WorksPhysics and Astronomy
Influence of Ni substitution on the phase transitions and magnetocaloric effect of NdCo2 at cryogenic temperatures
Abstract
We have investigated NdCo2-xNix cubic Laves compounds with 0 <= x <= 1 using neutron diffraction and bulk magnetization measurements to study the influence of partial Ni substitutions of Co on the phase transitions and the magnetocaloric effect. Upon cooling, NdCo2 undergoes a cubic to tetragonal transition at 100 K, and a tetragonal to orthorhombic transition at 42 K. The transitions are associated with long-range ferromagnetic ordering of the magnetic moments along the c axis and spin reorientation into the ab plane, respectively. Both transitions shift to lower temperatures as the Ni content x increases. For x >= 0.5, the orthorhombic phase is suppressed. Additionally, there was a reduction in the magnetic moment upon increasing the Ni substitution of Co. The magnetocaloric effect was determined both indirectly and directly, with good agreement between the methods. NdCo2 exhibits an adiabatic temperature change of 6.3 K for a field of 20 T, which is decreased to 4.9 K for NdCoNi for the same field strength due to the reduced magnetic moment upon Ni substitution.
00_Manuscript.pdf
Introduction
Magnetic refrigeration utilizing the magnetocaloric effect (MCE) offers superior cooling efficiency at cryogenic temperatures (20-77 K) compared to conventional gas compression-expansion methods [1]. Consequently, significant efforts have been made to discover materials with a large MCE in this temperature region. Laves compounds are a class of materials that have shown significant MCEs at low temperatures, but most high-performing candidates contain expensive and scarce heavy rare-earth elements with high magnetic moments, such as Ho, Gd, and Dy [1, 2, 3]. Therefore, there is a need to identify more cost-effective and abundant materials to scale up the technology. Replacing the heavy rare-earth elements in Laves compounds with more abundant light rare-earth elements at the expense of a lower MCE is one promising strategy to reduce cost [4].
NdCo is a well-studied light rare-earth cubic AB Laves phase (space group ) which is paramagnetic at room temperature. The compound exhibits a magnetostructural phase transition at 100 K associated with a tetragonal distortion (space group ) and ferromagnetic ordering (). The nature of this phase transition has been a subject of some debate in the literature. Most studies agree that the phase transition is of a second-order nature [5, 6, 7, 8, 9, 10]. However, studies based on perturbed angular correlation [11, 12] and resonant ultrasound spectroscopy [13] argue that the transition is first-order or close to tricritical in character. This discrepancy highlights the challenge of definitively characterizing magnetostructural transitions involving coupled magnetic and structural order parameters. Additionally, NdCo undergoes a first-order phase transition at 42 K associated with an orthorhombic distortion (space group ) and a spin reorientation () [5, 6, 7, 8, 9]. Since Co is a resource-critical metal, raising ethical and environmental concerns [14], partial substitution with other elements, such as Ni, can reduce the criticality of the compound. Our previous investigations of NdCoNi have revealed that the compound exhibits a at 36 K, but no [15, 16].
First-order magnetic transitions can exhibit high MCEs but are often accompanied by thermal hysteresis [17]. In contrast, second-order magnetic transitions typically result in both lower thermal hysteresis and MCE. The MCE is quantified by the magnetic entropy change and the adiabatic temperature change . To distinguish between the indirectly and directly determined in this work, we will denote these and , respectively. can either be determined from --diagrams constructed from heat capacity data, or as where is determined from field-dependent magnetization isotherms . The latter method can be unreliable for first-order phase transitions and at low temperatures, where approaches zero, and should be applied with caution to avoid an area of phase coexistence. can be determined by measuring the temperature change under application of a magnetic field. This direct is generally more reliable than the indirect , in particular for first-order magnetic transitions [18]. Pulsed-field measurements can be used to accurately determine by ensuring adiabatic conditions in the experiment [19].
In this work, with were studied to investigate the influence of Ni substitution on the MCE. The crystal structures, as well as local and bulk magnetic moments, were investigated in detail to understand how these properties are influenced by the partial substitution of Co with Ni. Studying the magnetic properties of a material is essential to understand the origin of its MCE. Therefore, neutron scattering was used to probe both magnetic and crystal structures of the compounds. These experiments are important for understanding the relationship between crystal structure, magnetic ordering, and MCE, to optimize the material performance.
Methods
Five (, 0.25, 0.50, 0.75, 1) compositions ( g batches) were synthesized by arc melting (Edmund Bühler, Compact Arc Melter MAM-1) under argon atmosphere ( %, Nippon Gases) from high-purity elements ( %, Goodfellow), with 3 wt.% excess of Nd to compensate for evaporation losses [20]. The chamber was evacuated and purged with argon gas three times, and left under a final pressure of 0.7 bar. The ingots were turned and remelted four times, followed by 72 hours of annealing at 1073 K and 72 hours at 833 K in evacuated and sealed quartz tubes to improve homogeneity. The final products were cooled to room temperature at a rate of 2 K/min and stored under argon atmosphere to prevent oxidation.
Microstructure and chemical compositions were characterized using a JEOL JSM-7900F scanning electron microscope (SEM) equipped with an Oxford Instruments AZtec Ultim Max 65 energy-dispersive x-ray (EDS) analyzer. Bulk pieces of the samples were embedded in ”polyfast” resin, ground, and polished. Back-scattered electron (BSE) mode was utilized to obtain elemental contrast, and EDS mode was used for elemental analysis, both using an acceleration voltage of 15 kV.
Synchrotron radiation powder x-ray diffraction (SR-PXD) data were obtained at BM01 at the Swiss-Norwegian Beamlines (SNBL) of the European Synchrotron Radiation Facility (ESRF) in Grenoble, France. Powder samples were prepared in 0.3 mm quartz capillaries and measured using the PILATUS2M detector [21] while rotating the capillary 360 degrees. All samples were measured at room temperature with a wavelength of 0.49537 Å. Additionally, was cooled to 80 K using a nitrogen blower, and measured with a wavelength of 0.71859 Å while heating with a rate of 1 K/min to 120 K, followed by 5 K/min to 300 K. The data were integrated with the SNBL Bubble software [21], and Rietveld refinements were performed using Topas V5 [22].
