Low temperature magnetism in the rare-earth perovskite GdScO3
Sheng Jie-Ming1, 2, 3, Kan Xu-Cai4, Ge Han1, Yuan Pei-Qian1, Zhang Lei1, Zhao Nan1, Song Zong-Mei1, Yao Yuan-Yin1, Tang Ji-Ning1, Wang Shan-Min1, Tian Ming-Liang4, 5, Tong Xin2, 3, ‡, Wu Liu-Suo1, §
Department of Physics, Southern University of Science and Technology, Shenzhen 518055, China
Institute of High Energy Physics, Chinese Academy of Sciences (CAS), Beijing 100049, China
Spallation Neutron Source Science Center, Dongguan 523803, China
School of Physics and Materials Science, Anhui University, Hefei 230601, China
High Magnetic Field Laboratory, Chinese Academy of Science (CAS), Hefei 230031, China

 

† Corresponding author. E-mail: tongx@ihep.ac.cn wuls@sustech.edu.cn

The work at SUSTech was supported by the National Natural Science Foundation of China (Grant No. 11974157). Part of this work was also supported by the National Natural Science Foundation of China (Grant No. 11875265), the Scientific Instrument Developing Project of the Chinese Academy of Sciences (3He-based neutron polarization devices), and the Institute of High Energy Physics, the Chinese Academy of Sciences. Kan X C and Tian M L were supported by the National Natural Science Foundation of China (Grant No. 51802002).

Abstract

The magnetic phase diagram of rare-earth perovskite compound, GdScO3, has been investigated by magnetization and heat capacity. The system undergoes an antiferromagnetic phase transition at TN = 2.6 K, with an easy axis of magnetization along the a axis. The magnetization measurements show that it exists a spin-flop transition around 0.3 T for the applied field along the a axis. The critical magnetic field for the antiferromagnetic-to-paramagnetic transition is near 3.2 T when temperature approaches zero. By scaling susceptibilities, we presume this point (B = 3.2 T, T = 0 K) might be a field-induced quantum critical point and the magnetic critical fluctuations can even be felt above TN.

1. Introduction

The family of rare-earth perovskite scandates RScO3 (R = rare earth) have attracted continued attention due to great variety of physical properties and potential applications. For example, the Ising-like antiferromagnet,[1] the giant magnetocaloric effect,[2,3] and the promising candidate materials for the replacement of SiO2 in silicon MOSFETs because of the large optical band gaps, high dielectric constants, andq chemical stabilities.[49] In addition, the RScO3 compounds with different lattice parameters have also been exploited to serve as substrates for engineering of highly strained ferroelectric and multiferroic thin films. The biaxial strain due to lattice mismatch between substrates and thin films can dramatically alter the properties of epitaxial films.[10] For instance, the SrTiO3 thin film exhibits a strain-induced ferroelectricity at room temperature when it is grown on DyScO3 substrate, compared with the paraelectric behavior of bulk unstrained SrTiO3.[11] The compressively strained BaTiO3 thin films grown on DyScO3 and GdScO3 substrates can also produce larger ferroelectric (FE) polarization and higher FE transition temperature.[12]

Apart from the effect of strain, the intrinsic properties of RScO3 substrates may also play important effect on the properties of thin films. For example, the magnetic properties of RScO3 should be took into account when they are used as substrates for magnetic oxide thin films.[13] However, due to the low magnetic ordering temperatures of RScO3, normally only a few Kelvin degrees,[1315] the magnetic effects from the substrates to the magnetic thin films are usually neglected and have not been the subject of detailed investigation. However, some of this series compounds still exist strong magnetic correlation well above the magnetic transition temperature, which cannot be ignored even at dozens of Kelvin degrees, such as the short range magnetic order discovered in DyScO3 at 30 K.[1] Actually, it was argued that these strong fluctuations are enhanced by the one dimensionality of the rare earth magnetic spin chains. Indeed, the quantum critical Tomonaga–Luttinger liquid behavior has been reported in YbAlO3,[16] and the strong orbital couplings and crystal field effects was claimed to be crucial for the unexpected low dimensionality behaviors.[1,16,17]

On the other side, for GdScO3, with the electron configuration of Gd3+ is 4f7 and L = 0, no spin–orbit coupling is expected. Thus following the arguments in previous studies,[1,16,17] the low-temperature magnetism of GdScO3 should behave very different, comparing with DyScO3 and YbAlO3. Despite that GdScO3 has been reported as a giant magnetocaloric system with an antiferromagnetic (AFM) transition, TN = 2.6 K,[3] the details of the magnetic phase diagram and Gd–Gd interactions still remained poorly understood. Moreover, the studying of the Gd–Gd interactions in GdScO3 may also be very helpful to understand the complex magnetism in GdFeO3 and GdMnO3.[18,19]

2. Experimental details

High-quality single crystals of GdScO3 used here are commercially available.[12,20] Magnetization measurements were performed using two magnetometers for different temperature ranges: (i) commercial vibration sample magnetometer Quantum Design Magnetic Property Measurement System (MPMS-VSM) for the high-temperature measurements T = 1.8 K–300 K; (ii) MPMS-3 with 3He insert for the temperatures T = 0.5 K–1.5 K. Heat capacity measurements were performed on the Physical Property Measurement System (PPMS) in the temperature range 1.8 K–300 K.

