Structure and Magnetic Properties of the Ti-Doped Pyrochlore Molybdate Y₂Mo_2(1-x)Ti_2xO₇

Abstract

The structural, heat capacity, and magnetic properties of Y₂Mo_2(1-x)Ti_2xO₇ were investigated in this work. Y₂Mo_2(1-x)Ti_2xO₇ maintains the pyrochlore structure, but lattice constants are observed to decrease continuously with the Ti⁴⁺ doping. The magnetic measurements indicate a magnetic phase transition at low temperature for all Ti-doped samples, and the Ti⁴⁺ doping suppresses the transition temperature. This is ascribed to the diluted effect of nonmagnetic Ti⁴⁺ substitution of magnetic Mo⁴⁺. The heat capacity data confirm that the low-temperature magnetic phase is spin glass in the Ti⁴⁺-doped Y₂Mo_2(1-x)Ti_2xO₇ system. The M(H) data reveals the coexistence of antiferromagnetic and ferromagnetic interactions at low temperatures.

1 Introduction

Pyrochlore oxides have attracted a lot of scientific interest over the past two decades [1,2,3,4]. In pyrochlore oxides A₂B₂O₇, the rare earth cations and the transition metal cations generally occupy A and B sites, respectively. And the A- and B-site cations form the three-dimension array of corner-shared tetrahedra, respectively. When magnetic metal cations with the inherent antiferromagnetic (AFM) interaction occupy the vertexes of tetrahedra, spins encounter the magnetic frustration. This spin frustration originating from the structure leads to highly degenerate ground states. In some geometrically frustrated antiferromagnets, the degenerate magnetic ground states are lifted by the internal or external factors such as the lattice distortion, the strain, the next-nearest-neighbor interaction, and the ionic doping, which generates a rich variety of magnetic phases including spin glass, spin liquid, spin ice, antiferromagnetism [5,6,7].

In Y₂Mo₂O₇, nonmagnetic Y³⁺ and magnetic Mo⁴⁺ ions occupy A and B sites, respectively. The magnetic interaction between the nearest neighbor Mo⁴⁺ ions at B sites is AFM, so the magnetic Mo⁴⁺ ions in Y₂Mo₂O₇ meet the geometrical spin frustration [8]. In the past years, physical properties of Y₂Mo₂O₇ and its doping compounds have been investigated in some references [9,10,11]. It has been established that Y₂Mo₂O₇ behaves as a nearly ideal spin glass from nonlinear susceptibility, heat capacity, AC susceptibility, elastic and inelastic neutron scattering, muon-spin relaxation, and neutron spin echo data [10,11,12,13]. The other metal ions doping in Y₂Mo₂O₇ would change magnetic property due to the change of structure and generation of chemical disorder [14]. The magnetic property in the Ti⁴⁺-doped Y₂Mo₂O₇ was investigated in reference [15]. In this work, the effect of Ti⁴⁺ doping in Y₂Mo₂O₇ was investigated. The Ti⁴⁺ doping does not change the structure except slightly increasing the lattice constant in the system. Spin glass is confirmed to persist in the Ti⁴⁺-doped Y₂Mo₂O₇, but the transition temperature is suppressed with the Ti⁴⁺ doping.

2 Experimental

Polycrystalline samples of Y₂Mo_2(1-x)Ti_2xO₇ were prepared by using the standard solid-state method. Firstly, the stoichiometric reactants of Y₂O₃ (99.99 wt%), MoO₂ (99 wt%), and TiO₂ (99.99 wt%) were thoroughly mixed. Then the mixtures were pressed into sheets and pre-fired at 1073 K under the flowing N_2. Finally, they were sintered at 1523 K under the same atmosphere with the intermediate grinding. The structure was checked by using the x-ray diffractometer (XRD) with the Cu K_α radiation at room temperature. The XRD data were recorded in the 2θ range from 20° to 80°. The XRD data have been analyzed using Rietveld refinement program (GSAS). The FRIR and Raman spectra were employed to check the change of microstructure in the Ti⁴⁺-doped samples. The magnetic properties were measured by employing a superconducting quantum interference device magnetometer. Temperature dependence of magnetization was measured in the zero-field-cooling (ZFC) and the field-cooling (FC) modes under the magnetic field of 100 Oe. The AC susceptibilities at different frequencies were measured at the zero DC magnetic field and a constant AC field of 3.5 Oe from 4 to 200 K. The heat capacity at low temperatures was measured by using physical properties measurement system.

