Hydrothermal synthesis and multiferroic properties of Y₂NiMnO₆

Abstract

The double perovskite Y₂NiMnO₆ was prepared by a mild hydrothermal method. The room temperature powder X-ray diffraction analysis shows that it crystallizes in the monoclinic perovskite lattice with space group P2₁/n. Analysis by bond valence sum and X-ray photoelectron spectroscopy reveal its oxide states to be Ni²⁺/Mn⁴⁺. Magnetic behaviours were characterized by DC and AC magnetic susceptibility, as well as magnetization–hysteresis. After the as-synthesized sample was annealed at 1273 K, the order of the arrangement between Ni²⁺ and Mn⁴⁺ increased. Close to the magnetic ordering temperature, a dielectric anomaly was observed at 84 K. The ferroelectric polarization of about 35 μC m⁻² at 77 K was determined by Positive-Up-Negative-Down method, which is induced by up-up-down-down magnetic arrangement as expected theoretically.

---

Introduction

Multiferroic materials that are both ferroelectric and magnetic have attracted much research interest due to their potential applications in multi-state data storage and electric-field controlled spintronics.^1,2 Among all the well-studied multiferroic systems, a large number of them are transition metal oxides, especially with perovskite-related structures.^3–7 According to the “d⁰-ness rule”, transition metal ions having partially filled d shells cannot induce the ferroelectricity directly. Their magnetic orders can sometimes break the spatial inversion symmetry and produce a spontaneous ferroelectric polarization (P) so that the problem of chemical incompatibility for achieving simultaneous ferroelectric and magnetic orderings could be overcome. For example, non-collinear spiral magnetic order in Tb(Dy)MnO₃ is responsible for the generation of spontaneous polarization below the corresponding magnetic transition temperature (T).^3,8 The reverse Dzyaloshinskii–Moriya interaction between non-collinear canted spins in Sm(Y)FeO₃ also gives rise to ferroelectric polarization.^6,9 The exchange striction interaction between Gd(Dy) and Fe ions in Gd(Dy)FeO₃ results in the emergence of ferroelectricity.^10,11 More recently theoretical calculations suggest that some double perovskites are multiferroics, in which collinear up-up-down-down magnetic arrangement along the alternating B–B′ chains can break the space inversion symmetry via an exchange striction mechanism and produce a net ferroelectric polarization along the chain direction.¹² Similar up-up-down-down (↑↑↓↓) spin configuration driven ferroelectricity has also been observed in the Ising chain magnet Ca₃MnCoO₆ and double perovskite R₂CoMnO₆ (R = Y and Lu).^13–15

Recently, multiferroicity has been predicted to exist in the low temperature magnetic phase of double perovskite Y₂NiMnO₆ with the up-up-down-down spin configuration.¹² Booth et al. reported a synthesis by solid state reaction, but the dielectric constant was only measured at room temperature.¹⁶ Maiti et al. synthesized nanocrystal Y₂NiMnO₆ by sol–gel method.¹⁷ However, the ferroelectric transition was observed at 537 K, far above the Curie temperature. This is probably due to the possible existence of mixed valence Mn³⁺/Mn⁴⁺ or chemical inhomogeneity in the nanocrystal Y₂NiMnO₆. In contrast, hydrothermal synthesis has been successfully applied to prepare advanced solid materials, advantageous to form perfect crystals and introduce specified oxidation states into a solid.^18,19 In this study we report hydrothermal synthesis and characterization of the double perovskite Y₂NiMnO₆ and confirm the existence of ferroelectric transition occurring at T_c = 84 K which agrees well with the magnetic transition temperature below which an up-up-down-down spin configuration presents.

Experimental section

Material synthesis

In a typical synthetic procedure, 10 mL Y(NO₃)₃ (0.4 mol L⁻¹), 5 mL NiSO₄ (0.4 mol L⁻¹) and 0.1739 g MnO₂ were initially mixed in a beaker by continuously stirring to form a solution. Then, KOH was added on stirring and cooling to room temperature. The final mixture was transferred into a 45 mL Teflon-lined stainless steel autoclave with a filling capacity of 60%. Crystallization was carried out under autogenous pressure at 533 K for 3 days. After the autoclave was cooled down and depressurized, the sample was washed thoroughly with distilled water and sonicated by the direct immersion of a titanium horn (Vibracess, 20 kHz, 200 W cm⁻²). The dark crystals were obtained at the bottom of the beaker.

