WorksPhysics and Astronomy
Anisotropic Magnetic and Transport Properties of RAlGe (R = Y, Gd-Tm, Lu)
Abstract
We grew single crystals of the RAlGe family for R = Y, Gd-Tm, Lu and characterized them with powder X-ray diffraction, temperature dependent specific heat measurements as well as temperature and field dependent resistance and magnetization measurements. These RAlGe materials crystallize in an orthorhombic Cmcm crystal structure in which the R atoms occupy a position with 2 point symmetry. We find that when R is a moment bearing rare earth atom, the RAlGe materials order antifferomagnetically at ranging from 5 K (TmAlGe) to 39 K (TbAlGe). A second, lower temperature antiferromagnetic transition is also detected in the range 6-8 K for R = Tb-Ho, but no lower transition is observed (for T above 1.8 K) for R = Er, Tm. The R = Tb-Tm members show evidence for strong crystal field effects influencing the physical properties, where the crystal field favors an easy -axis for R = Tb-Ho and an easy -axis for R = Er, Tm. Of these compounds, strongly uniaxial behavior is observed for DyAlGe and TmAlGe. At T = 2 K, rich metamagnetism is observed when the external field is applied along the -axis for R = Tb–Ho, whereas a single metamagnetic transition is observed (up to 70 kOe) with the field along the -axis when R = Er and Tm. All RAlGe materials show metallic transport behavior and have moderate positive magnetoresistance up to 60-150 % at 2 K superimposed on top of metamagnetic features. The field dependent magnetization measurements on non-moment bearing YAlGe show clear de Haas-van Alphen oscillations, indicating the Fermi surface includes small, high mobility pockets.
Introduction
The AlGe ( = rare earth element) materials exhibit an attractive combination of unique magnetotransport properties and potentially flexible topological states. The light rare-earth ( = La–Gd) members of the series crystallize in a noncentrosymmetric, tetragonal structure, and following early predictions that suggested the broken inversion symmetry may enforce type-II Weyl nodes [1, 2], were found to exhibit a wealth of interesting transport and magnetic phenomena [3] [4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18]. Highlights include the potential observation of a topological Hall effect in CeAlGe [7], and high anomalous Hall and Nernst effects that are in agreement with values predicted from the intrinsic Berry curvature in PrAlGe and NdAlGe [10, 12–15]. CeAlGe also displays complex angular magnetoresistance [8], and when alloyed with Si, CeAlGeSi, the magnetoresistance undergoes enormous spikes exceeding 1000% when the applied field is aligned within 1° of the - or -axis [9]. This unusual behavior was suggested to arise from an interaction between magnetic domain walls and the small Fermi surface associated with the Weyl nodes.
The rare earth ( = Y, Gd–Tm, Lu) based AlGe compounds crystallize in a different, orthorhombic, crystal structure with space group [19], except for GdAlGe, which can adopt either a high temperature tetragonal or low temperature polymorph. Figure 1 shows that the crystal structure of the orthorhombic variant consists of corrugated Al-Ge layers and alternating triangles of atoms that fill the space between the Al-Ge layers. The Al-Ge layers are composed of a sheet of Al atoms that span the plane and which are linked above and below, in an alternating manner, by Ge atoms that form chains extending along the -axis. The atoms occupy a position with relatively low point symmetry in the space between the layers, sitting at the vertices of alternating isosceles triangles related to each other by inversion symmetry.

Figure 1. Crystal Structure of orthorhombic, , RAlGe, where = Y, Gd–Tm, Lu. The solid black lines denote the unit cell.
Compared to the light rare earth based RAlGe compounds, the heavy RAlGe materials remain relatively understudied. Existing data indicate antiferromagnetic ordering occurs between –40 K for = Gd–Er, [20–24] but to the best of our knowledge, no data concerning the physical properties of TmAlGe have been published. Very recent work on GdAlGe suggests the presence of strongly dispersive, Dirac-like bands that support a substantial Fermi velocity on the order of m/s, much greater than typically observed in magnetic topological semimetals. [25]
Here, we grow single crystals of AlGe for = Y, Gd–Tm, Lu and provide characterization of their magnetic behavior, electronic transport properties, and specific heat. We find all samples (besides YAlGe and LuAlGe) order antiferromagnetically, with between 5 K (for TmAlGe) up to 39 K (for TbAlGe). When = Gd, Tb, Dy, and Ho, additional lower temperature antiferromagnetic transitions are detected under 10 K, whereas = Er, Tm members do not show signatures of a second transition down to 1.8 K. For = Tb–Ho, field dependent magnetization measurements at 2 K show multiple metamagnetic transitions when the field is applied along the -axis. A single, K, metamagnetic transition is detected when the field is along the -axis for = Er and Tm. All samples have metallic resistivity and show clear anomalies associated with the magnetic transitions. The heavy RAlGe compounds generally have moderate positive magnetoresistance on the order of 60–150% depending on the atom, with peaks or anomalies corresponding their metamagnetic transitions. Density functional theory calculations suggest the electronic structure of nonmagnetic RAlGe may support Dirac nodal lines, and presence of quantum oscilation in the magnetization of YAlGe suggests the Fermi surface contains small, low-mass bands. This work therefore provides an overview of the basic physical properties of the RAlGe family for heavy rare earths ( = Gd-Lu, except Yb) and serves as a basis for more detailed investigation of each member.
Experimental details
Crystal growth
Single crystals of RAlGe were grown from a high temperature solution [26] based on excess aluminum. Starting with elemental rare earths (all Ames Lab, 99.99 %), Al chunks (Alfa-Aesar 99.99 %), and Ge pieces (Alfa-Aesar 99.99 %), the elements were weighed in a nominal molar ratio of and loaded into a 3 piece alumina Canfield crucible set. [27, 28] The crucibles were sealed under vacuum inside of fused silica tubes, loaded into a box furnace, and heated to 1150°C. After dwelling at 1150°C for 1 h, the furnace was cooled to 700°C over 40 h, upon which the tubes were rapidly removed from the furnace, inverted, and the excess flux decanted in a centrifuge equiped with metal cups. Opening the crucibles revealed shiny, well faceted, three dimensional crystals with lengths of –6 mm. Thus far, our attempts to produce a hypothetical YbAlGe using the same or similar procedure have not been successful.
X-ray diffraction
To determine the phase of the crystals, samples were analyzed with powder X-ray diffraction (PXRD). Several crystals from each batch were ground to a powder, sifted through a 35 micron mesh sieve, and diffraction patterns were collected at room temperature on a Rigaku Miniflex-II instrument operating with Cu-K radiation ( Å) at 30 kV and 15 mA. To determine the lattice parameters, the powder data was refined by the Rietveld method using GSAS-II software [29].