Powder neutron diffraction (PND) measurements were performed using the high-resolution powder diffractometer for thermal neutrons (HRPT) at the Swiss Spallation Neutron Source (SINQ) at the Paul Scherrer Institute (PSI), Villigen, Switzerland. Unless otherwise stated, measurements were performed using a wavelength of 1.886 Å and the high-intensity mode (40′ primary beam collimator). Measurements were performed between 1.5 and 120 K. At 1.5 K, additional measurements were performed using the medium-resolution mode (12′ primary beam collimator), and at a wavelength of 1.154 Å to determine the thermal displacement parameter. The medium-resolution measurements were also performed at 45 K for and . Refinements were performed using Mag2Pol [23], and the magnetic symmetry analysis was done using ISODISTORT [24, 25].
Magnetic and magnetocaloric properties were characterized using a Physical Property Measurement System (PPMS). Temperature-dependent magnetization was measured on zero-field-cooled samples during heating and cooling at a rate of 2 K/min, with applied fields of 0.01, 1, and 5 T, respectively. Hysteresis loops were obtained at 5 K between and T with a magnetic field-sweep rate of 0.3 T/min. measurements were performed during heating at 0 and 5 T, respectively, using the approach. and were determined from the vertical and horizontal difference between the 0 and 5 T curves in an --diagram, respectively. Additionally, isotherms were measured from 0 to 5 T with 4 K intervals in the temperature region of interest to calculate for comparison with the -based results. A comprehensive description of the and methodologies is provided in the Supplementary Material [26].
was measured at the Dresden High Magnetic Field Laboratory (HLD) at Helmholtz-Zentrum Dresden-Rossendorf (HZDR), Germany. Measurements were performed between 10 and 120 K for pulsed magnetic fields of 5, 10, and 20 T. For a comprehensive description of the experimental setup, see Ref. [19].
Results
Structural properties and phase transitions
Initially, the samples were characterized with SR-PXD and SEM-EDS at room temperature. All five compositions of the (, 0.25, 0.50, 0.75, 1) series were found to be of high purity, with minor secondary phases. This was quantified by Rietveld refinements and SEM-EDS analysis, see the Supplementary Material for details [26]. Both methods primarily show the C15 cubic AB Laves phase, but identify small amounts of AB, AB, or oxide impurities (1.0-3.3 wt.%). Note that none of these phases were identified by PND (see below). Rietveld refinements of the PND patterns for and are presented in Figure 1. No additional phases other than the main C15 phase were observed or included in the refinements. The refinements for the other studied compounds are presented in the Supplemental Material [26]. The expected tetragonal and orthorhombic phase transitions induce small structural deviations from the cubic cell, resulting in minor but resolvable peak splittings. The tetragonal and orthorhombic distortions can be most efficiently identified from the (620) Bragg peak. This peak is marked with a star in Figure 1 and shown in the insets. The evolution of the peak measured with SR-PXD from 80 to 300 K for is shown in the Supplemental Material [26]. The small peak splitting shows that a careful analysis is required to assign the correct symmetry.

Figure 1. Experimental data (blue circles), calculated diffraction pattern (red lines), and difference (grey lines) for the Rietveld refinements of the PND patterns collected with Å of NdCo (a)-(c) and NdCoNi (d)-(f) of the cubic (120 K, space group ), tetragonal (45 K, space group ), and orthorhombic (1.5 K, space group ) phases, respectively. Error bars are within the size of the markers. Vertical bars indicate Bragg reflections for the refined AB phase. The inset shows the (620) Bragg peak, marked with a star in the plot. Rietveld refinements for NdCoNi, NdCoNi, and NdCoNi are presented in the Supplemental Material [26].
All compounds exhibit the cubic structure (space group ) at 120 K. This can, e.g., be seen by the absence of splitting of the (620) Bragg peak. In the cubic phase, the Nd atoms occupy the 8a sites, while the Co and Ni atoms occupy the 16d sites [5]. The unit cell axis of at 120 K [Fig. 2] agrees with previously reported values [5]. The axis at this temperature shows a small and linear contraction with increasing Ni content (Supplemental Material [26]), following a Vegard’s law trend [27]. This is consistent with the larger lattice constants of (7.350 Å at room temperature [28]) compared to (7.270 Å at room temperature [29]), and the similar behavior previously reported for compounds [30].
Upon cooling, all compounds undergo a structural transition into a tetragonal phase (space group ), where the Nd atoms occupy the 4b sites, while the Co and Ni atoms occupy the 8c sites [5]. The tetragonal transition occurs at 100 K for , consistent with previous findings [5, 6, 9], and the transition temperature decreases with increasing Ni content. To avoid a jump in the plot of the unit cell axes resulting from the transitions from a face-centered cubic unit cell to a body-centered tetragonal unit cell, the pseudo-cubic lattice constants of are presented in Figure 2, where . The difference between and increases with decreasing temperature and decreases with increasing Ni substitution [Fig. 2]. The tetragonal splitting can be observed from the (620) Bragg peak in Figure 1 at 45 K, which is clearly split into two distinct peaks for . However, for , only a small shoulder and no distinct splitting is visible.
Figure 2. Unit cell axes obtained by Rietveld refinements of PND data at selected temperatures in the temperature interval 1.5 – 120 K. For the tetragonal and orthorhombic phases, the pseudo-cubic unit cell axes are shown for and (, , and ). The error bars indicate the error of the refinements. Vertical gray lines indicate the temperatures of the phase transitions.
Upon cooling to 1.5 K, , , and remain in the tetragonal symmetry, while and undergo another phase transition into an orthorhombic phase with space group [5]. For this phase, the Nd atoms occupy the 4e sites, while the Co/Ni site is split into the 4a and 4d sites [5], as also supported by Mössbauer studies [8]. As for the tetragonal phase, the pseudo-cubic and are shown in Figure 2. The differences between , , and are larger for than [Fig. 2]. The (620) Bragg peak [Fig. 1] is split into two peaks for , while for the orthorhombic distortion is only visible in the peak asymmetry and width.
Magnetic structures from neutron diffraction
PND was used to study the magnetic properties across the phase transitions. The cubic phases are paramagnetic, while the tetragonal and orthorhombic phases display long-range ferromagnetic order, described by the propagation vector . All the magnetic scattering is thus located on nuclear Bragg peaks. This can, e.g., be seen at the low- (111) peak at approximately Å, which shows a magnetic contribution and increasing intensity with decreasing temperature across all compounds. However, the structural deformation is too small to use low- peaks to distinguish the direction of the magnetic moments. Thus, we used the high- (620) Bragg peak at approximately Å for this purpose, although the peak exhibits a moderate magnetic contribution due to its high value [31].