Material characterization

Sample compositions were determined by inductively coupled plasma spectroscopy (ICP) on a Perkin-Elmer Optima 3300DV spectrometer. The X-ray diffraction (XRD) patterns were obtained from 20° to 120° in a step of 0.02° with a counting time of 2 s per step on a Rigaku D/Max 2500 V/PC X-ray diffractometer with Cu Kα radiation (λ = 1.5418 Å) at 40 kV and 200 mA. The morphology of the sample was checked with a JEOL JSM-6700F scanning electron microscope (SEM). The valence states of the sample were determined using X-ray photoelectron spectroscopy (XPS, ESCALAB250, America). DC and AC magnetic measurements were performed in a superconducting quantum interference device vibrating sample magnetometer (Quantum Design). For electric measurements, the sample was ground, pelletized and then annealed at 1273 K for 10 hours. The silver paste was attached on the two faces of the pellet as electrodes. Capacitance measurement was carried out with LCR meter Agilent E4980A using homemade setup. The polarization–electric field (P–E) hysteresis loop was measured using Positive-Up-Negative-Down (PUND) method by a Radiant Precision II.

Results and discussion

Synthesis

In hydrothermal process, precursor, mineral reagent, alkaline concentration, appropriate reaction temperature and time are key factors for the growth of transition metal oxides.^19,20 In one respect, the amount of alkaline is a key factor in producing the double perovskite Y₂NiMnO₆. In our specific case, when the alkalinity was less than 3 M, we obtained a mixture containing target crystals and impurities including Y(OH)₃, NiO, and K_xMnO₂. In the other respect, once alkalinity is above 12 M, the high alkalinity will not be able to remove small amount of NiO in the products. When alkaline concentration was only within the range between 3 and 10 M, pure dark crystals could be attained. KOH not only maintains the alkalinity in the solution but also acts as mineralizing agent. We also tried to use NaOH or the combination of KOH and K₂CO₃ as mineralizer. But impurities mentioned above cannot be removed. Therefore, pure KOH is necessary to obtain the desired products. In addition, the reaction temperature and time are responsible for the formation of the title compound. When the temperature is below 533 K, and time over 5 days, Y(OH)₃ and NiO impurities were also present in the products. These facts indicate there is competition between thermodynamically and kinetically stabilized phases. Zhang et al. synthesized R₂NiMnO₆ (R = Pr, Sm and Ho) using MnCl₂ as manganese source which is not favorable for the formation of Y₂NiMnO₆ with smaller rare earth radius.²¹ We also tried to substitute MnCl₂ & KMnO₄ at a molar ratio of 1 : 4 for MnO₂ as manganese source, which results only in nonstoichiometric sample with the Y : Ni : Mn ratio 1 : 0.36 : 0.46.

Structure and thermal stability analysis

Fig. 1a shows the powder X-ray diffraction (XRD) pattern of the as-grown sample. The XRD peaks can be indexed as a single phase monoclinic structure with space group P2₁/n. The Rietveld refinement was done with GSAS/EXPGUI system.^22,23 The lattice parameters obtained from the analyses (Table 1) are in good agreement with those reported previously.¹⁶

Fig. 1 Rietveld refinement of the XRD powder patterns of (a) Y₂NiMnO₆, (b) Y₂NiMnO₆ annealed at 1273 K. Orange circles, black lines, blue lines and red lines show the observed, calculated, background and the difference patterns respectively. The sticks correspond to Bragg positions. (c) The magnetic (left) and crystal (right) structure of Y₂NiMnO₆.

Table 1 Lattice parameters, structure refinement parameters and bond valence sum (BVS) of Y₂NiMnO₆ with P2₁/n space group

	As grown	After annealing
Lattice parameters
a	5.2247(2)	5.2192(1)
b	5.5739(2)	5.5485(1)
c	7.4863(3)	7.4781(1)
â	89.788(3)	89.666(1)
V	218.00(1)	216.60(1)
Rwp	0.0788	0.0569
Rp	0.0578	0.0441
÷^2	4.804	2.584
Bond valence sum
BVS Ni	1.89	2.12
BVS Mn	4.16	4.05
BVS Y	3.1	3.05

---

To investigate the thermal stability and whether the degree of order between Ni²⁺ and Mn⁴⁺ could be enhanced by annealing, the samples were annealed at selected temperatures within the range of 573–1673 K. As shown in Fig. 2a, there is considerable Y₂O₃ in the product when annealing temperature was above 1573 K, but only minor NiO impurity at 1473 and 1373 K. When the annealing temperature was between 773 K and 1273 K, no impurities can be observed but the pattern shifts to higher angles. For annealing temperature at 573 K, there is no obvious change of XRD pattern from that of as-grown sample, implying no evidence of structure change.