To test for possible site disorder, [22] additional single crystal X-ray diffraction was performed on a HoAlGe crystal using a Rigaku XtaLab Synergy-S diffractometer with Ag radiation (0.56087 Å) operating at 65 kV and 0.67 mA. The samples were held in a nylon loop with vacuum grease, and the data was collected at room temperature. The total number of runs and images was based on the strategy calculation from the program CrysAlisPro (Rigaku OD, 2023). The data integration and reduction were also performed using CrysAlisPro, and a numerical absorption correction was applied based on Gaussian integration over a face-indexed crystal. The structures were solved by intrinsic phasing using SHELXT software and were refined with SHELXL. The crystallographic data and refinement results are given in Tables B.4-B.2 in the appendix, and the atomic positions and thermal displacement parameters are also listed in the appendix in Table B.3.
Prior to measuring the physical properties, the crystals were oriented using a Multiwire Laue Camera operating with X-rays from a W source ( Å) at 10 mV and 30 mA with cm between the sample and detector.
Physical property measurements
The magnetization measurements were performed in a Quantum Design Magnetic Property Measurement System SQUID magnetometer. The samples were mounted on a kelF disk, and a blank background using the bare disk was first measured and the values subtracted. Unless noted, all temperature dependent measurements were field cooled, and the field dependent isotherms collected on increasing .
The temperature and magnetic field dependent electrical resistivity were measured in a Quantum Design Physical Property Measurement System. The samples were prepared by cutting the crystals in suitable bars. The contacts were 25 m thick annealed Pt wire that were either spot welded onto or attached with silver epoxy (EPO-TEK H20E) to the RAlGe samples in standard four point geometry. The spot welded contacts were covered with a small portion of silver epoxy to ensure mechanical stability. Typical contact resistances were on the order of 1 .
Specific heat was measured in a Quantum Design Physical Property Measurement System. Prior to each measurement, an addenda was measured using the blank PPMS puck with a small amount of N-grease. For the moment bearing RAlGe samples, the magnetic specific () heat was estimated by subtracting the data for LuAlGe, which served as a reference for the electronic and phonon contributions. The magnetic entropy was determined by integrating . For the integration, the specific heat data below our base temperature (1.8-1.9 K) were extrapolated with a spline fit (using Origin 2026 software) to at K.
Electronic band structure calculations
Band structure calculations in density functional theory [30, 31] (DFT) were performed with the PBE [32] exchange-correlation functional using a plane-wave basis set and projector augmented wave method, [33] as implemented in the Vienna Ab-initio Simulation Package [34, 35] (VASP). We use a kinetic energy cutoff of 240.3 eV, a -centered Monkhorst-Pack [36] k-mesh, and a Gaussian smearing of 0.05 eV. For the accurate calculation of density of states (DOS), we use a dense k-mesh with the tetrahedron method. [37] To calculate Wilson loop with Wannier charge centers, maximally localized Wannier functions (MLWF) [38, 39] and the tight-binding model were constructed with Y , Al and Ge orbitals to reproduce closely the band structure around the Fermi energy () within eV.
Results and discussion
Structural characterization
Figure 2a shows powder X-ray diffraction patterns obtained from the RAlGe samples and Figure 2b shows a representative Rietveld refinement for GdAlGe (analogous refinements were performed for each sample’s pattern). The observed reflections are in excellent agreement with the pattern anticipated for the YAlGe type structure (illustrated in Figure 1), and the refinements give consistently reasonable quality fits to the experimental data, with –13 and GOF = 1.8-3.3. We note that several patterns (in particular that for YAlGe) show very weak peaks (with intensity less than 100 counts) that are most likely from Y-Al or Y-Ge binary phases that form from excess flux which was not fully decanted off the crystals. Given such week intensities attributable to secondary phases, the PXRD patterns support successful growth of (essentially) single phase RAlGe crystals. The refined lattice parameters for each material are given in Figure 2c-Figure 2f as a function of the atomic number of the respective rare earth atoms. For Gd–Tm, Lu, all three lattice parameters decrease monotonically as the rare earth element moves from Gd to Lu, in line with the expected Lanthanide contraction. The lattice parameters for YAlGe are close to, but smaller than those found for TbAlGe with the -lattice parameter being the closest among the -lattice parameter be further below, but still closer to that found for TbAlGe than for DyAlGe.

Figure 2. (a) Powder X-ray diffraction patterns for the RAlGe samples. The patterns are normalized to the highest intensity (1 3 0) peak and offset for clarity. (b) Rietveld refinement of GdAlGe using the Cmcm structural model. (c)-(e) Refined lattice parameters and (f) unit cell volume of RAlGe as a function atomic number . Note the x-axis break between Y and the 4f elements (Gd-Lu). For each plot (c-f), the YAlGe data point falls between those for TbAlGe and DyAlGe.
Whereas the powder diffraction data discussed above support the YAlGe type structure for Y, Gd–Tm, Lu, a recent publication on DyAlGe and HoAlGe suggested that these two members of the heavy RAlGe series adopt a disordered variant of this structure, with the same symmetry but in which the , Al, Ge all have mixed occupancy on each crystallographic site. [22] Because the disordered structural model was based on fits to powder diffraction patterns, we investigated the possibility of a disordered structure by collecting and analyzing single crystal X-ray diffraction data on our HoAlGe crystals. The results of our structure solutions and refinements are presented in Tables B.4-B.3 in the appendix. In short, we find no strong evidence for a disordered structure, as the fully ordered arrangement, with atomic coordinates given in Table B.2 (and illustrated in Figure 1) gives an excellent fit to the diffraction data, and attempts to introduce mixed occupancy onto the respective positions yielded refinements in which the occupancies of the expected atoms changed by less than 1 %. Whereas we only measured single crystal diffraction on HoAlGe, the temperature dependent resistivity datasets, discussed in the following section, have residual resistance ratios, RRR = , in the range 10–20 for each member of the series, and we furthermore observe de Haas-van Alphen oscillations in the magnetization isotherms for YAlGe. Both the high RRR values and observed quantum oscillations point to weak carrier scattering, inconsistent with substantial site disorder, and coupled with the single-crystal X-ray diffraction data for HoAlGe, we conclude that there is no (significant) site-occupancy disorder in the heavy the RAlGe family.
