Anomalous magnetic properties of GdCrTiO₅ nanoparticles

The interest in materials with sizable magnetocaloric effect (MCE) has grown significantly because of the possibilities of their future uses as solid refrigerants in magnetic cooling systems.¹ Magnetic cooling has been recognized as a promising alternative to conventional vapour-cycle refrigeration because of its low power consumption, improved efficiency, and environmental friendliness.^1–4 Magnetic refrigeration is an entropically-driven thermodynamic phenomenon, which couples magnetic entropy with lattice thermal energy and permits cooling from room temperature to the sub-Kelvin temperature region.⁵ 

Multiferroic materials consist of multiple ferroic states,^6,7 and it is envisioned that the co-dependence of the order parameters will contribute towards spintronics and innovative memory devices.^8–10 In the RCrTiO₅ series, one can expect to find exceptional and fascinating magnetic behavior because of the co-existence of two magnetic sublattices, Cr³⁺ and R³⁺.^11–14 There are limited reports on this group of materials, mostly on basic magnetic properties in polycrystalline bulk materials.^6–13 Few reports systematically studied magnetic properties of bulk GdCrTiO₅.^9–11 Recently, a study demonstrated magnetic transitions such as spin reorientation (SR) and compensation in bulk DyCrTiO₅, as well as in HoCrTiO₅, and also observed exchange bias (EB) effects.¹³ Due to this exchange coupling, a shift in the hysteresis loop across the applied field axis originates in ferromagnetic-antiferromagnetic (FM-AFM) systems.¹⁵ The EB effect due to exchange coupling among different atoms of an oxide compound having a rare-earth constituent depicts interesting magnetic behavior.^13–15 Nowadays, the application of nanostructured materials is enhanced due to their improved properties, which is a consequence of the increase in the surface-to-volume ratio and uncompensated surface spins present on the grain surfaces.^16,17 The uncompensated surface spins present at the surface, significantly modify magnetic properties and other physical properties.^16,17 The change in the structural properties subsequently brings about concomitant changes in the magnetic nature of these materials.^17–22 Zheng et al.,²³ observed the EB effect in YMnO₃, which is ascribed to the uncompensated surface spins of the nanoparticles. Previously, the MCE has been studied in bulk GdCrTiO₅ compound.⁶ In this contribution, GdCrTiO₅ nanoparticles were synthesized, and their magnetic properties were studied.

The GdCrTiO₅ specimen was synthesized by a cost-effective sol-gel technique.^14,16,17 Stochiometric amounts of Gd(NO₃)₃, Cr(NO₃)₃ and Ti(OC₃H₇)₄ were added to absolute ethanol. The prepared solution was stirred continuously by adding a few drops of HNO₃. After that, 50 mg of cetyltrimethylammoniumbromide (CTAB) was added to stirring solution, subsequently 10 ml of distilled water was poured into the solution while stirring was continued for another 30 minutes. The prepared reaction mixture was allowed to age for a day. The aged solution was dried, and then the obtained powder was consequently crushed and calcined at 800 °C for three hours in a box furnace. The structural and morphological features of the calcined powders were explored by x-ray diffraction (XRD) technique using Cu-Kα radiation (λ = 1.5406 Å), and transmission electron microscope (TEM). A vibrating sample magnetometer (VSM) using a Cryogenic, Cryogen Free Physical Properties Measurement System, with a vibrating sample magnetometer insert was used in order to study the magnetic behaviour as a function of temperature and applied magnetic field.

XRD pattern of GdCrTiO₅ sample along with the profile fitting using FULLPROF²⁴ are plotted in Fig. 1(a). The simulated XRD pattern is obtained by considering the orthorhombic crystal structure with a Pbam space group. The cell parameters a, b and c are found to be 7.381 ± 0.002, 8.679 ± 0.002 and 5.863 ± 0.002 Å, respectively, while the cell volume found from the refinement parameters is 375.6 ± 0.2 ų. An earlier report on a polycrystalline bulk GdCrTiO₅ sample, which has been prepared by the solid-state reaction method, shows an orthorhombic phase with the cell parameters a, b and c given as 7.4385, 8.5814, and 5.7782 Å, respectively.⁶ In a comparison of the lattice parameters of the synthesized powder with that of the bulk sample,⁶ it is observed that the lattice size is increased, which might be the consequence of the decreased size of the particles.²⁵ It is observed that as the size of the crystallites is reduced, the lattice parameter increases.²⁶ Minor peak shifting has appeared in a few peak positions as a result of internal stresses in nano-dimensional material.²⁷ Broadening of the XRD peaks is observed when the size of the particle is reduced.²⁸ In the present study peaks are broadened compared to the bulk,⁶ because of the reduction in size of the particle to nanoscale as confirmed by the TEM. Certain XRD peaks are merged when samples are in the nano form.¹⁴ As a result of peak shifting and merging of XRD reflections, a difference is observed between the experimental and calculated data is obtained in the present case.^27,28 Furthermore, the crystallite size of the sample was calculated using Scherrer’s formula¹⁷ and found to be 33.0 ± 0.2 nm.