Fig. 2 (a) Room temperature XRD patterns of the as-grown sample and the samples annealed up to 1673 K. Inset shows the M–H hysteresis loops at 4 K of as-grown sample (black), sample annealed at 773 (red), 973 (blue), 1073 (green), and 1273 K (pink). (b) XRD patterns around 33.5° of the as-grown sample and the samples annealed up to 1673 K.

Fig. 2b shows the main (112) and (−112) peaks around 33.5° for different annealed samples. In the as-grown sample, the two Bragg peaks locate almost at the same angle, indicating a very small deviation in structure from the high symmetric Pnma space group. For the samples annealed between 773 K and 1273 K, such two peaks are clearly splitting (Fig. 2b), indicating the occurrence of octahedral distortion. The diffraction pattern of the sample annealed at 1273 K (Fig. 1b) could be indexed as a monoclinic structure with space group P2₁/n. As seen in Table 1, the volume of the unit cell and β angle are decreased due to annealing. The above differences in terms of lattice structure actually point to a difference in the degree of order in La₂NiMnO₆.^24,25 The volume and β angle are 233.65 ų and 90.01° for partially ordered La₂NiMnO₆, whereas two parameters decrease to 233.173 ų and 89.899° for fully ordered sample respectively. Therefore, it seems that the proper annealing condition can enhance the degree of order in our Y₂NiMnO₆ samples.

The promotion of order arrangement by annealing is further evidenced by the magnetic field (H) dependent magnetization (M) measurement (inset of Fig. 2a). With increasing the annealing temperature up to 1273 K, the saturation magnetic moment increases monotonically. The sample annealed at 1273 K exhibits the highest magnetic moment, indicating higher degree of order compared with those annealed at lower temperatures. It has been reported that in ordered double perovskite Ni²⁺ and Mn⁴⁺ ions are alternatingly located in corner-shared octahedral environments and a ↑↑↓↓ arrangement along c axis forms as elucidated in Fig. 1c. In order to analyze in detail the degree of order, oxide states and magnetic behaviours for the as-grown and annealed samples, XPS and magnetic measurements were performed as following.

Oxide states analysis

The XPS spectra of Mn 2p and Ni 2p core level can determine the oxide states of Mn and Ni ions in Y₂NiMnO₆.

Both the as-grown and the 1273 K annealed sample are investigated to check whether the oxide states of Mn and Ni ions have been changed or not after 1273 K annealing procedure, as shown in Fig. 3.

Fig. 3 XPS spectra of Mn 2p (a) and Ni 2p (b) for the as-grown (upper lines) and the 1273 K annealed sample (lower lines), respectively.

The XPS spectra of the two samples suggest the Mn 2p_3/2 peaks at about 642.6 eV. Since the 2p_3/2 core level appears at 641.8 eV and 642.4 eV for Mn³⁺ and Mn⁴⁺, respectively,²⁶ our data reveal unambiguously that the oxide states of Mn ion is +4 for two samples. The examined Ni 2p_3/2 peaks of the two samples are at 855.4 eV, in agreement with the value reported for Ni²⁺.²⁷ Furthermore, bond valence sum (BVS) calculated by software SPuDs is consistent with the XPS results (Table 1).²⁸

Joy et al. prepared LaMn_0.5Ni_0.5O₃ through low-temperature glycine-nitrate method.²⁹ However, charge disproportionation of Mn⁴⁺ + Ni²⁺ → Mn³⁺ + Ni³⁺ occurs when the as-grown sample was annealed at high temperature. XPS spectrum and BVS of our sample which was also synthesized at relative lower temperature clearly show that the oxide states still keeps Ni²⁺/Mn⁴⁺ after the sample was annealed at 1273 K. Therefore, hydrothermal method provides an effective route to synthesize double perovskite R₂NiMnO₆ with stable Ni²⁺/Mn⁴⁺ oxide states.