Physical properties
YAlGe and LuAlGe
We begin our discussion of the physical properties of the RAlGe family with YAlGe and LuAlGe, which contain non-moment-bearing Y or Lu atoms. Figure 3a and 3e show the temperature dependent magnetic susceptibility, here defined as , of YAlGe and LuAlGe measured at 10 kOe. Both compounds have small , on the order of –, that is nearly temperature independent, but clearly anisotropic, above K. For YAlGe, the magnetization is negative when the field is aligned along the - and -directions and positive along the -axis. This indicates the Landau and/or core diamagnetic contributions to the susceptibility are dominant in the former orientations and the Pauli paramagnetic contribution is dominant in the latter. In LuAlGe, the Pauli paramagnetism dominates the susceptibility in each sample orientation. In both YAlGe and LuAlGe, the susceptibility increases rapidly at low temperatures beginning below K, which can likely be attributed to the presence of magnetic impurities in the samples. Figure 3b and Figure 3f show the temperature dependent electrical resistivity for each compound. Both YAlGe and LuAlGe behave as normal metals, with resistivity that decreases approximately linearly on cooling before saturating at the lowest temperatures. The residual resistance ratios () are 16.7 and 12.9 for YAlGe and LuAlGe respectively, attesting to good crystal quality and low levels of disorder. The resistance data has no evidence for transitions, consistent with the magnetic data and with the expectations for non-moment bearing Y and Lu. The magnetoresistances are presented in Figure 3c and Figure 3g. Both compounds have similar behavior, in which increases with a sub-linear field dependence to modest, unsaturated, values of 35 % and 24 % at 55 kOe for YAlGe and LuAlGe respectively. Magnetization isotherms measured at base temperatures of 1.8 K and 2 K are presented in Figure 3d and Figure 3h. de Haas–van Alphen oscillations are clearly present in the data for YAlGe, and their amplitude increases at higher fields, becoming most apparent above 40 kOe. No quantum oscillations are observed for LuAlGe, and instead the magnetization is weakly field dependent above 10 kOe. At lower fields, the magnetization isotherms for LuAlGe show a rapid increase, which, considering the minute () values of , likely reflects the polarization of magnetic impurities extrinsic to the bulk sample.

Figure 3. (a) Temperature dependent magnetic susceptibility (), (b) temperature dependent resistivity, (c) field-dependent resistivity (MR), and (d) Field-dependent magnetization for YAlGe. (e-h) show the same data for LuAlGe.
The presence of quantum oscillation in the magnetization isotherms for YAlGe suggests the Fermi surface contains small/low mass and high mobility pockets. To quantify this further, Figure 4a and Figure 4b plot the oscillatory part of the magnetization, as a function of inverse magnetic field, where is the difference between the total measured magnetization and the non-oscillatory component, the latter of which was estimated using a third order polynomial fit to the isotherms. We note that to achieve appropriate data density, the data in Figure 4 was obtained from separate measurements than

Figure 4. Analysis of the De Haas–van Alphen oscillations for YAlGe (a) and (b) show the oscillatory component of the magnetization plotted against inverse field for and orientations, respectively. (c) and (d) show the fast-Fourier transforms of the associated data in (a) and (b).
that shown in Figure 3d and were only measured between 40-55 kOe where the oscillations are strong. Figure 4c and Figure 4d show Fast-Fourier transforms (FFTs) of the quantum oscillations in 4a and 4b respectively. In both orientations, the FFT reveals one strong frequency, at 354 T for and 272 T when . From the frequencies, we calculated the associated Fermi-surface cross sections using the Lifshitz-Onsager relation:
where are the quantum oscillation frequencies corresponding to each pocket, is Planck’s constant, and is the electron charge. The 354 T and 272 T components for and are both consistent with small, low mass pockets, and represent cross sections of and , accounting for 1.9 % and 1.4 % of the respective – and – planes.
GdAlGe
The temperature dependent magnetization for GdAlGe is shown in Figure 5a. The susceptibility is effectively isotropic at high temperatures, as expected for Gd with orbital angular momentum . decreases on cooling and reaches a peak near 35 K, indicating the onset of antiferromagnetic order. The resistivity, shown in Figure 5b, is consistent with the magnetic data, with a clear loss of spin disorder scattering observed below K. Close inspection of the low-temperature data (insets in Figure 5a and Figure 5b) indicate there are two relatively closely spaced transitions, and based on the extrema in the respective derivatives and shown in Figure 5c and Figure 5d [40, 41], we assign transition temperatures of K and K, with good agreement between magnetic and transport data. The data presented here show good agreement with the previous publications on GdAlGe [25], however, the lower temperature transition was not observed in the polycrystalline samples.

Figure 5. Physical properties of GdAlGe. (a) Temperature dependent magnetic susceptibility (left axis). The right axis shows the average of the individual inverse curves for each orientation (see text). The solid line shows the Curie-Weiss fit to the data (open points). (b) Temperature dependent resistivity. The inset in (b) gives a close-up view of the data between 20–40 K. (c) Derivative of the temperature-susceptibility product (), (d) derivative of the resistivity (), (e) field dependent magnetization at K, (f) magnetoresistance at K, (g) total and magnetic contributions to the heat capacity. The data for LuAlGe is also shown for reference. (h) Estimated magnetic entropy.
The right axis of Figure 5a shows the inverse of the average of the three curves, where . Above the Néel temperature, the inverse susceptibility is linear with temperature, consistent with thermally disordered local moments, and we therefore fit the to the Curie-Weiss law over 100-300 K, where
in which is the Curie constant, the temperature, and the Weiss temperature. From the fit, we estimate K, suggesting antiferromagnetic correlations, and an effective moment , very close to the value of 7.94 expected for Gd.
The field dependent magnetization measured at 2 K is given in Figure 5e, and the results are also consistent with an antiferromagnetic ground state. For each orientation, increases linearly over the measured range of 0–55 kOe. The magnetization with the field oriented along and axes are nearly identical, and the field dependence has a smaller slope with , which is consistent with the smaller susceptibility, in the ordered state, for this orientation implied by Figure 5a. At 55 kOe, the magnetization only reaches /Gd, far from the expected saturated moment of for Gd. In their recent work, Laha et al. observe a metamagnetic transition at 62 kOe; however, the step in the magnetization is small () and remains unsaturated up to at least 70 kOe. [25] The magnetoresistance of GdAlGe is shown in Figure 5f, and increases in an approximately linear fashion to a maximum of 120 % at 55 kOe. Like the 2 K isotherm, there is no indication of saturation up to the highest measured field.
Finally, Figure 5g shows the specific heat of GdAlGe measured between 1.8–50 K. Both magnetic transitions are apparent as well-resolved peaks at K and K, consistent with the transition temperatures determined by magnetic and transport data. To estimate the magnetic contribution of the specific heat (), we measured and subtracted the specific heat of LuAlGe, which serves as a reference for the electronic and phonon contributions to for the heavy RAlGe series. The data for LuAlGe is shown as a gray curve in Figure 5g and the estimated of GdAlGe is the green dataset. We integrated to estimate the magnetic entropy associated with the transitions. The results are shown in Figure 5h and show that approximately 90 % of the full is recovered by 50 K. This entropy deficiency strongly suggests that LuAlGe is only an approximate of the non-magnetic contributions to the specific heat of GdAlGe. It is also possible that some magnetic correlations persist above . As such, we can only extract qualitative statements about entropy associated with other members of the RAlGe series, likely with increasing accuracy as the moment bearing R gets closer to Lu.
TbAlGe
The data collected on TbAlGe single crystals are presented next in Figure 6. The magnetization is anisotropic in the paramagnetic state, with the strongest response observed when . The inverse is linear above K. Curie-Weiss fits over this temperature range give K, suggesting antiferromagnetic correlations, and an effective moment of /Tb, agreeing well with the expected for Tb.