In Fig. 1(b) the TEM micrograph for GdCrTiO₅ is shown. The morphology indicates that the particles are nearly spherical in shape. The particle size histogram with a log-normal fit²⁹ to the histogram is shown in Fig. 1(c). The average particle size for the synthesized sample is found to be 38.0 ± 0.4 nm, with particle size ranging from 15 to 90 nm. The particle size and crystallite size are nearly equal, which indicates single crystalline nature of the particles. In this contribution, the nanoparticles of GdCrTiO₅ were obtained after three hours of calcination at a temperature of 800 °C, while the bulk GdCrTiO₅ sample was prepared at a sintering temperature of 1400 °C over several days.⁶ The selected area electron diffraction (SAED) pattern is depicted in Fig. 1(d). The dotted pattern indicates the crystalline sample with single crystal nature.¹⁷ The elements present in the sample are confirmed using energy-dispersive x-ray spectroscopy (EDS). The EDS spectrum is shown in Fig. 1(e), indicating that the sample consists of the constituent elements Gd, Cr, Ti and O. The presence of C and Cu in the spectra (Fig. 1(e)) is from the carbon-coated Cu grid used for the microscopy.

To understand the magnetic properties of the synthesized GdCrTiO₅ sample, magnetic moment as a function of temperature, M(T), was measured in zero-field-cooling (ZFCW), field-cool-cooling (FCC) and field-cool-warming (FCW) cycles at the applied magnetic field, 0.01, 0.05 and 0.1 T.^30–38 The M(T) curves as a function of the indicated applied magnetic fields for the GdCrTiO₅ sample are shown in Fig. 2(a)-(c). The M(T) curves diverge on reducing the temperature from 150 to 140 K. This type of behaviour is attributed to the competing magnetic interactions in the material.^14,39 The temperature range 0 to 200 K is shown in Fig. 2(a)-(c), as the sample remains paramagnetic beyond 200 K. The nature of the M(T) curves shows the irreversible nature in all three measurement cycles at low field. The complicated magnetic properties in this system are expected because of the competing Cr³⁺-Cr³⁺, Gd³⁺-Gd³⁺ and Gd³⁺-Cr³⁺ exchange interactions.²⁹ From Fig. 2(a), a magnetic transition associated with the Néel temperature, T_N = 190 ± 2 K, is observed. In the temperature range T < T_N, the magnetization curves show very complex and interesting features, as the magnetization drastically changes in this region. The antiferromagnetic (AFM) moment is slightly canted from their arrangement and a ferromagnetic (FM) component is emerges because of the anti-symmetric exchange interaction of the Cr³⁺-Cr³⁺ ions, called as Dzyaloshinskii-Moriya (DM) interaction.^13–15 In addition, at low temperatures, an anisotropic exchange develops between the Gd³⁺ and Cr³⁺ spins, which generates an effective field at the Gd³⁺ site, showing a moment opposite to Cr³⁺.^13,15 With an increase in the temperature from 2 K, the magnetization decreases rapidly due to the faster decrease of Gd³⁺ moment.¹³ In FCC and FCW cycles, all the local moments of Gd³⁺ and Cr³⁺, cancel each other at 9 K. As a result, the net magnetic moment becomes zero, termed as the compensation temperature, referred to as T_comp1.²⁹ Above T_comp1, the resultant magnetization vector orients in the reverse direction to the external applied magnetic field with increasing temperature. As a result, negative magnetization is observed, and this extends up to a minimum value at T_SR of about 20 K. At this temperature, a new magnetic phase transition occurs because of spin reorientation of Cr³⁺ ion.¹³ Increasing the temperature further, the net magnetization enhances and again compensates at a temperature of 115 K, referred to as T_comp2.²⁹ In between T_comp1 and T_comp2 the sample shows negative magnetization.