Since the thermal displacement parameter exhibits a similar dependence, accurate determination of is essential to obtain a reliable magnetic moment from the refinement of the PND data. However, the decrease of the magnetic form factor is faster than the decrease of as a function of . Therefore, short wavelength PND measurements ( Å), providing a higher range, were used to refine at 1.5 K. This was locked when refining the magnetic moment and . At 120 K, all compounds were paramagnetic, and was refined, but the magnetic moment. For intermediate temperatures, extrapolated values were used to refine the magnetic moment. The temperature-dependence of for all compounds is shown in the Supplemental Material [26].
Figure 3 shows the cubic, tetragonal, and orthorhombic unit cells for the compounds, with corresponding magnetic moment vectors. The cubic cell exhibits no magnetic moment. The irreducible representation describes the tetragonal deformation of the cubic cell, while the magnetic order follows , giving a magnetic space group [No. 141.557 Belov–Neronova–Smirnova (BNS)] in the tetragonal phase [32]. Our refinements show that the easy magnetization direction in this phase is along the [1] or -direction, as reported previously based on PND [5] and Mössbauer [8] data. Magnetic moments are present on the Nd and the Co/Ni sites.
Upon further symmetry reduction to orthorhombic symmetry, the irreducible representation describes the structural distortion, and the magnetic symmetry is equivalently described by or , both corresponding to (BNS: 74.559) [32]. The magnetic moments are then reoriented to the plane, mark-
Figure 3. (a) Paramagnetic cubic unit cell, (b) tetragonal unit cell with magnetic moments aligned along the axis, and (c) orthorhombic unit cell with magnetic moments aligned along the axis of the NdCoNi compounds. The figures are created using Mag2Pol [23].
ing the tetragonal-to-orthorhombic transition as a spin-reorientation transition. The origin of the spin rotation is based on changes in the crystal electric field [7, 8]. The space group does not distinguish between magnetic moments along the two short axes, as this corresponds to a positive or negative sign for the strain mode . This is supported by Mössbauer measurements, where the easy magnetization direction in the orthorhombic phase of NdCo is determined to be in the [110] direction of the cubic unit cell, i.e., the plane of the orthorhombic cell [8].
A previous study on the refinement of PND measurements of NdCo has concluded that only the magnetic moments refined in the -direction yield stable results [5]. Figure 4 shows the differences between the refinement results in the present work when the magnetic moments are oriented along the -, -, and -direction. The result for the -direction is clearly different from the - and -directions, which are almost identical. Table I compares the magnetic moments and reliability factors for refinements of magnetic moments oriented along the - and -directions in the orthorhombic structure in the present work.
Figure 4. Simulated PND patterns of the (620) Bragg peak using Mag2Pol with magnetic moments oriented in the , , and directions, respectively, compared to experimental data of the NdCo compound measured at 1.5 K. The difference between and is also shown.
| Sample | Parameter | axis | axis |
| NdCo | () | 4.37(8) | 4.24(8) |
| 1.07 | 1.08 | ||
| (%) | 3.37 | 3.40 | |
| NdCoNi | () | 3.77(9) | 4.05(5) |
| 1.16 | 1.16 | ||
| (%) | 7.49 | 7.48 |
Table I. PND refinements of the magnetic moments oriented along the and unit cell axes in the orthorhombic phase. is the magnitude of the total magnetic moment, while and provide the reliability factors of the refinements.
Based on our data and stable refinements, we cannot unambiguously determine whether the moments lie along or in the orthorhombic phase, as both solutions exhibit reasonable magnetic moments and errors. However, our refinements exclude the possibility of magnetic moments orienting along the -direction in the orthorhombic phase, confirming a spin-reorientation transition from to across the tetragonal-to-orthorhombic phase transition, as previously reported [7]. Thus, the
magnetic moments were oriented along the -direction to be consistent with previous reports [5].
Figure 5 shows the temperature dependence of the atomic magnetic moments of the compounds, obtained from refinements of the PND data. The refinements were performed with magnetic moments oriented along the axis in the tetragonal cells and along the axis in the orthorhombic cells, as discussed above. All compounds are paramagnetic above the Curie temperature , with increasing magnetic moments below this temperature. The magnetic moment of Nd is larger than that of Co/Ni for all compounds. Both the magnetic moments of Co/Ni and Nd decrease with increasing Ni substitution. There is a slight change in the magnitude of the magnetic moments of NdCo around .
Figure 5. Atomic magnetic moments (/atom) and total magnetic moment for the compounds (/f.u.) as a function of temperature, derived from refinements of the PND data. Since Co and Ni occupy the same site, one cannot differentiate between them using PND. Therefore, the Co/Ni curves show an average between these two for each position. The total magnetic moment is calculated as the sum of the Nd and two times the Co/Ni magnetic moment. The vertical gray dotted lines mark for NdCo and NdCoNi.
The magnetic moments for NdCo are in line with those previously determined using PND [5, 9]. The magnetic moment of Nd in NdCo at 1.5 K is 2.67 , and as expected, somewhat lower than the theoretical value of . The magnetic moment of Co in NdCo at 1.5 K is 2·0.75 /f.u. and the Ni/Co moment in NdCoNi is 2·0.51 /f.u. Assuming a constant Co moment across the series gives 0.27 /f.u. for the moment of Ni in NdCoNi. The reduction of in PrCoNi has been attributed to a strong decrease in the exchange interactions from PrCo to PrCoNi as a result of the Co atoms being replaced by the more weakly magnetic Ni as the nearest neighbors to Pr [30]. A similar explanation could apply to the trends observed in this work, but additional studies are needed to confirm this.
Bulk magnetic properties
Figure 6 shows bulk magnetization results. Values for the saturation magnetization , remanent magnetization , and coercivity determined from these measurements are given in the Supplemental Material [26], along with temperature-dependent magnetization measurements performed at fields of 1 and 5 T, respectively. All compounds exhibit a small magnetic hysteresis at low fields and magnetization. NdCo exhibits the smallest hysteresis, and NdCoNi the largest.
The bulk of NdCo (3.99 ) is slightly lower than the total magnetic moment derived from PND refinements (4.24 ), which can be attributed to a lack of saturation at the applied field of 5 T. The bulk decreases drastically upon substitution of only Ni, from 3.99 to 3.05 . There is a smaller but still significant decrease upon substitution to Ni (to 2.37 ), while subsequent Ni substitution has a negligible effect on . The reduction in upon Ni substitution of Co is significantly stronger than expected based on their difference in magnetic moments according to Hund’s rules [33], and the PND results Figure 5. However, a similar strong change in magnetic moment has been observed in bulk magnetization measurements of PrCoNi as a function of [30]. It is known that Co couples ferromagnetically with light rare-earth elements in cubic Laves compounds [34].