Figure 6. Physical properties of TbAlGe. (a) Temperature dependent magnetic susceptibility (left axis). The right axis shows the polycrystalline average of the individual inverse curves for each orientation. The solid line shows the Curie-Weiss fit to the data (open points). (b) Temperature dependent resistivity. The inset gives a close-up view of the data between 1.8–50 K. (c) Derivative of the temperature-susceptibility product , (d) derivative of the resistivity , (e) field dependent magnetization at K, (f) magnetoresistance at K, (g) total and magnetic contributions to the heat capacity. The data for LuAlGe is also shown for reference. (h) Estimated magnetic entropy. The horizontal dashed lines show the entropy expected for different multiplets.
In prior works, TbAlGe was reported to order antiferromagnetically at K and K [21, 23, 24], and our measurements are consistent with the earlier results. Figure 6a and Figure 6b, and the respective derivatives in 6c and 6d show that antiferromagnetic transitions are observed at K and K in our magnetic and transport data. Given the step-like drop of at 7 K when , the lower transition may be first-order. The data and isotherms (discussed next) suggest that Tb spins are primarily aligned along the -axis below .
Figure 6e shows the field dependent magnetization of TbAlGe measured at 2 K. When , undergoes two step-like jumps, at kOe and 35 kOe, reaching /Tb at 70 kOe. Both transitions are hysteretic, suggesting a first order nature. At the highest fields (–70 kOe), appears to slightly increase, hinting that a third metamagnetic transition may occur just above 70 kOe. The possibility of another transition at higher field is consistent with the relatively low values of compared to the anticipated saturated moment of 9 for Tb ions; however, higher field measurements are needed to confirm this. The magnetization for and both show a linear field dependence with no indication of metamagnetism up to 70 kOe. Our data is in reasonable agreement with recent work on TbAlGe single crystals reported by Gouda et al. [23].
The magnetoresistance at 2 K is given in Figure 6f. With the field applied along , the resistance shows a peak at 19 kOe and a broader increase beginning at kOe. Both features correspond reasonably well with the two metamagnetic transitions observed in the magnetization isotherms. Whith the field applied along and , the magnetoresistance increases monotonically to modest values near 70 % at 55 kOe and shows no signatures of a transition, in agreement with the magnetic data for these orientations.
Both AFM transitions show strong peaks in the specific heat data (Figure 6g), and the transition temperatures K and K are in good agreement with the magnetic and transport data. The sharp, spike-like feature observed at may suggest the lower temperature transition is first-order, and this interpretion is consistent with the steplike drop observed in the magnetic data in Figure 6a. Figure 6h shows that by 70 K, the magnetic entropy recovers of the expected for Tb local moments. Considering the clear magnetic anisotropy visible in the data up to 300 K, this is consistent with some degree of CEF splitting above the magnetic ordering temperatures.
DyAlGe
DyAlGe has qualitatively similar properties to TbAlGe but with stronger -axis anisotropy (at low temperatures) and transitions that are shifted to lower temperatures. In agreement with the data on polycrystalline samples, [22] our magnetic susceptibility and resistivity measurements indicate DyAlGe undergoes two transitions, at K and K, as shown in Figure 7a–Figure 7d. has a prominent peak at (28 K) when . Broad humps with maxima at K are observed when the field is oriented along and , likely attributable to low lying crystal field levels. This behavior shows DyAlGe to have strongly uniaxial behavior in which the moments are aligned along the -axis below . The susceptibility curve shows a second, very sharp transition at K that is only weakly resolved with the field along the other principle axes but is clearly observed in the resistivity (see inset to 7b). Curie-Weiss fits to the high temperature data give /Dy, in very good agreement with the expected 10.6 for Dy local moments. The paramagnetic temperature is -23.4(5) K, consistent with the observed antiferromagnetic order.

Figure 7. Physical properties of DyAlGe. (a) Temperature dependent magnetic susceptibility (left axis). The right axis shows the polycrystalline average of the individual inverse curves for each orientation. The solid line shows the Curie-Weiss fit to the data (open points). (b) Temperature dependent resistivity. The inset in (b) gives a close-up of the data between 1.8–35 K with arrows marking the transitions. (c) Derivative of the temperature-susceptibility product , (d) derivative of the resistivity , (e) field dependent magnetization at K, (f) magnetoresistance at K, where the inset shows a zoomed-in view near the 60 kOe transition, (g) total and magnetic contributions to the heat capacity. The data for LuAlGe is also shown for reference. (h) Estimated magnetic entropy. The horizontal dashed lines show the entropy expected for different multiplets.
The field dependent magnetization of DyAlGe at 2 K is shown in Figure 7e. Like TbAlGe, rich metamagnetism is observed with the applied field along the -axis. The data provides evidence for four metamagnetic transitions occurring below 70 kOe, that occur at approximately 9 kOe, 16 kOe, 38 kOe, and 53 kOe when increasing the field. The isotherm shows clear hysteresis between 14–40 kOe and for the high field transition at 52–57 kOe. For fields aligned along the and axes, the magnetization increases linearly up to 70 kOe, with a greater slope when . Given that only reaches /Dy and has yet to saturate at 70 kOe, additional metamagnetic transitions are likely to be found at higher fields. As such, DyAlGe may be like TbNiGe, [42] providing another example of a strongly uniaxial moment with a devil’s staircase-like .
The magnetoresistance is shown in Figure 7f. When the field is applied along the -axis, metamagnetic transitions occur at kOe and 60 kOe. These transitions approximately align with the first and last metamagnetic transitions observed in the magnetic isotherms, and the slightly higher onset fields likely come from imperfect sample orientation. Whereas the intermediate-field metamagnetism detected in Figure 7e is not observed in the transport data, the magnetoresistance shows a broad hysteretic region over approximately the same fields, between 10–45 kOe. The magnetoresistance increases monotonically with field when , consistent with the magnetic data, and reaches modest maximum values of 120 % at 90 kOe.
Figure 7g shows that both transitions are observed in the specific heat, and the transition temperatures agree well with the magnetic and transport results. While in part accounted for by an imperfect non-magnetic background subtraction (LuAlGe), there is substantial magnetic entropy remaining above , and the peak associated with magnetic order at 28.5 K appears superimposed on a relatively broad Schottky-like maxima in . This is consistent with and magnetization data, and is associated with thermal population of the crystal
field levels. Given that an analogous feature is not observed in TbAlGe, which also shows strongly uniaxial behavior in the paramagnetic state (see Figure 13 in the Appendix) but much weaker anisotropy at 2 K, it is likely that the ground state CEF multiplet is more energetically isolated from the next lowest energy states in DyAlGe than in TbAlGe. Inelastic neutron data would be needed to confirm this picture. By 70 K, the integrated magnetic entropy reaches nearly 80 % the full Rln16 expected for trivalent Dy moments.