For T > T_comp2 the sample shows positive magnetization. Another important feature is that higher M(ZFC) values are observed compared to the M(FC) values, in this case at 0.01 and 0.05 T applied field. A previous report stated that various interactions like FM, AFM, and canted spin structure in the material result in higher magnetization in the ZFC cycle than FC.⁴⁰ The magnetization as a function of temperature was previously studied in bulk GdCrTiO₅ at 0.005 T, showing a decrease in magnetization with increasing temperature, but T_SR and T_comp has not observed.⁶ Therefore, for the first time, the spin reorientation and negative magnetization features are reported here for the nanoparticles of GdCrTiO₅. Furthermore, a loop is observed in FCC and FCW cycles below the temperature 60 K (Fig. 2(a)-(c)). This nature of FCC and FCW cycles was previously observed in some manganites and metallic alloy systems.^30–38 However, this behavior is not observed in the GdCrTiO₅ system.^6–13 The obtained feature indicates an FM-AFM phase transition at low temperature.^36,37 The observed loop has been suppressed with the increase in probing magnetic field, which is completely suppressed at the higher fields.^30,32 This phenomenon is termed kinetic arrest.^30–38 In the present case, a similar suppression of the loop between FCC and FCW magnetization data has been observed (Fig. 2(c)) when the probing field becomes 0.1 T.^30,32

A further effect of the magnetic field on the magnetization of the material is explored. Fig. 3(a) shows the five-segment M$μ0H$ loop of GdCrTiO₅, recorded at 2, 10 and 50 K. Considering the low field region shown in Fig. 3(b), all three measurement cycles show small coercivity, which indicates the FM nature of the material due to DM interaction in Cr³⁺ moment.¹⁵ The obtained loops are asymmetric in nature, referred to as the EB effect.^13–15 The EB field is evaluated by using the relation, H_ex = H₁+H₂/2, where H₁ and H₂ are the positive and negative intercepts of the M$μ0H$ curve on the field axis.¹⁴ The obtained value of H_ex for the nanoparticles GdCrTiO₅ is 0.008 T. To examine the magnetocaloric nature of the material; isothermal magnetic measurement is carried out from 0 to 7 T in the temperature range 2 to 50 K, shown in Fig. 3(c). At low temperatures, the M$μ0H$ loops are nonlinear but do not attain saturation, whereas, at high temperatures, these curves become linear. This change in the behaviour of the M$μ0H$ loops is observed because of the broken exchange interactions within Gd³⁺ ions.⁴ 

In order to better understand the critical behaviour of spins in a magnetic phase transition, Arrott plots are plotted at different temperatures, shown in Fig. 3(d). According to the literature,^4,39 Arrott plots are broadly used to establish Curie or Néel temperature in magnetic materials and order of the magnetic transition. If the transition is first-order, then the curve’s slope is negative, and for the second-order transition, a positive slope is expected.^4,39,41 In the present case, all Arrott plots exhibit a positive slope, indicating the second-order nature of the material.⁴ The magnetocaloric effect (MCE) of the material is examined by utilizing the isothermal magnetization data. In this report, the magnetization is measured at different intervals of temperature and field. Therefore, the change in isothermal magnetic entropy ΔS_m is numerically calculated with the following equation,⁶ 

where, M_i+1 and M_i is the magnetization at temperature T_i+1 and T_i for the change in the magnetic field ΔH_i. The temperature-dependent ΔS_m is calculated using the expression, and data is plotted in Fig. 3(e). A large negative value of ΔS_m is obtained for this system similar to that observed in other Gd-based systems.^2–5 The maximum value of the change in isothermal magnetic entropy (−ΔS_m) of 22 ± 3 J.kg^-1.K^-1 is found below 10 K, for a 7 T change in the field. The MCE in these GdCrTiO₅ nanoparticles is associated with degenerated frustrated magnetic states due to the competition between DM interaction and spatial anisotropy.⁶ When a magnetic field is applied, the degeneracy is removed and frustration in magnetic moment appears in the material; consequently a change in magnetic entropy occurs.

This is the first report on GdCrTiO₅ nanoparticles, synthesized using a simple sol-gel technique to explore the role of particle size on structural and magnetic properties. The sample was stabilized with an orthorhombic crystal structure (space group Pbam). The single crystalline nature of the particles with a size of 38.0 ± 0.4 nm is evidenced from the TEM analysis. From the temperature-dependent magnetization measurements, different magnetic transitions, including T_SR and T_N, are observed. Magnetization as a function of field measurement showed the good MCE (−ΔS_m = 22 ± 3 J.kg^-1.K^-1) in the synthesized sample. The observed magnetic properties of the material are attributed to the coupling between Cr³⁺ and Gd³⁺ ions.

Financial aid from the SANRF (Grant No. 120856), as well as the University of Johannesburg (UJ) URC and FRC (Grant Nos. 282810, 083135), as well as Spectrum is acknowledged. The use of the NEP Cryogenic Physical Properties Measurement System at UJ, obtained with the financial support from the SANRF (Grant No. 88080) and UJ, RSA, is acknowledged. The use of the Spectrum facility in the FoS at UJ is acknowledged.

The authors have no conflicts to disclose.

The data that support the findings of this study are available from the corresponding author upon reasonable request.