The Curie temperature was determined from the maximum of the derivative of the temperature-dependent magnetization, as shown in Figure 6(b) inset, using an applied magnetic field of 0.01 T to minimize field-induced shifts. The of NdCo (100 K) and NdCoNi (34 K) are consistent with previous reports [5, 15], while the intermediate stoichiometries exhibit values between those two end members (56, 48, and 42 K, respectively). A similar trend is observed in PrNiCo [30], where decreases with increasing Ni substitution. This can be attributed to weaker exchange interactions as the Ni content, which strongly influences in RCo compounds, where is a rare-earth element [6].
Figure 6. Bulk magnetization for the five selected compositions of the NdCoNi system. (a) Field-dependent magnetization () measured at 5 K, where the inset displays the magnetization around the origin. (b) Temperature-dependent magnetization (5 K 150 K) at 0.01 T, where the inset shows the derivative of the magnetization as a function of temperature. The maximum of the derivative was used here to determine . The arrows indicate the magnetization data corresponding to heating and cooling.
A small thermal hysteresis can be observed in the tetragonal phase Figure 6(b), the cubic phase is paramagnetic, and the orthorhombic phase exhibits no visible hysteresis. Thus, the spin reorientation associated with the tetragonal-to-orthorhombic phase transition can be observed in the bulk magnetization data at the low-temperature end of the hysteresis for NdCo (42 K) and NdCoNi (22 K). This can be observed as a small bump at the temperature-dependent curves measured at 1 T, whereas no transition is visible at 5 T, as the high external field suppresses it (see Supplemental Material [26]).
Magnetocaloric effect
To determine the order of the phase transition, an exponent analysis was performed using the curves obtained from the data. As reported in [17], a local exponent greater than 2 signifies a first-order magnetic transition. This analysis confirms that the transitions at of NdCo and NdCoNi (at 42 and 22 K, respectively) are first-order phase transitions, while the transitions of all compounds at are second-order phase transitions.
The MCE was characterized by and for all samples. Additionally, was determined by pulsed-field measurements for NdCo, NdCoNi, and NdCoNi. and at 5 T extracted from - diagrams are shown in Figure 7, while the at 5, 10, and 20 T are shown in Figure 8. A comparison of the results obtained using the different methods is presented in the Supplemental Material [26]. The agreement between and is satisfactory, with the direct method yielding slightly higher values. should be the more reliable method in particular for the first-order phase transitions, i.e., [18]. This is particularly evident for the NdCo sample in the present work.

Figure 7. (a) and (b) results obtained at 5 T from S-T diagrams based on measurements.

Figure 8. (a) Pulsed-field measurements of around of NdCo (100 K), NdCoNi (56 K), and NdCoNi (34 K), and of NdCo (42 K) and NdCoNi (22 K). (b) Time-dependent during the magnetic field pulse applied to the NdCo and NdCoNi compounds at 105 K.
A strong MCE is associated with of all compounds, and with of NdCo and NdCoNi. The MCEs at exceed those observed at . The Ni-substituted compounds exhibit MCEs that span most of the temperature region needed for magnetocaloric hydrogen liquefaction (20 to 77 K), whereas the MCE of NdCo appears at higher temperatures. A slight decrease in is observed with increasing Ni substitution, but does not follow this trend, except for NdCo displaying the highest value. NdCo has previously been reported to exhibit, at , a of 7.3 J/kgK and a of 3.1 K at 5 T [6]. In comparison, the we obtained in the present study (9.0 J/kgK) is higher, while the is comparable when determined from measurements (3.1 K). We have previously studied the MCE of NdCoNi [15], yielding slightly higher values for and compared to the present work using the method but slightly lower values using the method, respectively.
Figure 8(a) shows the evolution of at (NdCo and NdCoNi) and (NdCo, NdCoNi and NdCoNi) for fields up to 20 T. As expected, is lower than for heavy rare-earth-based compounds measured using similar techniques, including Co Laves compounds where heavy rare-earth elements were used [18, 35, 36]. returned reversibly to 0 K by 150 ms for all applied field [Figure 8(b)]. The MCE at increases more steeply than at with increasing field strength. This difference was quantified by calculating . At 5 T, this ratio is 58 and 46% for NdCo and NdCoNi, respectively, while at 20 T the values are 45 and 39%, respectively. Additionally, there is a difference between the field dependence of of the different compounds. Considering , NdCoNi exhibits a slightly higher than NdCoNi at 5 T, while at 20 T the of NdCoNi is higher. The at of NdCo exhibits an even stronger dependence on applied field.
Discussion
Ni substitution of Co in NdCo reduces the tetragonal and orthorhombic distortions [Fig. 1 and Fig. 2]. This observation can be attributed to differences in the atomic radii of Co and Ni. The same trend is seen for NdCoFe, where the larger Fe atoms increase the difference between and in the tetragonal unit cell, from 0.032 Å for NdCo to 0.048 Å for NdCoFe [6].
Furthermore, the Ni substitution shifts both the tetragonal and the orthorhombic phase transitions towards lower temperatures. Figure 9 shows the phase diagram of the NdCoNi system, depicting the various crystalline phases present over the temperature range from 1.5 to 105 K. Since both transitions are coupled magnetostructural, their shifts can be attributed either to a shift in the exchange constant driving [4], or the difference in atomic radii of Ni and Co driving the structural distortion. A similar shift in the transition temperature was seen in NdCoFe, attributed to the
Figure 9. Phase diagram over the paramagnetic (PM) cubic, ferromagnetic (FM) tetragonal, and FM orthorhombic phases, with the directions of the magnetic moments. The solid circles indicate measured transition temperatures, while the lines are interpolated between these.
Fe substitution increasing , and thus the tetragonal phase transition temperature [6].
The direction of the magnetic moments in the orthorhombic phase could not be determined unambiguously between the - and -directions from our PND data. There are minor differences between the simulated patterns of moments oriented along these two directions for the high (620) Bragg peak of NdCo [Fig. 4], where the peak splitting is observed for the orthorhombic phase. Unambiguous determination of the direction of the magnetic moments could be carried out with single-crystal neutron diffraction on an untwinned crystal. For compounds with Ni substitution, the peak splitting decreases with increasing substitution, requiring even higher resolution to determine the direction of the magnetic moments.