Figure 13. Inverse plotted against temperature for RAlGe for –Tm. The solid lines are Curie-Weiss fits to the respective data. Results for GdAlGe are not shown since Gd has , leading to negligible CEF effects and isotropic paramagnetic behavior. Note that in ErAlGe, the high temperature behavior of deviates from linearity above K for the data measured in the orientation. We believe this is an experimental artifact, which gives clearly erroneous values of and ; therefore, we performed Curie-Weiss fits between 100–200 K, where the data is linear with similar slope to the and data.
HoAlGe
Figure 8 shows the data collected on HoAlGe single crystals. The results are again qualitatively similar to those of TbAlGe and DyAlGe, with pushed to lower temperature as the R atom changes from Tb to Ho. The magnetic susceptibility (Figure 8a) shows two antiferromagnetic transitions, at K and K based on (Figure 8c). The resistivity, shown in Figure 8b, shows good metallic behavior, a RRR of 11.5, and two transitions in excellent agreement with the magnetic results (see in Figure 8d). These results are again roughly consistent with 11 K and 8 K antiferromagnetic transition temperatures previously found in polycrystalline HoAlGe. [22] Fitting the high temperature susceptibility to the Curie-Weiss law gives K, consistent with antiferromagnetic interactions. The estimated effective moment is 10.3(2) /Ho, matching the expected value of 10.3 for trivalent Ho local moments.
Figure 8. Physical properties of HoAlGe. (a) Temperature dependent magnetic susceptibility (left axis). The right axis shows the polycrystalline average of the individual inverse curves for each orientation. The solid line shows the Curie-Weiss fit to the data (open points). (b) Temperature dependent resistivity. The insets in (a) and (b) give zoomed view of the data between 1.8–20 K. (c) Derivative of the temperature-susceptibility product , (d) derivative of the resistivity , (e) field dependent magnetization at K with insets showing a close-up view near the metamagnetic transitions, (f) magnetoresistance at K, (g) total and magnetic contributions to the heat capacity. The data for LuAlGe is also shown for reference. (h) Estimated magnetic entropy. The horizontal dashed lines show the entropy expected for different multiplets.
The field dependent magnetization at 2 K is shown in Figure 8e. When , Figure 8e shows three metamagnetic transitions at approximately 14 kOe, 25 kOe, and 32.5 kOe. Here the critical fields were determined by maxima in , for the increasing field-sweep. The upper inset in Figure 8e shows clear hysteresis between ≈ 12–25 kOe, whereas the higher field transition at 30 kOe is not hysteretic. Above the transitions, the magnetization saturates at ≈ 8.5 at 70 kOe. When , the magnetization first increases linearly at low fields and then undergoes two metamagnetic transitions at approximately 38 kOe and 48 kOe. Both transitions display measurable hysteresis upon raising and lowering the field, and above the second transition, the magnetization continues to increase with field, reaching nearly 8 at 70 kOe. As this is smaller than the expected 10 expected for Ho, we anticipate additional metamagnetism at yet higher fields. Finally, when , the magnetization increases linearly to ≈ 4 at 40 kOe, where the curve begins to saturate towards 5 as approaches 70 kOe. The presence of metamagnetism in multiple field orientations is consistent with the temperature dependent behavior, indicating that while still anisotropic, the magnetism of HoAlGe is substantially less anisotropic than in DyAlGe. The magnetoresistance is shown in Figure 8f and undergoes a step-like jump at 13 kOe when . Above the metamagnetic transition, the resistance increases nearly linearly with field until 36 kOe, where it drops rapidly, finally reaching a minimum at 58 kOe, above which it increases with field and is not saturated by 90 kOe. Like with DyAlGe, the sharp changes in behavior shown in the magnetoresistance for HoAlGe correspond reasonably well with the metamagnetism found in the isotherms.
Figure 8g shows the heat capacity of HoAlGe. Both transitions are evident at 7.7 K and 10.9 K, in good agreement with magnetic and transport results. The lower, 7.7 K, transition produces a very narrow spike in , indicating a first order nature. This is consistent with the sharp, step-like, features observed in the magnetic and resistivity datasets at (see insets to Figure 8a and Figure 8b). The magnetic entropy, shown in Figure 8h, recovers nearly the full expected for free Ho ions by 40 K, suggesting HoAlGe has smaller CEF splitting compared to and Dy members which is consistent with the weaker magnetic anisotropy observed for HoAlGe.
ErAlGe
The magnetic behavior of the RAlGe series substantially changes when crossing from to Er. The temperature dependent magnetization and resistivity, and their respective derivatives shown in Figure 9a-Figure 9d indicate that the Er moments order antiferromagnetically at K, with no additional transitions observed at lower temperature. Figure 9a and its inset show that whereas reaches a maximum at the Néel temperature when , the transition is only weakly observed when the field is applied along - or -axes. Unlike samples with -Ho, which likely have local moments oriented primarily along the -axis, for , the behavior above and below suggests the local moments have an easy -axis. It is very likely that the change in anisotropy observed here in the orthorhombic RAlGe family reflects a change in sign of the leading terms in the crystal field Hamiltonian, which are normally expected to switch between between and in an isostructural series owing to the change in sign of the second-order Steven’s factor [43]. This is discussed in more detail in the discussion section.

Figure 9. Physical properties of ErAlGe. (a) Temperature dependent magnetic susceptibility (left axis). The right axis shows the average of the individual inverse curves for each orientation. The solid line shows the Curie-Weiss fit to the data (open points). (b) Temperature dependent resistivity. The insets in (a) and (b) give a zoomed view of the data at low temperature. (c) Derivative of the temperature-susceptibility product (), (d) derivative of the resistivity (), (e) field dependent magnetization at K, (f) magnetoresistance at K, (g) total and magnetic contributions to the heat capacity. The data for LuAlGe is also shown for reference. (h) Estimated magnetic entropy. The horizontal dashed lines show the entropy expected for different multiplets.
The right axis of Figure 9a shows the inverse of of ErAlGe. The Curie-Weiss fits suggest an average effective moment of , in excellent agreement with the expected value of for local moments. The average Weiss temperature is K, suggesting antiferromagnetic interactions above , which is also in agreement with the experimental observations.
Figure 9e shows the field dependent magnetization at 2 K. is isotropic until kOe, upon which rapidly jumps from to between 20–30 kOe when . Above 20 kOe, when , appears to be saturating towards , and when , continues to gradually increase, approaching at 55 kOe, which is lower than the expected for local moments. The isotherm shows no hysteresis when measuring with increasing and decreasing fields, differing from the metamagnetism observed for Tb–Ho based samples. The magnetoresitance, shown in Figure 9f, has a prominent peak at kOe, in agreement with the metamagnetic transition, and otherwise increases with field up to % at 90 kOe.