3. Results and analysis

As shown in Fig. 1(a), GdScO3 crystallizes in a distorted orthorhombic perovskite structure with a = 5.4862 Å, b = 5.7499 Å, and c = 7.9345 Å in the Pbnm (No. 62) space group at room temperature.[21] Gd3+ ions are located between 8 distorted ScO6 octahedrons. Due to the distortion, the point symmetry of the Gd site is decreased from Oh in the ideal cubic structure to Cs, which is only one mirror plane normal to the c axis, in the orthorhombic structure. Figure 1(b) illustrates the T dependence of specific heat of GdScO3 single crystal. A sharp anomaly appears near 2.6 K indicating a magnetic transition similar to GdAlO3.[22] The temperature evolution of the magnetic susceptibility χ(T) with an applied magnetic field of 0.3 T along the three principle crystal axis is shown in Fig. 1(c). The anomalies observed around 2.6 K confirm the AFM phase transition. The AFM order is mainly from a competition between exchange interactions and dipolar magnetic interactions for the neighbouring Gd3+ ions.[23] The Néel temperature TN = 2.6 K of GdScO3 is lower than the isostructural GdAlO3 (TN = 3.98 K).[22] This may be due to Sc3+ ion has larger ionic radius than Al3+ ion, which leads to larger distance between the neighbouring Gd3+ ions and produces smaller magnetic interactions. Above TN, the susceptibilities χ(T) along the three principle crystal axis almost overlap together except the small T region near TN, which indicates GdScO3 only exists very weak anisotropy. This is consistent with the S = 7/2, L = 0 electron configuration of Gd3+ ion. Below TN, it is seen that the susceptibility for the field parallel to axes b and c varies very little with temperature, whereas that when the field is parallel to the a axis decreases steeply (see the inset of Fig. 1(c)). Unlike the magnetic Ax structure preferred by the direct dipole–dipole interaction only, the magnetization with easy axis along a is consistent with the AF magnetic structure (Gx) as observed in GdAlO3,[22] where the exchange interaction between the large Gd3+ S = 7/2 spins in ab plane is dominant. This is very different from the cases of DyScO3 and YbAlO3.[1,16,17] Figure 1(d) displays the inverse susceptibility plotted against the temperature for magnetic field applied along the c axis. Fits of 1/χc for the paramagnetic (PM) region (15 K < T < 300 K) to a Curie–Weiss law yield Curie temperature θCW of –4.85 K and effective magnetic moment μeff of 7.45 μB/Gd. The negative value in θCW confirms AFM interactions between the neighboring Gd3+ ions. The large μeff = 7.45 μB/Gd compared with the theoretical value of 7.94 μB assuming a spin of 7/2 for the Gd3+ ions.

Fig. 1. (a) Crystal structure of GdScO3 in space group Pbnm, where a Gd3+ ion is surrounded by eight distorted ScO6 octahedrons. (b) T dependence of specific heat Cp(T). (c) T dependence of magnetic susceptibility χ(T) in an applied magnetic field of 0.3 T parallel to the a, b, and c axes, respectively. The inset: χ(T) near the AFM transition. (d) T dependence of inverse magnetic susceptibility 1/χ(T) with 0.3-T magnetic field along the c axis. The blue solid line is the fit using the Curie–Weiss law above the transition temperature.