3 Results and Discussion

The experimental and the Rietveld refined XRD data of Y₂Mo_2(1-x)Ti_2xO₇ with x = 0.05, 0.10, 0.20, and 0.30 at room temperature were shown in Fig. 1. All Y₂Mo_2(1-x)Ti_2xO₇ samples present in the cubic phase with Fd3m symmetry and no impurity peaks are observed, which indicates that the Ti⁴⁺ doping does not change the crystal structure in this series of samples. Compared with the standard XRD spectra of Y₂Mo₂O₇, the (222) peaks of Y₂Mo_2(1−x)Ti_2xO₇ with x = 0.05, 0.10, 0.20, and 0.30 shift − 0.004°, + 0.053°, + 0.144°, and + 0.155°. For x = 0.05, the (222) peak unexpectedly and slightly shifts to the low degree, which may ascribe to measurement inaccuracy from the large measurement step (0.033° per step). The Ti⁴⁺ doping shifts the (222) peak to the high degree, which means that the lattice constant becomes small. The experimental XRD data for all samples is well simulated by the Rietveld refinement. As shown in Table 1, the lattice constant (D) for Y₂Mo_2(1-x)Ti_2xO₇ decreases monotonically upon the substitution of Ti⁴⁺. This is due to that the ionic radius of Ti⁴⁺ (0.605 Å) is smaller than that of Mo⁴⁺ (0.65 Å). The bond lengths between different ions from Rietveld refinements for Y₂Mo_2(1−x)Ti_2xO₇ are shown in Table 2. When Ti⁴⁺ substitutes Mo⁴⁺, the bond lengths between most different ions in Y₂Mo_2(1−x)Ti_2xO₇ vary slightly (≤ 3.3%) except the Ti–O bond (up to 12%). The large change of the Ti–O bond length is related with the microstructure crystalline distortion resulting from the different ionic radii between Mo⁴⁺ and Ti⁴⁺.

Fig. 1

The XRD data along with Rietveld analysis are shown for Y₂Mo_2(1−x)Ti_2xO₇ with x = 0.05, 0.10, 0.20, and 0.30

Table 1 Structural parameters from the Rietveld refinement of XRD data for Y₂Mo_2(1−x)Ti_2xO₇ with x = 0.05, 0.10, 0.20, and 0.30 at room temperature

Table 2 The bond lengths between different ions from Rietveld refinements for Y₂Mo_2(1−x)Ti_2xO₇ with x = 0.05, 0.10, 0.20, and 0.30

Figure 2 shows magnetization as a function of temperature [M(T)] for Y₂Mo_2(1-x)Ti_2xO₇ measured under 100-Oe magnetic field. The ZFC and FC M(T) curves are coincident in the high-temperature region and show irreversible at low temperatures. The irreversibility temperature (T_irr) of Y₂Mo_2(1-x)Ti_2xO₇ is 20.2, 18.1, 14.1, and 12.0 K for x = 0.05, 010, 0.20, and 0.30, respectively. The inset of Fig. 2 exhibits that T_irr decreases almost linearly with the doping content of Ti⁴⁺. Obviously, the doping of Ti⁴⁺ suppresses the irreversibility temperature. In Y₂Mo₂O₇, the observance of magnetic irreversibility is considered to originate from the thermodynamic spin glass [16, 17]. The generation of spin glass in Y₂Mo₂O₇ is also confirmed by neutron powder diffraction, the heat capacity, and density functional theory calculations, and suggested to be the results of the spin-orbital coupling and random fluctuations in the Mo environment at the local level [18]. However, magnetic irreversibility behavior is also found in the antiferromagnetic and ferromagnetic systems. So, it could not be concluded that spin glass occurs in the Ti⁴⁺-doped Y₂Mo_2(1-x)Ti_2xO₇ system. Temperature dependence of the inverse magnetic susceptibility [χ⁻¹ = (M/H)⁻¹] is shown in Fig. 3. The curves are fitted through the Curie-Weiss law χ = C/(T − θ_CW), where C is the Curie constant and θ_CW is the Curie-Weiss temperature. The obtained θ_CWs for all samples are all negative, which indicates that the main magnetic interaction in Y₂Mo_2(1-x)Ti_2xO₇ is antiferromagnetic.

Fig. 2

Temperature-dependent magnetization of Y₂Mo_2(1-x)Ti_2xO₇ under the field of 100 Oe. Filled and unfilled symbols represent the ZFC and FC data, respectively. The inset shows the variation of irreversibility temperature with Ti⁴⁺ substitution

Fig. 3

Temperature-dependent inverse susceptibility deduced from magnetization data of Y₂Mo_2(1−x)Ti_2xO₇ with x = 0.05, 0.10, 0.20, and 0.30. The solid lines show the fitting curves with the Curie-Weiss law

To further clarify the magnetic state in the Y₂Mo_2(1−x)Ti_2xO₇ system, we studied the isothermal magnetization as a function of a magnetic field [M(H)] measured at 5 K for Y₂Mo_2(1−x)Ti_2xO₇ (Fig. 4). The M(H) curves show almost linear behavior, which reflects the strong AFM interaction in this series of samples. But the M(H) curves also show a slight magnetic hysteresis in the low-field region. The typical M(H) curve of x = 0.20 as shown in the inset of Fig. 4. The magnetic hysteresis indicates that there is a weak ferromagnetic (FM) interaction in the system. Figure 5 shows the corresponding Arrott plot (M² vs H/M) from the M(H) data [19]. For the materials with spontaneous magnetization, the Arrott plot will give a positive intercept in the high-field regime. On the other hand, a negative intercept in the Arrott plot implies an antiferromagnetic state [19, 20]. As seen in Fig. 5, the intercepts for all samples are negative, which indicates that the AFM interaction is predominated rather than the FM interaction in Y₂Mo_2(1-x)Ti_2xO₇. Therefore, the AFM interaction is predominated but the weak FM interaction also exists in Y₂Mo_2(1-x)Ti_2xO₇.