Magnetic analysis

The temperature dependence of zero field cooled (ZFC) and field-cooled (FC) under 100 Oe applied magnetic field were measured in temperature range of 4 to 300 K (Fig. 4).

Fig. 4 Temperature dependence of ZFC and FC susceptibility of (a) as grown Y₂NiMnO₆, and (b) Y₂NiMnO₆ annealed at 1273 K. Inset of (a) and (b) show χ⁻¹ versus temperature with Curie–Weiss fitting.

A ↑↑↓↓ antiferromagnetic arrangement has been theoretically predicted in Y₂NiMnO₆ and confirmed by neutron diffraction experiment in R₂CoMnO₆ (R = Y and Lu). FC curves of the as-grown and annealed samples show ferromagnetic-like behaviour. This implies ferromagnetic instability accessed by magnetic-field cooling, as proposed by S. Yáñez-Vilar et al.¹⁴ The Curie temperature was extracted from the ZFC curves, i.e., the inflection point where dM/dT is a minimum. For the both samples, the Curie temperatures are 84 K and a cusp appears at 76 K. The ZFC and FC magnetizations have the same values around and above T_c, below which the irreversibility is observed. This feature could be attributed to either the low measuring field or spin glass behaviour. The Curie–Weiss fitting (inset of Fig. 4a and b) yields C = 3.77 emu mol⁻¹ Oe⁻¹ K⁻¹, T_cw = 90 K for the as-grown sample and C = 3.82 emu mol⁻¹ Oe⁻¹ K⁻¹, T_cw = 100 K for the annealed sample. The positive Curie–Weiss temperature indicates long range canted-spin correlation.¹⁵ The extrapolated effective moments increases from 5.49 μ_B to 5.58 μ_B after the sample was annealed, which are essentially consistent with the theoretical magnitude 5.9 μ_B,^25,30 corresponding to one Ni²⁺ (S = 1) and one Mn⁴⁺ (S = 3/2) per formula.

Fig. 5 shows the M–H curves at different temperatures, in the range of −30 kOe to +30 kOe. The magnetizations of both samples tend to reach saturated at 4 K and a 30 kOe applied field. At temperatures higher than T_c, the M–H behaviours are Brillouin-function-like, corresponding to a paramagnetic state. The highest magnetic moment at 4 K and a 30 kOe magnetic field is 3.37 μ_B for the as-grown sample, indicating the presence of about 16% antisite disorder. It is noteworthy that the magnetic moment significantly increases to 5.11 μ_B after the sample was annealed at 1273 K, which is close to the summation of full ordered Ni²⁺ (S = 1) and Mn⁴⁺ (S = 3/2) magnetic moments in a formula unit.

Fig. 5 The M–H hysteresis of (a) Y₂NiMnO₆ and (b) Y₂NiMnO₆ annealed at 1273 K.

In order to shed more light on the magnetic behaviours of the two samples, we performed temperature dependence of susceptibility measurements in AC mode. Fig. 6 presents the frequency dependence of ac susceptibility, including the real component χ′(T) and imaginary components χ′′(T). Although it is not evident in the χ′(T), there exists a peak at 84 K corresponding to T_c in the χ′′(T), which is visible at higher frequency. For the as-grown sample, the susceptibilities show frequency dependent peak below the Curie temperature. In addition, a shoulder appears in the as-grown sample at about 68 K, which is absent in the annealed sample. As discussed above, the as-grown sample is partially ordered and short interaction exists in it.³⁰ Thus we believe an antiferromagnetic matrix coexists with some ferromagnetic clusters in the as-grown sample. However, for the annealed sample, the peak shows no shift with frequency (Fig. 6d), indicating the elimination of antisite disorder and a greater degree of long range antiferromagnetic order. Besides, small non-interacting ferromagnetic clusters, which may still exist in the annealed sample, give rise to the ferromagnetic-like susceptibility (Fig. 4b). Similar behaviour may be also resulted from the antiferromagnetically ordered spins that are slightly canted in low field.