The determined from PND refinements Figure 5 is similar to the determined from bulk magnetization measurements Figure 6(a) for NdCo, 4.24 and 3.99 , respectively. However, upon substitution of Ni, decreases more rapidly than . For NdCoNi, is 3.38 , while is 2.42 . Part of the discrepancy might be attributed to the measurement temperature of (5 K) being closer to than the measurement temperature of (1.5 K), in particular for compounds with high Ni substitution where is low. The field-dependent bulk magnetization Figure 6(a) of all the compounds is not completely saturated at 5 T, but increases by approximately 0.05 /T. As we have seen from x-ray magnetic circular dichroism experiments of NdCoNi, the increase can be attributed to van Vleck paramagnetism of Nd [16, 37].
The exponent analysis applied to the data agrees with most of the present published data that the phase transition at of NdCo is of the second-order while the transition at is of the first-order (Supplemental Material [26]). As the Ni substitution increases along the NdCoNi series, the exponent approaches 2 more closely, although the differences between the compositions are small. The exponent does not overshoot above 2 for any of the compounds, which would indicate a first-order phase transition [17]. decreases away from 2 in high fields, where the transition is suppressed [35].
The decrease in MCE from NdCo upon Ni substitution is attributed to the reduced magnetic moment. The highest MCE was observed at , rendering the suppression of at high Ni content irrelevant for applications. The compounds studied in this work cover the lower end of the temperature range required for magnetocaloric hydrogen liquefaction, while there is a gap at the higher end, near 77 K. The Ni substitution allows tuning of , enabling the possibility of covering the entire magnetocaloric hydrogen liquefaction region from 20 to 77 K by fine-tuning the stoichiometry of NdCoNi with .
Conclusions
The structural and magnetic phase transitions, as well as the magnetocaloric effects, of five NdCoNi ( 0, 0.25, 0.50, 0.75, 1) cubic Laves compositions have been studied for possible magnetic refrigeration applications in a temperature range relevant for hydrogen liquefaction. All compounds display a phase transition from a paramagnetic cubic phase to a ferromagnetic tetragonal phase with magnetic moments oriented along the -direction upon cooling. The Co-rich samples undergo a spin-reorientation transition associated with a further symmetry reduction to an orthorhombic space group. Both the tetragonal and orthorhombic distortions and transition temperatures decreased with increasing Ni substitution. Bulk magnetization measurements show a larger decrease than neutron diffraction measurements. The phase transition orders were verified using exponent analysis. The magnetocaloric effect decreased slightly with increasing Ni substitution due to a reduced magnetic moment. The adiabatic temperature change was measured directly at 5, 10, and 20 T and was observed to scale more strongly with the applied field at the Curie temperature than at the spin-reorientation temperature.
Acknowledgments
This work was financed by the Research Council of Norway under project number 336403. The authors gratefully acknowledge support from the Clean Hydrogen partnership and its members within the framework of the project HyLICAL (Grant No. 101101461). The authors also acknowledge NcNeutron, financed by the Research Council of Norway under project number 245942. The authors thank the staff at SNBL (BM01 and BM31) of the ESRF for their skilled assistance during the measurements. Part of this work was performed at the Swiss Spallation Neutron Source (SINQ), Paul Scherrer Institut (PSI), Villigen, Switzerland. We acknowledge support from Deutsche Forschungsgemeinschaft (DFG) through CRC/TRR 270 (Project-ID 40553726) and through Würzburg-Dresden Cluster of Excellence on Complexity and Topology in Quantum Matter—ct.qmat (EXC 2147, Project No. 390858490) and the support of the HLD at HZDR, member of the European Magnetic Field Laboratory (EMFL).
Data Availability
The data that support the findings of this article are openly available [38].
Supplemental material for ”Influence of Ni substitution on the phase transitions and magnetocaloric effect of NdCo at cryogenic temperatures”
Vilde G. S. Lunde; Øystein S. Fjellvåg; Allan M. Döring; Marc Straßheim; Vladimir Pomjakushin; Konstantin P. Skokov; Oliver Gutfleisch; Tino Gottschall; Joachim Wosnitza; Anja O. Sjåstad; Bjørn C. Hauback; Christoph Frommen
The supplemental material contains scanning electron microscope-energy-dispersive x-ray (SEM-EDS) results [Section I], synchrotron radiation powder x-ray diffraction (SR-PXD) data [Section II], powder neutron diffraction (PND) data [Section III], as well as bulk magnetization and magnetocaloric effect data [Section IV].
00_Supplemental.pdf
SEM-EDS ANALYSIS
SR-PXD DATA
Figure A2. SR-PXD data obtained of the (620) Bragg peak of NdCo at temperatures ranging from 80 to 300 K. The sample was cooled down using a nitrogen blower and measured during heating. Below 100 K, peak asymmetry is observed, and below 95 K, the peak is clearly split. .

Figure A3. Rietveld refinement (red lines) of SR-PXD data (blue circles) obtained at BM01 of all compounds at room temperature using Å. Grey lines indicate the difference between the measured data and the refinement. Vertical bars indicate Bragg reflections for the main AB cubic Laves phase (C15), as well as the secondary AB (space group ) and AB (space group ) phases with their respective weight percentages as determined from the refinement. A is Nd, and B is the stoichiometric transition metal composition for the given compound.
Neutron Scattering

Figure A4. Experimental (blue circles) and calculated diffraction patterns (red lines), and difference (grey lines) for Rietveld refinements of PND data of NdCoNi, NdCoNi, and NdCoNi in a cubic (120 K, space group ) and tetragonal (45 K, space group ) phase, using Å. Error bars are within the size of the markers. Vertical bars indicate Bragg reflections for the refined AB phase. The inset shows the (620) Bragg peak, marked with a star in the plot.
Figure A5. Lattice constants in of the cubic unit cell axis obtained by Rietveld refinements of the C15 cubic Laves phase using PND data at 120 K using Mag2Pol. Error bars indicate the error of the refinement. Values are tabulated in Table A2.
Figure A6. Unit cell volume () obtained by Rietveld refinements of PND data at selected temperatures in the temperature interval from 1.5 to 120 K. The error bars indicate the error of the refinements. Vertical gray lines indicate the phase transitions. The pseudo-cubic volume is shown for the tetragonal and orthorhombic structures, which are twice those of the corresponding tetragonal and orthorhombic unit cell volumes, respectively.

Figure A7. Thermal displacement parameter , determined at 1.5 and 120 K through refinements of PND data, while interpolated values using these two data points were used for intermediate temperatures. The error bars indicate the calculated error of the refinements. Tabulated values are given in Table A2.