Figure 2(a) illustrates the isothermal magnetizations of GdScO3 as a function of applied field along the a axis ranging from 1.0 K to 5.0 K. The magnetizations closely approach to saturation around M = 6.5 μB/Gd above 5 T, which is consistent with the reported saturated moment of Gd3+ ion.[22] An enlarged view of the low-field region of M(B) curves is shown in Fig. 2(b). All of the magnetization curves for T < TN show an abrupt change around the field between 0.2 T and 0.45 T. This suggests the occurrence of a spin flop (SF) type transition, similar as the transition reported in GdAlO3,[22] where the spin direction of two sublattice Gd3+ ions suddenly turn perpendicular to the easy axis direction. This is due to the lower Zeeman energy of Gd3+ spins for the applied field along the easy axis until reaching the critical field of SF transition Bc1. In addition, the observed hysteresis loop shown in the inset of Fig. 2(a) indicates the SF transition is a subsequent field-induced first-order transition. With increasing of temperature, Bc1 also increases. This behavior is slightly different with GdAlO3, in which the critical field is almost independent of T.[22] Above Bc1, the magnetizations increase linearly with the field, and break off at a point which gives an AFM-to-PM critical field Bc2. The T-dependent magnetization of GdScO3 with field applied along the a axis is shown in Fig. 2(c). Two anomalies are observed at 1.97 K and 2.55 K for B = 0.35 T, which correspond SF transition and AFM-to-PM phase transition, respectively. For the former, as the applied magnetic field increases, the phase transition T also increases, while the latter has the opposite trend (see the dash lines in Fig. 2(c)). Figure 2(d) shows the contour plot of magnetic phase diagram for applied field along the a axis. The intensity corresponds to the susceptibility χ = d M/d B. Below TN, the critical field of SF transition divides the phase diagram into two regions. At the low-field region, GdScO3 is an antiferrromagnet, in which the spin directions of Gd3+ ions are dominantly aligned in the a axis, and each Gd3+ ion in one sublattice might be surrounded by six nearest neighbours Gd3+ ions with opposite moments direction belonging to the other sublattice like GdAlO3[23] (the magnetic structure of GdScO3 should be confirmed by neutron diffraction). Above Bc1, the moments turn perpendicular to the a axis due to the lower ground energy. A further increase of field rotates the spins continuously, until an AFM–PM critical field Bc2 is reached where all the moments are polarized along the applied field.

Fig. 2. (a) Field dependence of magnetization M of GdScO3, measured at different T, with field along the a axis. (b) The magnetization discontinuity at several T corresponds the SF phase transition. (c) T dependence of magnetization M of GdScO3 measured at several fields along the a axis. (d) The contour plot of the magnetic phase diagram of GdScO3 with applied field along the a axis.

The measurement of isothermal magnetizations of GdScO3 with applied field along the b and c axes from 1.8 K to 3.0 K is shown in Figs. 3(a) and 3(c), respectively. Both the saturation moments and critical fields along the two axes are very similar with the a axis, which is consistent with the S-state of Gd3+ ions. In addition, since the spins direction of Gd3+ ions are dominantly along the a axis, there is no SF transition occurred when the field along the two axes. Figures 3(b) and 3(d) illustrate the magnetic phase diagram for field parallel to the b and c axes, respectively. With the increasing field, the AFM arrangement of spins of two-sublattice Gd3+ ions are gradually polarised, forming a PM magnetic state.

Fig. 3. (a) and (c) Field dependence of magnetization M of GdScO3, measured at different T, with field along the b and c axes. (b) and (d) The contour plot of the magnetic phase diagram of GdScO3 with applied field along the b and c axes. The AFM to PM phase boundary is given by the red open circles.

It usually believed that the AFM phase will directly enter into trivial PM phase in GdScO3 with the applied field increased. However, since the AFM–PM transition is of second order and the transition temperature is gradually suppressed by the applied field, it might be existed a quantum critical field Bc2 at 0 K, where the AFM state transforms into the PM state. Around the QCP, the magnetic fluctuations are quantum mechanical in nature, which can exhibit scale invariance in both space and time. Unlike classical critical points, where the critical fluctuations are limited to a narrow region around the phase transition, the influence of the QCP can be felt over a wide range of temperatures above the QCP, so the effect of quantum criticality is felt without ever reaching 0 K. Figure 4(a) illustrates the field-dependence of the magnetic susceptibilities χ = dM/dB measured below and above TN. Around the QCP, susceptibilities with different T can be scaled as

where d is the spatial dimension, ν is the scaling exponent related to the magnetic field, z is the dynamical exponent, and (Bc2≃ 3.2 T at 0 K ). The relationship between d, ν, and z can be obtained by scaling the susceptibility curves as shown in Fig. 4(a) and minimizing the deviation:

The equations give

yielding the interesting result that the spatial dimension d is approximately equal to the dynamical exponent z in GdScO3. The susceptibility curves scales well for temperatures from 1.0 K to 4.2 K and the scaling behaviors observed are consistent with a two dimensional QCP, where d = 2, z ≈ 2, and ν ≈ 1/2. This is not very surprising.

Fig. 4. (a) Field dependencies of the magnetic susceptibility χ = d M/d B from 1.0 K to 4.2 K with field along a axis. (b) Critical scaling of the magnetic susceptibility in panel(a).

