Hydrostatic pressure effect on magnetic phase transition and magnetocaloric effect of metamagnetic TmZn compound

The magnetocaloric effect (MCE) is an intrinsic thermal response of all magnetic solids which has a direct and strong correlation with the corresponding magnetic phase transition. It has been well recognized that the magnetic phase transition can be tuned by adjusting applied pressure. Therefore, we perform the high hydrostatic pressure magnetization measurements (up to 1.4 GPa) on a recently reported giant MCE material of TmZn. The results indicate that the Curie temperature of TC increases from 8.4 K at the ambient pressure to 11.5 K under the pressure of 1.4 GPa. The field-induced first order metamagnetic transition is getting weak with increasing pressure, which results in a reduction of MCE. The hydrostatic pressure effect on the magnetic phase transition and MCE in the metamagnetic TmZn is discussed.

In recent years, the magnetocaloric effect (MCE) in magnetic materials has been well investigated, not only due to their potential applications for active magnetic refrigeration but also enable to understand the related fundamental properties of these materials [1][2][3][4][5][6][7][8] . MCE is an intrinsic thermal response of all magnetic solids which manifests as the isothermal magnetic entropy change (Δ S M ) and the adiabatic temperature change (Δ T ad ) when the magnetic field is applied or removed. Magnetic refrigeration technology based on MCE is an alternative technology over the commercial gas compression/expansion refrigeration because of its promising advantages (high energy efficiency, environmental conservation, small noise, etc.) [1][2][3][4] .
The MCE is the essential result of the magnetic entropy change due to the coupling of a magnetic spin system with magnetic field, and it is significant around the magnetic phase transition. Despite it has been well recognized that the magnetic phase transitions can be tuned by pressure [9][10][11] , only a few works are related to the hydrostatic pressure effect on MCE 12-17 . Morellon et al 12 . found that the external pressure can tune the magnetic phase transition and induce a giant MCE in Tb 5 Si 2 Ge 2 , whereas the MCE in Gd 5 Ge 2 Si 2 decreases evidently with increasing pressure 13 . The peak position of the magnetic entropy change Δ S M for La 0.69 Ca 0.31 MnO 3 shifts to higher temperatures gradually, while the maximum value of − Δ S M is almost unchanged with increasing pressure 14 . As a matter of fact, a weak pressure dependence on MCE has also been reported in some MCE materials, such as, GdCo 2 B 2 15 and GdCr 2 Si 2 16 compounds. Very recently, a giant reversible MCE in metamagnetic TmZn compound was reported 18 . To further understand the magnetic phase transition and its correlation with MCE, in this paper, we have further performed the high hydrostatic pressure magnetization measurements on TmZn.  18,19 . The magnetic properties of TmZn have been extensively investigated thirty years ago by the specific heat, resistivity, magnetization and neutron diffraction measurements [18][19][20][21][22] . The results indicated that  the strong field and temperature dependence of magnetic moment in TmZn cannot be described by a simple Rdderman-Kittel-Kasuya-Yosida (RKKY) model, and the low temperature ferromagnetic state in TmZn is probably due to the soft longitudinal spin fluctuations 23 , since the low temperature magnetization is not saturate even under fields approaching 10 T 18 .

Results and Discussion
To investigate the pressure effect on MCE in TmZn, a set of magnetic isothermal M(H) curves under the hydrostatic pressures of 0, 0.60 and 1.40 GPa with increasing and decreasing magnetic field up to 5 T for TmZn are measured. No obvious hysteresis can be observed under all the pressures for the whole temperature range. To ensure the readability of the figure, only several selected isotherms with increasing field for TmZn under 0, 0.60 and 1.40 GPa are presented in Fig. 2 for a comparison, and the corresponding Arrott plot (H/M versus M 2 ) curves are given in Fig. 3. Except some differences in values, the magnetic isotherms and the Arrott plots show a similar behavior under all the present pressures. I. e., a field-induced metamagnetic transition appears in a certain temperature range (around and above T C ), and the critical field shifts to higher magnetic fields with increasing temperature. Based on the Banerjee criterion 24 , the magnetic transition is first order if some of the H/M versus M 2 curves show negative slope at some points. Therefore, the present TmZn under all the present pressures reveal a typical field-induced first order metamagnetic transition, since a clear S-shape can be observed in the Arrott plots under all the pressures (as given in Fig. 3). In details, the magnetization jump during the metamagnetic transition and the temperature range of the metamagnetic transition is getting smaller with increasing pressure. Additionally, the slop of the Arrott plot related to the strength of first order transition is getting weak with increasing pressure. These behaviors indicate that the first order metamagnetic transition of TmZn is suppressed gradually with increasing hydrostatic pressure but not breakdown up to 1.40 GPa.

Conclusions
In summary, the magnetic phase transition and magnetocaloric effect in metamagnetic TmZn have been systematically investigated by magnetization measurements under high hydrostatic pressure up to 1.4 GPa. The Curie temperatures of T C are determined to be 8.4, 9.1 and 11.2 K under the pressures of 0, 0.60 and 1.40 GPa, respectively. The field-induced first order metamagnetic transition in TmZn is suppressed gradually with increasing hydrostatic pressure but not breakdown up to 1.40 GPa. The MCE in TmZn decreases gradually with increasing

Methods
The polycrystalline sample of TmZn was prepared by induction melting of the high purity Tm and Zn elements in a sealed Ta-tube. Firstly, high purity Tm and Zn with stoichiometric amounts were weighted and arc-welded in a Ta-tube under an argon pressure of ca. 75 kPa. Then the Ta-tube was placed in a water-cooled sample chamber of an induction furnace and heated up to 1250 K for five minutes, following by two hours annealing at 950 K. The sample was proved to be single phase by X-ray powder diffraction and Energy Dispersive X-ray Spectroscopy. The magnetization measurements under various hydrostatic pressures with DC magnetic fields up to 5 T were performed with a commercial superconducting quantum interference device (SQUID) magnetometer by Quantum Design (MPMS-5S) from 2 to 32 K. The sample was compressed in a homemade micro-CuBe pressure cell which was filled with the mixture of Florinerts 70 and 77 as the pressure transmitting medium. The hydrostatic pressure inside the cell was determined by the superconducting transition temperature of Sn.