Table A2. Results from the refinements of the PND data using Mag2Pol. The first part of the table shows the results for the compounds in their cubic state at 120 K. The second part shows the results for the compounds in their tetragonal states: 45 K for NdCo and NdCoNi, and 1.5 K for the remaining compounds. The last part of the table shows the results for NdCo and NdCoNi in their orthorhombic state at 1.5 K. The magnetic moments are refined along the axis in the tetragonal phase and along the axis in the orthorhombic phase.
| 120 K: Cubic | NdCo | NdCoNi | NdCoNi | NdCoNi | NdCoNi |
| (Å) | 7.2832(2) | 7.27837(3) | 7.2764(3) | 7.2745(3) | 7.27066(2) |
| (Å) | 386.34(4) | 385.56(5) | 385.26(6) | 384.95(5) | 384.34(3) |
| (Å) | 0.28(2) | 0.53(2) | 0.44(2) | 0.47(2) | 0.51(1) |
| (Å) | 0.45(3) | 0.67(3) | 0.78(3) | 0.71(2) | 0.64(2) |
| 2.22 | 2.12 | 1.97 | 2.18 | 2.21 | |
| 45 K / 1.5 K: Tetragonal | NdCo | NdCoNi | NdCoNi | NdCoNi | NdCoNi |
| (Å) | 5.15719(6) | 5.1467(3) | 5.1450(2) | 5.1428(2) | 5.1394(3) |
| (Å) | 7.2659(1) | 7.2684(3) | 7.2636(2) | 7.2627(3) | 7.2600(3) |
| (Å) | 193.24(1) | 192.53(5) | 192.27(4) | 192.08(4) | 191.76(5) |
| (Å) | 0.18(2) | 0.36(2) | 0.31(2) | 0.28(1) | 0.32(1) |
| (Å) | 0.31(3) | 0.41(3) | 0.31(2) | 0.33(2) | 0.35(1) |
| () | 2.53(3) | 2.30(4) | 2.72(3) | 2.61(3) | 2.36(3) |
| () | 0.75(2) | 0.55(3) | 0.63(3) | 0.63(3) | 0.51(2) |
| 1.86 | 1.92 | 1.90 | 2.03 | 2.08 | |
| 1.5 K: Orthorhombic | NdCo | NdCoNi | - | - | - |
| (Å) | 5.1504(5) | 5.1476(4) | - | - | - |
| (Å) | 5.1427(5) | 5.1385(4) | - | - | - |
| (Å) | 7.2965(5) | 7.2832(4) | - | - | - |
| (Å) | 193.26(5) | 192.64(4) | - | - | - |
| 0.3739(8) | 0.3727(5) | - | - | - | |
| (Å) | 0.13(2) | 0.26(2) | - | - | - |
| (Å) | 0.23(3) | 0.25(3) | - | - | - |
| () | 2.67(6) | 2.80(4) | - | - | - |
| () | 0.75(4) | 0.63(2) | - | - | - |
| 1.08 | 1.16 | - | - | - |
Figure A8. Curve fitting using Gaussian functions of the (620) Bragg peak at 1 and 45 K measured using PND. The first row shows the data of the tetragonal structures, and the second row shows the data of the orthorhombic structures.

(a) (111) Bragg peak

(b) (620) Bragg peak
Figure A9. Intensities and integrated peak intensities of the (a) (111) and (b) (620) Bragg peaks for all the five NdCoNi () compounds at selected temperatures in the range from 1.5 to 120 K. The integrated intensities are approximated by subtracting the background around the given peaks. The low- (111) peak shows a strong correlation with the magnetic moment, and a drop in the integrated peak intensity can be seen upon heating above . The high- (620) peak was used to observe the tetragonal and orthorhombic distortions at low temperatures.
Bulk Magnetization, Magnetocaloric, and Heat Capacity Measurements
Detailed Indirect Magnetocaloric Measurement Methodology
This section presents a detailed description of the methodology for the measurement and analysis of the data used to determine the magnetocaloric effect indirectly in this work. and were determined using three different methods (, , and directly), as presented together in Figure A12 c) and d).
Figure A12. Temperature-dependent bulk magnetization measured at (a) 1 T and (b) 5 T, respectively. (c) and (d) obtained at 5 T from curves, S-T diagrams from data, and direct pulsed-field measurements. (e) Curie-Weiss fitting if the inverse susceptibility at 1 T, used to determine and .
From : Illustrated in Figure A10(a). measurements provide iso-magnetic entropy versus temperature diagrams. This is obtained by integrating over temperature for each field, in this work: T and T. From these curves, we can calculate and because:
Figure A10. Illustration of the determination of and on NdCoNi data. (a) data (upper) used to find the total entropy (lower) at 0 and 5 T. The vertical and horizontal distances between these determine and , respectively. (b) curves (upper) used to determine (lower) at different fields.
From : Illustrated in Figure A10(b). Field-dependent magnetization curves were measured at constant temperatures from 0 to 5 T in 4 K increments. From these curves, is calculated:
Figure A11. The markers show the heat capacity data measured at 0 and 5 T with the approach. Lines connect data points to guide the eye.
Figure A13. exponent analysis of data shown for different magnetic fields between 1 and 5 T. is marked with a horizontal dotted line since an exceeding 2 indicates a first-order phase transition. Markers show the data points, while the lines connect these data points to guide the eye.
Figure A14. Pulsed-field results obtained at HLD. (a) results using 5, 10, and 20 T. (b) measured at 105 K as a function of magnetic field for pulsed-fields of 5, 10, and 20 T, showing the evolution of these. (c) of the (42 K for NdCo, 22 K for NdCoNi) relative to the (100 K for NdCo, 56 K for NdCoNi) as a function of field.
| Compounds | (/f.u.) | (/f.u.) | (/f.u.) | (/f.u.) | (mT) | (K) | (K) | |
| NdCo | 1.84 | 3.62 | 7.17 | 3.99 | 0.19 | 10 | 95 | 100 |
| NdCoNi | 1.84 | 3.62 | 5.81 | 3.05 | 0.47 | 44 | 55 | 56 |
| NdCoNi | 1.84 | 3.62 | 5.42 | 2.37 | 0.56 | 55 | 46 | 48 |
| NdCoNi | 1.84 | 3.62 | 4.36 | 2.30 | 1.17 | 62 | 47 | 42 |
| NdCoNi | 1.84 | 3.62 | 4.28 | 2.42 | 0.80 | 39 | 39 | 34 |
Table A3. Calculated and experimental magnetic properties of the studied compounds. The de Gennes factor and the theoretical effective magnetic moment are calculated for Nd. The experimental effective magnetic moment and the paramagnetic Curie temperature were determined through Curie–Weiss fitting using a field of 1 T. The saturation magnetization and the Curie temperature were determined from temperature-and field-dependent measurements, respectively, using a field of 5 T and 0.01 T, respectively. Remnant magnetization and coercivity were determined from field-dependent measurements at 5 K.