Fig. 6 Real χ′ and imaginary χ′′ parts of AC susceptibility of (a) Y₂NiMnO₆, and (b) Y₂NiMnO₆ annealed at 1273 K. A zoomed view of AC susceptibility from 70 to 90 K of (c) Y₂NiMnO₆, and (d) Y₂NiMnO₆ annealed at 1273 K.

Dielectricity and ferroelectricity

The dielectricity and P–E measurement were performed on sample annealed at 1273 K. For ferroelectricity, the transition from ferroelectricity to paraelectricity is always accompanied by the changes in the dielectric constant ε with temperature. The dielectric constant ε as a function of T is shown in Fig. 7.

Fig. 7 Temperature dependence dielectric constant at 1 MHz.

The ε(T) exhibits an anomaly at 84 K, in agreement well with the Curie temperature. To check whether the sample is ferroelectricity, PUND measurement^5,31 was performed at 77 K. This method could subtract non-ferroelectric components, such as the accumulated charge at the grain boundaries and leakage current.⁵ The P–E loop is presented in Fig. 8, showing typical electric hysteresis and giving a saturated polarization of 35 μC m⁻², which is comparable with that of magnetic related multiferroics.

Fig. 8 P–E hysteresis loop measured at 77 K.

To understand the mechanism of the ferroelectricity driven by magnetical order, we propose a model similar to double perovskite Lu₂CoMnO₆ and Y₂CoMnO₆ with identical structure with Y₂NiMnO₆, as discussed subsequently.^14,15 As shown in Table 2, we compare the Ni–O–Mn bond angles of Y₂NiMnO₆ annealed at 1273 K with that of full ordered La₂NiMnO₆. The bond angles of Y₂NiMnO₆ are smaller than that of La₂NiMnO₆. With decreasing the bond angles, the superexchange interaction between Ni²⁺ and Mn⁴⁺ weakens. When the next-nearest neighbour Ni–O–Ni antiferromagnetic interaction is comparable with the nearest neighbour Ni–O–Mn ferromagnetic interaction, a frustrated ↑↑↓↓ magnetic arrangement is stabilized. Such ↑↑↓↓ arrangement of Ni²⁺ and Mn⁴⁺ breaks inversion symmetry and generates ferroelectricity, as theoretically predicted by Kumar et al.¹²

Table 2 Bond angles obtained from the room-temperature refinement^a

	Bond angle (deg)
a The parameters of La2NiMnO6 were taken from ref. 26.
Y2NiMnO6	Ni–O1–Mn	144.43(26)
	Ni–O2–Mn	148.31(32)
	Ni–O3–Mn	144.49(32)
La2NiMnO6	Ni–O1–Mn	161.7(3)
	Ni–O2–Mn	162.3(3)
	Ni–O3–Mn	158.5(2)

---

Conclusion

In summary, double perovskite Y₂NiMnO₆ was synthesized by mild hydrothermal method. Bond valence sum and XPS show the transition metals are Ni²⁺ and Mn⁴⁺. After the sample was annealed at 1273 K, the degree of order is greatly enhanced and approach to fully ordered state, as proved by the magnetic measurements. For the annealed sample, a dielectric constant anomaly at magnetic transition temperature was observed, and a spontaneous polarization about 35 μC m⁻² was determined by PUND method.

Acknowledgements

The authors are grateful to Yisheng Chai for the capacitance measurement.