As previous discussed, in GdScO3, the orbital contribution (L = 0) is absent, and the magnetic ground states of Gd3+ are isotropic eight-degenerated S = 7/2 states. Due to the large spin values of S = 7/2, the in plane exchange interactions of the form JSS are greatly enhanced, and become dominant in GdScO3 and GdAlO3.[23]

On the other hand, although sharing the same crystal structure, the situations in YbAlO3 and DyScO3 are completely different. Due to the large orbital contributions, the degenerated spin states of YbAlO3 and DyScO3 are lifted by the crystal field effect, and the ground states are well separated Kramer doublets. These doublet ground states could be treated as effective spin S = 1/2, which turns out to be crucial for observed unconventional one dimensional physics in YbAlO3 and DyScO3 and related systems.[16,17]

4. Conclusion

In summary, we investigated the magnetism and heat capacity of single crystal GdScO3 at low temperatures. GdScO3 is an uniaxial antiferromagnet with the easy axis of magnetization along the a axis. The Néel temperature is found to be 2.6 K. The magnetization measurements show it exists an SF phase transition when the applied field along the a axis. The field-temperature phase diagrams for B along the three principle crystal axes are established. By scaling the magnetic susceptibilities, we demonstrate there might be a quantum critical point (QCP) at B = 3.2 T and T = 0 K corresponding to the AFM-to-PM phase transition. Compared to the one-dimensional physics observed in YbAlO3 and DyScO3, the magnetic structure and the critical scaling indicate that the in-plane exchange interactions are more important in GdScO3. It is found that quantum critical like fluctuations affect the physical properties over a large temperature and field area in GdScO3, and additional caution may need be paid at low temperatures, when it was treated as a simple PM substrates for the epitaxial growth of magnetic thin films.

Reference
[1] Wu L S Nikitin S E Frontzek M Kolesnikov A I Ehlers G Lumsden M D Shaykhutdinov K A Guo E J Savici A T Gai Z 2017 Phys. Rev. 96 144407
[2] Wu Y D Qin Y L Ma X H Li R W Wei Y Y Zi Z F 2019 J. Alloys Compd. 777 673
[3] Jia J H Ke Y J Zhang X X Wang J F Su L Wu Y D Xia Z C 2019 J. Alloys Compd. 803 992
[4] Schlom D G Haeni J H 2002 MRS Bull. 27 198
[5] Lim S G Kriventsov S Jackson T N Haeni J H Schlom D G Balbashov A M Uecker R Reiche P Freeouf J L Lucovsky G 2002 J. Appl. Phys. 91 4500
[6] Lucovsky G Zhang Y Whitten J L Schlom D G Freeouf J L 2004 Microelectronic Engineering 72 288
[7] Lucovsky G Hong J G Fulton C C Zou Y Nemanich R J Ade H Scholm D G Freeouf J L 2004 Physica Status Solidi 241 2221
[8] Delugas P Fiorentini V Filippetti A Pourtois G 2007 Phys. Rev. 75 115126
[9] Christen H M Jellison G E Jr Ohkubo I Huang S Reeves M E Cicerrella E Freeouf J L Jia Y Schlom D G 2006 Appl. Phys. Lett. 88 262906
[10] Schlom D G Chen L Q Eom C B Rabe K M Streiffer S K Triscone J M 2007 Ann. Rev. Mater. Res. 37 589
[11] Haeni J H Irvin P Chang W Uecker R Reiche P Li Y L Choudhury S Tian W Hawley M E Craigo B et al. 2004 Nature 430 758
[12] Choi K J Biegalski M Li Y L Sharan A Schubert J Uecker R Reiche P Chen Y B Pan X Q Gopalan V et al. 2004 Science 306 1005
[13] Ke X Adamo C Schlom D G Bernhagen M Uecker R Schiffer P 2009 Appl. Phys. Lett. 94 152503
[14] Plaza I Palacios E Bartolomé J Rosenkranz S Ritter C Furrer A 1997 Physica B 234 635
[15] Raekers M Kuepper K Bartkowski S Prinz M Postnikov A V Potzger K Zhou S Arulraj A Stüßer N Uecker R et al. 2009 Phys. Rev. 79 125114
[16] Wu L S Nikitin S E Wang Z Zhu W Batista C D Tsvelik A M Samarakoon A M Tennant D A Brando M Vasylechko L et al. 2019 Nat. Commun. 10 698
[17] Wu L S Nikitin S E Brando M Vasylechko L Ehlers G Frontzek M Savici A T Sala G Christianson A D Lumsden M D et al. 2019 Phys. Rev. 99 195117
[18] Tokura Y Seki S Nagaosa N 2014 Rep. Prog. Phys. 77 076501
[19] Kimura T Lawes G Goto T Tokura Y Ramirez A P 2005 Phys. Rev. 71 224425
[20] Schubert J Trithaveesak O Petraru A Jia C L Uecker R Reiche P Schlom D G 2003 Appl. Phys. Lett. 82 3460
[21] Liferovich R P Mitchell R H 2004 J. Solid State Chem. 177 2188
[22] Blazey K W Rohrer H 1968 Phys. Rev. 173 574
[23] Quezel S Mignod J R Tcheou F 1982 Solid State Commun. 42 103