TABLE A4: Magnetocaloric effect results. represents the magnetic transition temperatures, both and . is determined from isotherms, and the relative cooling power is calculated from . is determined from and , while is measured directly using pulsed magnetic fields. Determining the of the peaks was problematic due to the small and broad nature of the peaks with tails. Thus, this parameter was only determined at .
| Parameter | (K) | (J/kgK) | (J/kg) | (K) | (K) | ||
| Technique | / / Direct | / | Direct | ||||
| Field | 0.01 T | 5 T | 5 T | 5 T | 5 T | 10 T | 20 T |
| 100 | 9.0 / 7.8 / 7.8 | 222 | 3.1 / 2.7 | 2.7 | 4.0 | 6.3 | |
| 42 | 2.2 / 1.0 / 4.3 | - | 0.8 / 0.4 | 1.6 | 2.0 | 2.5 | |
| 56 | 7.2 / 7.0 / 8.0 | 188 | 2.1 / 2.1 | 2.4 | 3.0 | 5.1 | |
| 22 | 1.6 / 1.8 / 2.4 | - | 0.9 / 0.9 | 1.1 | 1.4 | 2.0 | |
| 48 | 6.2 / 6.1 / - | 184 | 1.9 / 1.9 | - | - | - | |
| 42 | 5.9 / 6.2 / - | 177 | 1.9 / 2.0 | - | - | - | |
| 34 | 5.9 / 6.2 / 6.9 | 173 | 2.0 / 2.2 | 2.5 | 3.3 | 4.9 |
Table A4.

Figure A1. Representative back-scattered electron image (left) and EDS measurements (right) of the NdCoNi compound, illustrating the microstructure typical for the compounds in this study. The main phase is the AB Laves phase. There is a small oxygen contamination because the SEM-EDS sample preparation was performed in ambient air. The lighter areas are a NdNi secondary phase, thus Co-deficient. Percentages of secondary phases for all analyzed compounds are tabulated in Table A1.
| Primary AB phase | Secondary phase | area % |
| NdCo | NdO | 0(0) |
| NdCoNi | NdO | 0.3(1) |
| NdNi | 3(1) | |
| NdCoNi | NdO | 0.2(1) |
| NdNi | 1.5(4) | |
| NdCoNi | NdO | 0.3(1) |
| NdNi | 1.5(6) | |
| NdCoNi | Nd(CoNi) | 2.1(5) |
Table A1. Composition and quantity (area %) of secondary phases according to SEM-EDS analysis, determined as illustrated in Figure A1. The remaining percentage corresponds to the primary AB Laves phase, where A is Nd and B is the stoichiometric transition metal composition of the given compound. The area percentage of the secondary phases is calculated from the average and standard deviation across three BSE images acquired at different positions within the same cross-section, all using the same scale.
References
- [1]V. Franco, J. Blázquez, J. Ipus, J. Law, L. Moreno-Ramírez, and A. Conde, Magnetocaloric effect: From materials research to refrigeration devices, Progress in Materials Science 93, 112 (2018).
- [2]P. B. D. Castro, K. Terashima, T. D. Yamamoto, Z. Hou, S. Iwasaki, R. Matsumoto, S. Adachi, Y. Saito, P. Song, H. Takeya, and Y. Takano, Machine-learning-guided discovery of the gigantic magnetocaloric effect in HoB₂ near the hydrogen liquefaction temperature, NPG Asia Materials 12, 35 (2020).DOI
- [3]I. Tereshina, A. Karpenkov, A. Kurganskaya, V. Chzhan, S. Lushnikov, V. Verbetsky, E. Kozlyakova, and A. Vasiliev, Effects of composition variation and hydrogenation on magnetocaloric properties of the (Gd₁₋ₓTbₓ)Ni (x = 0.1; 0.9) compounds, Journal of Magnetism and Magnetic Materials 574, 170693 (2023).
- [4]W. Liu, T. Gottschall, F. Scheibel, E. Bykov, A. Aubert, N. Fortunato, B. Beckmann, A. M. Döring, H. Zhang, K. Skokov, and O. Gutfleisch, A matter of performance and criticality: A review of rare-earth-based magnetocaloric intermetallic compounds for hydrogen liquefaction, Journal of Alloys and Compounds 995, 174612 (2024).DOI
- [5]Y. Xiao, Q. Huang, Z. Ouyang, J. Lynn, J. Liang, and G. Rao, Crystal and magnetic structures of Laves phase compound NdCo₂ in the temperature range between 9 and 300K, Journal of Alloys and Compounds 420, 29 (2006).
- [6]A. Murtaza, W. Zuo, J. Mi, Y. Li, A. Ghani, M. Yaseen, M. Tahir Khan, C. Hao, K. Li, Z. Dai, S. Yang, and Y. Ren, Magnetocaloric effect and critical exponent analysis around magnetic phase transition in NdCo₂ compound, Journal of Physics D: Applied Physics 53, 345003 (2020).
- [7]G. Dublon, M. Kroupp, M. P. Dariel, and U. Atzmony, Crystal Field Induced Spin Rotation Observed by Magnetization Study of Polycrystalline NdCo₂, physica status solidi (b) 76, 669 (1976).
- [8]U. Atzmony, M. P. Dariel, and G. Dublon, Crystal-field-induced spin rotations in NdCo₂ and HoCo₂: A Mössbauer study, Physical Review B 14, 3713 (1976).
- [9]Z. W. Ouyang, F. W. Wang, Q. Huang, W. F. Liu, Y. G. Xiao, J. W. Lynn, J. K. Liang, and G. H. Rao, Magnetic structure, magnetostriction, and magnetic transitions of the Laves-phase compound NdCo₂, Physical Review B 71, 064405 (2005).
- [10]J. Herrero-Albillos, F. Bartolomé, L. M. García, F. Casanova, A. Labarta, and X. Batlle, Nature and entropy content of the ordering transitions in RCo₂, Physical Review B 73, 134410 (2006).
- [11]M. Forker, S. Müller, P. de la Presa, and A. F. Pasquevich, Perturbed angular correlation study of the magnetic phase transitions in the rare-earth cobalt Laves phases RCo₂, Phys. Rev. B 68, 014409 (2003).
- [12]M. Forker, S. Müller, P. De La Presa, and A. Pasquevich, Comment on “Nature and entropy content of the ordering transitions in RCo₂”, Physical Review B 75, 187401 (2007).