Fig. 4

Magnetic field–dependent magnetization of Y₂Mo_2(1-x)Ti_2xO₇ at 5 K. The inset shows the detailed M(H) curves of Y₂Mo_1.40Ti_0.60O₇

Fig. 5

The Arrott plot (M² vs H/M) of Y₂Mo_2(1-x)Ti_2xO₇

The coexistence of the AFM and FM interactions in this series of samples would be possible to induce spin glass in the system. To demonstrate whether or not spin glass occurs in Y₂Mo_2(1-x)Ti_2xO₇, we further measured temperature dependence of AC susceptibility under zero DC magnetic field and a constant AC field of 3.5 Oe from 4 to 200 K with different frequencies (f) of 10 Hz and 100 Hz. The real part of AC susceptibility curves [χ’(T)] with different frequencies are shown in Fig. 6. A pronounced peak is observed in the χ’(T) curves for all samples, and the corresponding temperature is the transition temperature. When frequency change from 10 to 100 Hz, the χ’(T) curve at 100 Hz is coincident with that of 10 Hz and no peak shift is found for each sample. In a conventional spin glass, when increasing frequency, peak shifts to a higher temperature [8]. But in the Y₂Mo_2(1−x)Ti_2xO₇, no peak shift is found, which is unlike the conventional spin glass. In some spin glass system, the behavior of peak shift could not be detected due to that this shift is smaller than the measurement accuracy. Therefore, the spin glass could not be excluded. The transition temperature obtained from the AC susceptibility is 22, 20, 16, and 14 K for x = 0.05, 0.10, 0.20, and 0.30, respectively.

Fig. 6

The real part of the AC susceptibility Y₂Mo_2(1−x)Ti_2xO₇ with x = a 0.05, b 0.10,c 0.20, and d 0.30 at the frequency of 10 Hz and 100 Hz

We further made the heat capacity measurements as a function of temperature [C_p(T)] for Y₂Mo_2(1−x)Ti_2xO₇ with x = 0.05, 0.10, 0.20, and 0.30, shown in Fig. 7. The measured C_p(T) curves of Y₂Mo_2(1−x)Ti_2xO₇ (x = 0.05, 0.10, 0.20, and 0.30) is similar to that of Y₂Mo₂O₇ [18]. To get the magnetic specific heat, Y₂Ti₂O₇ was used as a subtraction from the lattice contribution at low temperatures because Y₂Mo₂O₇ presents a similar crystalline structure to Y₂Ti₂O₇. Figure 8 shows temperature dependence of the magnetic specific heat [C_m(T)] of Y₂Mo_2(1−x)Ti_2xO₇ (x = 0.05, 0.10, 0.20, and 0.30). A broad peak is observed at the transition temperature for all samples. Especially, the magnetic specific heat in the low-temperature region exhibits linear behavior. This phenomenon suggests a constant density of states of the low-temperature magnetic excitations and it is also claimed to be a common feature of spin glass [11, 21, 22]. A broad magnetic specific heat also extends up to high temperatures, which indicates that short-range order contributions persist at the high temperatures [11]. This also confirms spin glass. The coexistence of AFM and FM interactions in the frustrated magnet may be the origin of spin glass in this series of samples.

Fig. 7

The measured specific heat of Y₂Mo_2(1−x)Ti_2xO₇ with x = 0.05, 0.10, 0.20, and 0.30 at low temperatures

Fig. 8

Magnetic specific heat of Y₂Mo_2(1−x)Ti_2xO₇ with x = 0.05, 0.10, 0.20, and 0.30 at low temperatures. The solid lines show the linear fitting curves in the low-temperature region

4 Conclusion

In summary, we have investigated the structural, heat capacity, and magnetic properties of the pyrochlore Y₂Mo_2(1-x)Ti_2xO₇. No structural phase transition is induced, but the lattice constant decreases continuously with Ti⁴⁺ doping. Spin glass persists in the Ti⁴⁺-doped Y₂Mo₂O₇ in the low-temperature region. The Ti⁴⁺ doping suppresses the transition temperature. This is ascribed to the diluted effect of nonmagnetic Ti⁴⁺ substitution of magnetic Mo⁴⁺. The coexistence of AFM and FM interactions in the frustrated magnet may be the origin of spin glass.