References

1. J. F. Scott, Nat. Mater., 2007, 6, 256–257.
2. N. A. Spaldin and M. Fiebig, Science, 2005, 309, 391–392.
3. T. G. T. Kimura, H. Shintani, K. Ishizaka, T. Arima and Y. Tokura, nature, 2003, 426, 55–58.
4. S. Ishiwata, Y. Tokunaga, Y. Taguchi and Y. Tokura, J. Am. Chem. Soc., 2011, 133, 13818–13820.
5. Y. S. Chai, Y. S. Oh, L. J. Wang, N. Manivannan, S. M. Feng, Y. S. Yang, L. Q. Yan, C. Q. Jin and K. H. Kim, Phys. Rev. B: Condens. Matter Mater. Phys., 2012, 85, 184406.
6. J.-H. Lee, Y. K. Jeong, J. H. Park, M.-A. Oak, H. M. Jang, J. Y. Son and J. F. Scott, Phys. Rev. Lett., 2011, 107, 117201.
7. M. C. Weber, J. Kreisel, P. A. Thomas, M. Newton, K. Sardar and R. I. Walton, Phys. Rev. B: Condens. Matter Mater. Phys., 2012, 85, 054303.
8. S. S. Tokura Yoshinori, Adv. Mater., 2010, 22, 1554–1565.
9. M. Shang, C. Zhang, T. Zhang, L. Yuan, L. Ge, H. Yuan and S. Feng, Appl. Phys. Lett., 2013, 102, 062903.
10. Y. Tokunaga, N. Furukawa, H. Sakai, Y. Taguchi, T.-h. Arima and Y. Tokura, Nat. Mater., 2009, 8, 558–562.
11. Y. Tokunaga, S. Iguchi, T. Arima and Y. Tokura, Phys. Rev. Lett., 2008, 101, 097205.
12. S. Kumar, G. Giovannetti, J. van den Brink and S. Picozzi, Phys. Rev. B: Condens. Matter Mater. Phys., 2010, 82, 134429.
13. Y. J. Choi, H. T. Yi, S. Lee, Q. Huang, V. Kiryukhin and S. W. Cheong, Phys. Rev. Lett., 2008, 100, 047601.
14. S. Yáñez-Vilar, E. D. Mun, V. S. Zapf, B. G. Ueland, J. S. Gardner, J. D. Thompson, J. Singleton, M. Sánchez-Andújar, J. Mira, N. Biskup, M. A. Señarís-Rodríguez and C. D. Batista, Phys. Rev. B: Condens. Matter Mater. Phys., 2011, 84, 134427.
15. G. Sharma, J. Saha, S. D. Kaushik, V. Siruguri and S. Patnaik, Appl. Phys. Lett., 2013, 103, 012903.
16. R. J. Booth, R. Fillman, H. Whitaker, A. Nag, R. M. Tiwari, K. V. Ramanujachary, J. Gopalakrishnan and S. E. Lofland, Mater. Res. Bull., 2009, 44, 1559–1564.
17. R. P. Maiti, S. Dutta, M. Mukherjee, M. K. Mitra and D. Chakravorty, J. Appl. Phys., 2012, 112, 044311.
18. S. Feng and R. Xu, Acc. Chem. Res., 2001, 34, 239–247.
19. Y. Chen, H. Yuan, G. Li, G. Tian and S. Feng, J. Cryst. Growth, 2007, 305, 242–248.
20. J. J. Urban, L. Ouyang, M.-H. Jo, D. S. Wang and H. Park, Nano Lett., 2004, 4, 1547–1550.
21. G. Zhang, G. Li, F. Liao, Y. Fu, M. Xiong and J. Lin, J. Cryst. Growth, 2011, 327, 262–266.
22. A. Larsen and R. Von Dreele, Report LAUR 86-748, Los Alamos National Laboratory, New Mexico, USA, 1994.
23. B. H. Toby, J. Appl. Crystallogr., 2001, 34, 210–213.
24. D. Choudhury, P. Mandal, R. Mathieu, A. Hazarika, S. Rajan, A. Sundaresan, U. V. Waghmare, R. Knut, O. Karis, P. Nordblad and D. D. Sarma, Phys. Rev. Lett., 2012, 108, 127201.
25. N. S. Rogado, J. Li, A. W. Sleight and M. A. Subramanian, Adv. Mater., 2005, 17, 2225–2227.
26. J. C. Carver, G. K. Schweitzer and T. A. Carlson, J. Chem. Phys., 1972, 57, 973–982.
27. K. K. Lian, D. W. Kirk and S. J. Thorpe, J. Electrochem. Soc., 1995, 142, 3704–3712.
28. M. W. Lufaso and P. M. Woodward, Acta Crystallogr., Sect. B: Struct. Sci., 2001, 57, 725–738.
29. V. L. J. Joly, P. A. Joy, S. K. Date and C. S. Gopinath, Phys. Rev. B: Condens. Matter Mater. Phys., 2002, 65, 184416.
30. Y. Mao, J. Parsons and J. S. McCloy, Nanoscale, 2013, 5, 4720–4728.
31. M. Fukunaga and Y. Noda, J. Phys. Soc. Jpn., 2008, 77, 064706.

---

Footnote

† Electronic supplementary information (ESI) available: The structural parameters of the as-grown and annealed samples. See DOI: 10.1039/c4ra07099b

---