- [13]S. L. Driver, J. Herrero-Albillos, C. Bonilla, F. Bartolomé, L. García, C. Howard, and M. Carpenter, Multiferroic (ferroelastic/ferromagnetic/ferrimagnetic) aspects of phase transitions in RCo₂ Laves phases, Journal of Physics: Condensed Matter 26, 056001 (2014).
- [14]European Commission, Study on the critical raw materials for the EU 2023 – final report (2023).
- [15]V. G. Lunde, A. B. Møller, B. G. Eggert, A. M. Döring, J.-C. Grivel, R. Bjørk, F. Veillon, K. Skokov, O. Gutfleisch, A. O. Sjåstad, B. C. Hauback, and C. Frommen, Machine learning guided discovery and experimental validation of light rare earth Laves phases for magnetocaloric hydrogen liquefaction, Acta Materialia 297, 121282 (2025).
- [16]V. G. S. Lunde, B. S. Ofstad, S. Fjellvåg, P. Ohresser, A. O. Sjåstad, B. C. Hauback, and C. Frommen, Electronic and magnetic properties of light rare-earth cubic Laves compounds derived from x-ray magnetic circular dichroism, Physical Review B 112, 224413 (2025).
- [17]J. Y. Law, V. Franco, L. M. Moreno-Ramírez, A. Conde, D. Y. Karpenkov, I. Radulov, K. P. Skokov, and O. Gutfleisch, A quantitative criterion for determining the order of magnetic phase transitions using the magnetocaloric effect, Nature Communications 9, 2680 (2018).
- [18]G. Politova, I. Tereshina, A. Karpenkov, V. Chzhan, and J. Cwik, Magnetism, magnetocaloric and magnetostrictive effects in RCo₂ – type (R = Tb, Dy, Ho) Laves phase compounds, Journal of Magnetism and Magnetic Materials 591, 171700 (2024).
- [19]C. Salazar Mejía, T. Niehoff, M. Straßheim, E. Bykov, Y. Skourski, J. Wosnitza, and T. Gottschall, On the high-field characterization of magnetocaloric materials using pulsed magnetic fields, Journal of Physics: Energy 5, 034006 (2023).
- [20]R. T. Macaluso and B. K. Greve, Challenges in intermetallics: synthesis, structural characterization, and transitions, Dalton Transactions 41, 14225 (2012).DOI
- [21]V. Dyadkin, P. Pattison, V. Dmitriev, and D. Chernyshov, A new multipurpose diffractometer PILATUS@SNBL, Journal of Synchrotron Radiation 23, 825 (2016).DOI
- [22]R. E. Dinnebier, A. Leineweber, and J. S. O. Evans, Rietveld refinement practical powder diffraction pattern analysis using TOPAS, Journal of Applied Crystallography 52, 1238 (2019).
- [23]N. Qureshi, Mag²Pol: a program for the analysis of spherical neutron polarimetry, flipping ratio and integrated intensity data, Journal of Applied Crystallography 52, 175 (2019).arxiv.org/abs/1801.08431
- [24]H. T. Stokes, D. M. Hatch, and B. J. Campbell, ISODIS-TORT, ISOTROPY Software Suite, iso.byu.edu.
- [25]B. J. Campbell, H. T. Stokes, D. E. Tanner, and D. M. Hatch, ISODISPLACE: An internet tool for exploring structural distortions, Journal of Applied Crystallography 39, 607 (2006).
- [26]See Supplemental Material [url] for SEM-EDS and SR-PXD results, as well as additional PND and bulk magnetization data.
- [27]A. R. Denton and N. W. Ashcroft, Vegard’s law, Phys. Rev. A 43, 3161 (1991).pubmed.ncbi.nlm.nih.gov/9905387
- [28]E. Gratz, A. Lindbaum, A. S. Markosyan, H. Mueller, and A. Y. Sokolov, Isotropic and anisotropic magnetoe-lastic interactions in heavy and light RCo₂ Laves phase compounds, Journal of Physics: Condensed Matter 6, 6699 (1994).
- [29]J. H. Wernick and S. Geller, Rare-earth compounds with the MgCu₂ structure, Trans. Met. Soc. AIME 218 (1960).
- [30]A. Ermolenkо, A. Korolev, E. Gerasimov, V. Gaviko, P. Terentev, and N. Mushnikov, Compositional genesis of ferromagnetism in alloys PrNi₁₂−ₓCoₓ, Journal of Magnetism and Magnetic Materials 490, 165489 (2019).
- [31]M. Blume, A. J. Freeman, and R. E. Watson, Neutron magnetic form factors and x-ray atomic scattering factors for rare-earth ions, The Journal of Chemical Physics 37, 1245 (1962).DOI
- [32]M. Talanov, V. Shirokov, M. Pimenov, and V. Talanov, Magnetic phase diagrams of the pyrochlore-based magnets: Landau theory, Journal of Magnetism and Magnetic Materials 575, 170717 (2023).DOI
- [33]S. Blundell, Magnetism in Condensed Matter, Oxford Master Series in Condensed Matter Physics (Oxford University Press, Oxford; New York, 2001).
- [34]J. Farrell and W. E. Wallace, Magnetic Properties of Intermetallic Compounds between the Lanthanides and Nickel or Cobalt, Inorganic Chemistry 5, 105 (1966).DOI
- [35]E. Bykov, A. Karpenkov, W. Liu, M. Straßheim, T. Niehoff, K. Skokov, F. Scheibel, O. Gutfleisch, C. Salazar Mejía, J. Wosnitza, and T. Gottschall, Magnetocaloric effect in the Laves phases RCo₂ (R = Er, Ho, Dy, and Tb) in high magnetic fields, Journal of Alloys and Compounds 977, 173289 (2024).DOI
- [36]R. Dragland, C. Salazar Mejía, I. Hansen, Y. Hamasaki, E. A. C. Panduro, Y. Ehara, T. Gottschall, D. Meier, and J. Schultheiß, Relation between 4f-magnetism and the low-temperature magnetocaloric effect in multiferroic hexagonal manganites, Communications Materials 6, 95 (2025).DOI
- [37]W. Wallace and K. Mader, Magnetic characteristics of Pr₂Y₁−zNi₂ alloys and the nature of PrNi₂ at low temperatures, Inorganic Chemistry 7, 1627 (1968).
- [38]V. G. S. Lunde, Data for LIQUID-H (2025), Data-verseNO, doi: 10.18710/PT4R3F.DOI