The heat capacity of ErAlGe is shown in Figure 9g. A lambda-like peak is observed at 6.8 K, in reasonable agreement with the Néel temperature inferred from the magnetic and transport data. Below the transition, the heat capacity does not approach zero and is nearly constant on cooling at the lowest temperatures. This behavior indicates there is remaining entropy at our base temperature of 1.9 K and suggests that another magnetic transition would likely occur with further cooling. The magnetic entropy, shown in Figure 9h recovers nearly at the transition. Given a formal CEF quartet is impossible in orthorhombic symmetry, this may indicate the presence of a low-energy pseudoquartet, composed of two closely spaced doublets. By 70 K, the magnetic entropy recovers approximately 80 % the expected for trivalent Er moments. Considering that the low temperature specific heat is not trending towards the origin, the spline extrapolation to at that we used to estimate the specific heat below our lowest temperature data point (see experimental details section) is unreliable in ErAlGe. Because the true temperature dependence of was not measured below 1.9 K, the integration shown in Figure 9h likely results in an underestimate of the true magnetic entropy.
TmAlGe
The physical properties of TmAlGe are similar to those of ErAlGe but with much stronger -axis anisotropy. Like ErAlGe, the magnetization (shown in Figure 10a) shows a prominent peak at K when . The temperature dependence is considerably weaker for samples oriented with the field applied along and . In both orientations, a small local maxima is observed at 5 K followed by another, relatively broad, maxima between -70 K (shown in inset to Figure 10a). The latter is likely attributable to relatively low-lying crystal field multiplets. Fitting to the Curie-Weiss law gives and , in good agreement with the expectation of .

Figure 10. Physical properties of TmAlGe. (a) Temperature dependent magnetic susceptibility (left axis). The right axis shows the average of the individual inverse curves for each orientation. The solid line shows the Curie-Weiss fit to the data (open points). (b) Temperature dependent resistivity. (c) Derivative of the temperature-susceptibility product (). The insets in (a)–(c) give zoomed views of the data at low temperatures. (d) Derivative of the resistivity (), (e) field dependent magnetization at K, (f) magnetoresistance at K, (g) total and magnetic contributions to the heat capacity. The data for LuAlGe is also shown for reference. (h) Estimated magnetic entropy. The horizontal dashed lines show the entropy expected for different multiplets.
The resistivity for TmAlGe is shown in Figure 10b. Like the other RAlGe materials, ErAlGe shows metallic behavior and a moderate of 12.8. Loss of spin disorder scattering is evident at K (see in Figure 10d), in good agreement with the single transition found at 5 K in the magnetic data.
The field dependent magnetization for TmAlGe is presented in Figure 10e. When , a clear metamagnetic transition is observed at kOe, where the magnetization quickly jumps from 2 to 7 /Tm. Like ErAlGe, the metamagnetic transition does not have observable hysteresis when raising or lowering the field; however, unlike ErAlGe, TmAlGe shows substantial anisotropy at all fields, whereas the isotherms for ErAlGe are isotropic below its metamagnetic transition. Above 25 kOe, the magnetization for the orientation appears saturated at 6.9 up to 55 kOe, in reasonable agreement with the expected 7 for Tm ions. No metamagnetism is evident in the other two orientations. Instead, increases linearly over the measured field range when , reaching 1 /Tm at 55 kOe. increases weakly with the field when oriented with , only achieving 0.3 /Tm at 55 kOe.
Figure 10f gives the magnetoresistance of TmAlGe when and . Like ErAlGe, the magnetoresistance first increases with field, but falls sharply at 15 kOe, corresponding to the metamagnetic transition observed for in the field dependent magnetization data (Figure 10e). After the peak, the resistance increases nearly linearly up to 60% at 90 kOe.
Finally, Figure 10g shows the heat capacity of TmAlGe. A lambda peak is observed at 5.3 K, consistent with the Néel temperature inferred from magnetic and transport data. Like ErAlGe, the specific heat does not trend towards zero at our lowest temperature, which may point towards additional magnetic transitions at lower temperatures. The magnetic heat capacity furthermore shows a broad Schottky-like anomaly between –70 K, likely reflecting thermal population of the crystal field levels. This is consistent with the behavior of the magnetization when the field is applied along the - and -axis and in-line with the much stronger uniaxial anisotropy observed for TmAlGe than ErAlGe. The magnetic entropy is shown in Figure 10h and recovers nearly 75% Rln2 at the 5.3 K transition, which may suggest a ground state doublet. Since Tm is not a Kramer’s ion, such a doublet is not guaranteed. The integrated entropy recovers over 75% of the full Rln13 by 70 K. As was the case for ErAlGe, the magnetic entropy in Figure 10h is likely underestimated due to the small but apparent remaining entropy at 1.9 K, which is not accounted for by our extrapolation of the data to the origin and therefore neglected in our integration.
Discussion
de Gennes scaling and magnetic anisotropy
The key results from our measurements are outlined in Table 1, which contains the lattice parameters, magnetic transition temperatures, effective (paramagnetic) magnetic moment, Weiss temperatures, and de Gennes factor (()), with the Lande factor and the total angular momentum for the magnetic moment-bearing members of the RAlGe family. At sufficiently high temperatures ( K), the moment bearing RAlGe compounds (–Tm) show paramagnetic Curie-Weiss like behavior with estimated effective moments that agree well the the expected free-ion values. On cooling, the moment bearing members all show long-range antiferromagnetic order. When = Tb–Ho, the samples have two transitions in which the moments are aligned (at least primarily) along the -axis, with strongest anisotropy observed for DyAlGe. The first, higher temperature transition, decreases from 39 K in TbAlGe to 10 K in HoAlGe, whereas the second, lower temperature, transition occurs near ≈7 K in each. Based on the step-like features in the magnetization and sharp peaks observed in the specific heat, the lower transition in Tm-, Dy-, and Ho-based members is likely first-order in nature. The magnetic behavior changes for beyond Ho. Unlike the = Gd–Ho members, ErAlGe and TmAlGe each have only a single magnetic transition above 1.8 K and show -axis anisotropy that is considerably stronger in the case of TmAlGe. We note however that a second, lower temperature transition may also be found in these samples below 1.8 K, and that this notion is roughly supported by the small, but finite, entropy implied by the behavior of the specific heat at lowest temperatures. If the RKKY interactions mediates the coupling between the rare-earth based local moments through the conduction electrons, then in a mean field approximation, both the magnetic ordering and paramagnetic Weiss temperature should be proportional to the de Gennes factor. As shown in Figure 11 and

Figure 11. (a) Néel temperature as a function of the rare-earth ion de Gennes factor (dG) for the RAlGe. (b) Weiss temperature (), obtained from polycrystalline averaged , as a function of the rare-earth ion de Gennes factor for the RAlGe. The dashed lines show the values of and expected for ideal de Gennes scaling.
