings build up a geometrically frustrated substructure in the planes of Co2 and as the temperature is lowered, the combination between super-exchange and double-exchange causes the magnetic system to degenerate. Further, on cooling, the strength of the double-exchange diminishes, since the electrons do not have enough energy (heat) to hop between the Co atoms. At the same time, the super-exchange starts to couple in more directions and dominates at lower temperature. On the basis of above discussion it is established that the overall magnetic interactions can be described as those of a glassy disordered antiferromagnetic, where doubleexchange phenomenon dominates at higher temperatures and super-exchange at the lower ones and they compete through the whole measured temperature range on the frustrated lattice. It has also been exhibited for other systems, but the magnetic order in those cases was confined within the planes, prohibiting long-range order to occur. To get additional insight into this high state field we have carried out isothermal magnetization at three different temperatures for all the samples at 3 K, 50 K and 300 K and the results are shown in Fig. 7. The M(H) curve measured for all the three assumed temperatures is linear, and an irreversible characteristics have been found at all magnetic fields measured at about 0 T and 8 T. The M (H) measured data confirms that no hysteresis of magnetization has been found in these compounds. The magnetization value at these three different temperatures is small and does not change for the wide range of temperatures.

The effective moment value calculated from the fitting procedure for pure YBaCo₄₇ sample is 11.68 μ_eff (not shown here), which is larger than the value given by Lu et al. [33], i.e. 9.8 μ_eff. This variation is due to the difference in measured temperature range. In the present work we performed the measurements below 300 K and in [33] in the range of 300 to 400 K. Such large μ_eff value indicates that there exist some short range magnetic interaction in the high temperature in our measured paramagnetic (PM) range.