| RAlGe | (Å) | (Å) | (Å) | (Å) | (K) | (K) | (/R) | (K) | dG |
| GdAlGe | 4.0626(1) | 10.52(2) | 5.803(1) | 248.05(1) | 34.7, 35 [25] | 29.8, 30 [25] | 7.9(1) | -66(1) | 15.75 |
| TbAlGe | 4.0467(1) | 10.45(2) | 5.772(1) | 244.00(1) | 39.3, 39 [21], 38 [23], 40 [24] | 7.6, 7 [21], 7.6 [23], 8 [24] | 9.7(1) | -43.8(5) | 10.5 |
| DyAlGe | 4.0376(1) | 10.40(2) | 5.749(1) | 241.39(1) | 28.5, 30 [22] | 6.5 | 10.7(2) | -23.4(5) | 7.08 |
| HoAlGe | 4.0306(1) | 10.35(2) | 5.729(1) | 239.08(1) | 10.9, 11 [22] | 7.7 | 10.3(2) | -8.9(2) | 4.5 |
| ErAlGe | 4.0222(1) | 10.31(2) | 5.708(1) | 236.64(1) | 6.8, 7 [21] | - | 9.6(3) | -7.6(4) | 2.55 |
| TmAlGe | 4.0136(1) | 10.27(2) | 5.688(1) | 234.53(1) | 5.3 | - | 7.5(2) | -2.0(2) | 1.16 |
Table 1. Summary of key results for the moment-bearing members of the RAlGe family (R = Gd-Tm). Lattice parameters, antiferromagnetic transition temperatures, effective magnetic moments, averaged Weiss temperatures, and de Gennes factor (dG = ). The listed transition temperatures are inferred from the specific heat measurements. We also include references to the magnetic ordering temperatures reported in previous studies of RAlGe compounds.
Table 1, we find relatively poor, non-linear and non-monotonic scaling between and . Such poor de Gennes scaling of is likely associated with the crystal field, which constrains the moments to the -axis for = Tb–Ho members and to the -axis for = Er, Tm.
Strongly uniaxial behavior produces highly anisotropic interactions between magnetic moments, conditions under which de Gennes scaling is expected to be violated [44]. In the RAlGe series, this is supported by two observations. First, the deviation from the expected de Gennes scaling is greatest for = Tb, Dy, and Tm, which differ from the "ideal" values by a factor of 1.6, 1.8, and 2, respectively, whereas the values of HoAlGe and ErAlGe essentially fall on the expected line. If we take the magnetic anisotropy as an indicator of the crystal field splitting, the relative deviation from the values predicted for perfect de Gennes scaling is strongest for the RAlGe members with the greatest anisotropy (see Figure 6-Figure 10 and Figure 13 in the Appendix). Second, unlike the ordering temperatures, Figure 11b shows that the polycrystalline average Weiss temperatures scales well with the de Gennes factor. Whereas the anisotropic Weiss temperatures (, , and ) have contributions from both the molecular field (exchange) and the crystal field, the crystal field effects should be canceled by averaging such that , obtained from fits to , should reflect only the sign and magnitude of the exchange [45]. The substantially improved linearity between dG and support RKKY mediated exchange in the RAlGe series, where the deviations in the de Gennes scaling are the consequence of the crystal field restricting magnetic moments to - or -crystallographic axes for = Tb-Ho and = Er, Tm respectively.
The change from having an easy - to -axis that is observed on crossing from HoAlGe to ErAlGe is also an expected consequence of the crystal field. This can be demonstrated by considering the general crystal field Hamiltonian
where are the crystal field parameters and are the Stevens operators. In the RAlGe materials, the rare-earth atoms occupy a relatively low symmetry environment with point symmetry, producing a full CEF Hamiltonian in which all nine even terms from to are allowed. Despite the complexity, if we assume the leading terms are dominant, then the sign of will dictate whether the system shows an easy -axis or -plane, and a finite will further break the rotational symmetry within the -plane. Importantly, the second-order terms are expected to change sign between and , as can be seen by the general expression for the crystal field parameters [46]
where are related to the ligand environment, is the expectation value of the radial wavefunction, and are the Steven’s multiplicative factors [43, 47]. Because the point-symmetry and ligand environment are preserved across the RAlGe series, the sign of is generically expected to change between HoAlGe and ErAlGe owing to the fact the second-order Stevens parameter, , switches from negative values for – to positive for . [43] Consequentially, the magnetic easy-axis of the RAlGe compounds is anticipated to change between HoAlGe and ErAlGe, as is commonly observed in many rare earth-based series of materials. [42, 48, 49, 50, 51, 52]
As outlined in the Appendix, we attempted to make these statements more quantitative by using a simplified crystal field model, retaining only second-order terms in the CEF Hamiltonian [48, 53, 54], to estimate and from the anisotropic Weiss temperatures. Whereas this approach gives a rather poor estimate of the absolute value of the CEF parameters (i.e. a wide range of possible values depending on how the parameters are estimated), we indeed observe a change in sign of both and from positive values for – to negative values for . Here, the -axis of the crystal field will correspond to the crystallographic -axis of the RAlGe series, which is the primary 2-fold rotation axis in the space group. From this reference frame, the - and - crystal field axes will align with the crystallographic - and -directions. As discussed explicitly in the Appendix, the evolution of the estimated and is consistent with the change from samples having an easy -axis to easy -axis on crossing from HoAlGe to ErAlGe. Because the absolute values of the CEF parameters are poorly determined by our simplified approach, we do not attempt to propose explicit CEF splitting schemes for the RAlGe compounds; however, in each case, diagonalizing the approximate CEF Hamiltonians indicates that the scale of the CEF splitting is largest for DyAlGe and TmAlGe, on the order of 150 K, and smallest ( K) in HoAlGe and ErAlGe, results that are at least qualitatively consistent with the strength of the observed magnetic anisotropy in the RAlGe series. Finally, we note that point charge crystal field calculations, implemented in PyCrystalField [46], give qualitatively similar trends; however, the absolute values of the crystal field parameters and resulting eigenvalues are extremely sensitive to the charges assigned to the ligand atoms and therefor provide little quantitative guidance in determining appropriate crystal field schemes for the RAlGe. Ultimately inelastic neutron scattering data is needed to quantitatively determine and model the crystal field splitting.
Electronic band structure
Given the topological physics found in the tetragonal, light -based, RAlGe compounds, it is worth considering whether the orthorhombic, heavy members presented here may also support topological electronic states. This is further motivated by the observation of quantum oscillations in YAlGe (see Figure 3), which show that the non-magnetic band structure contains small pockets. To explore this question, we used DFT to calculate the electronic band structure of non-magnetic YAlGe. Figure 12a plots the projected density of states (PDOS) on atomic orbitals for YAlGe. While Ge orbitals dominate the lower energy states from -11.0 to -8.0 eV, other orbitals hybridize strongly from -7.5 to +3.0 eV. The Fermi energy sits inside a pseudo-gap region spanning eV with the valence bands being derived from Ge , Y and Al orbitals, while the conduction bands mostly from empty Y orbitals. This pseudo-gap region is also reflected in the band dispersion as plotted in Figure 12b without and 12d with SOC. Between the highest valence and lowest conduction band, there is a sizable energy separation along most of the high-symmetry directions in the BZ, giving a local DOS minimum and pseudo-gap region near the . Without SOC, Figure 12b shows that crossings along the –, –, –, – and – directions form 4-fold degenerate nodal loops. With SOC in 12d, these nodal loops are all gapped out, giving a continuous gap between the highest valence and lowest conduction band, but not a global band gap. This continuous gap between the two band manifolds gives a well-defined topological index. From the Wilson loops plotted in 12c, the odd crossings for the Wannier charge centers on both and planes but not on other planes indicate a weak topological insulator phase is hosted in the system, albeit still with a finite Fermi surface, similar to what is observed in RhBi. [55] The non-magnetic band structures of other RAlGe series are similar.

Figure 12. Electronic band structure of YAlGe. (a) Projected density of states (PDOS) on atomic orbitals. (b) Band structure without spin-orbit coupling (SOC). The inset shows the bulk Brillouin zone (BZ) and high-symmetry -points. (c) Wilson loop with Wannier charge centers (WCC) on and planes for (d) band structure with SOC. In (b) and (d), the highest valence (lowest conduction) band is colored in blue (red). (e)–(g) show the DFT calculated Fermi surface (FS) of the three bands crossing the Fermi level. The FS corresponds to the (e) next-valence, (f) highest-valence, and (c) lowest conduction bands, respectively. (h) is the combined FS from the three bands in (e–g).
To supplement the quantum oscillations observed in the magnetization isotherms, we also used DFT to calculate Fermi surfaces (FS). There are three bands crossing , as seen in Figure 12d and their Fermi surfaces are plotted in Figure 12e-Figure 12h, with the combined FS shown in Figure 12h. The next valence band below the highest energy valence band only crosses near the – direction and forms cone-shaped hole pockets as seen in Figure 12e. The valence band FS in Figure 12f is a skeleton spanning along the (or the -axis) direction. The different sizes of the cross sections of the skeleton give cyclotron extremal orbits for quantum oscillations when . Our calculation shows three main frequencies at 30.8, 172.0 and 393.1 T, with the highest frequency agreeing reasonably well with the experimentally measured 354 T pocket in Figure 4c. The conduction band’s FS, shown in Figure 12g, occupies the two ends of the Brillouin zone opposite to the valence band FS, with no overlap between them, confirming a continuous gap. When , there are two main cyclotron extremal orbits. One is the cone-shaped FS from the next valence band and the second is from the valence band, which form two concentric extremal orbits near the – direction (see Figure 12h) with the frequencies of 229.0 and 321.1 T, respectively. Given that the average (275 T) of these two frequencies matches very closely the 272 T experimental value with an asymmetric shape as seen in Figure 12d, it is possible this frequency corresponds to scattering between the two 229 T and 321 T calculated pockets. These results and calculations motivate more detailed study, reserved for future work, of the of the quantum oscillations and Hall effect to further study electronic band structure, especially in the magnetically ordered states of moment-bearing members of the RAlGe series.
Summary and conclusions
We present a detailed study of the magnetization, electronic transport, and specific heat of the RAlGe family, where Y, Gd–Lu (except Yb). Whereas the light rare earth RAlGe variants ( La–Gd) crystallize in a tetragonal structure, the heavy RAlGe materials adopt an orthorhombic crystal structure. Here, we find that the moment-bearing heavy RAlGe materials order antiferromagnetically at ranging from 5 K (TmAlGe) to 39 K (TbAlGe). We also find a second, low temperature antiferromagnetic transition occurs at 6–8 K when Gd–Ho. Magnetization isotherms measured at 2 K reveal multiple metamagnetic transitions (under 70 kOe) for for Tb–Ho, whereas a single metamagnetic transition is observed with the field along the -axis when Er and Tm. While not explored here, the 2 K isotherms suggest the Tb–Ho members will have rich – phase diagrams. Transport measurements show metallic behavior and moderate positive magnetoresistance of 60–150% at 2 K. The field dependent magnetization measurements on non-moment bearing YAlGe show clear de Haas–van Alphen oscillations, suggesting small, high mobility pockets at the Fermi level. DFT calculations indicate the nonmagnetic band structure contains nodal loops that are gaped by spin-orbit coupling, supporting a weak topological insulator phase. The overview of physical properties presented here lays the foundation for future, more detailed work on individual RAlGe members.
Acknowledgments
Work at Ames National Laboratory was supported by the U.S. Department of Energy, Office of Science, Basic Energy Sciences, Materials Sciences and Engineering Division. Ames National Laboratory is operated for the U.S. Department of Energy by Iowa State University under Contract No. DE-AC02-07CH11358. Some of the computation used resources of the National Energy Research Scientific Computing Center (NERSC), a DOE Office of Science User Facility. We thank Hanhbbit Rhee for beginning the crystal growth and characterization of the samples discussed in this paper.
*corresponding authors’ email: [email protected], [email protected]
Data Availability. The data that support the findings of this study will be made available in DataShare, an open-access repository at Iowa State University at DOI: 10.25380/ias-state.32226210.
Conflicts of Interest The authors have no conflicts of interest to declare.
Estimation of leading crystal field parameters
As outlined in the discussion section of the main text, the RAlGe ( Tb–Tm) compounds show evidence for strong crystal field effects influencing the magnetic anisotropy and de Gennes scaling of . Whereas the full CEF Hamiltonian for orthorhombic () RAlGe, with rare earth atoms in point symmetry, is exceedingly complex, we can use a simplified Hamiltonian to estimate the leading crystal field parameters. If only the second-order terms are assumed to be dominant, an approximate CEF Hamiltonian can be expressed as
In this approximation, the CEF parameters can be directly related to the anisotropic Weiss temperatures obtained from Curie-Weiss fits to the magnetic susceptibility as follows [48, 53, 54]:
where is the total angular momentum for each respective rare-earth ion and represents the paramagnetic Weiss temperature from only the molecular field. Here we assume the polycrystalline averaged gives a reasonable estimate of the magnetization without the influence of the crystal field so . The anisotropic curves and associated Curie-Weiss fits are shown in Figure 13.
Equations (A.2) provide several routes to estimate the leading crystal field parameters. Using to determine , can then be independently estimated using either or . Alternatively, and can be used to determine and without use of . The results of both approaches are plotted in Figure 11c and Figure 11d and reveal two main conclusions. First, trends monotonically across the series and changes sign from positive for –Ho to negative for and Tm. Second, the absolute value of both parameters, especially , are poorly determined by this approach, with extreme results from differing calculations differing by an factor of nearly four for TbAlGe and TmAlGe. The wide range of possible values may reflect either unaccounted for uncertainties in the determination of the Weiss temperatures and/or the limitations of using a Hamiltonian truncated at second-order terms.
The absolute value of both parameters, especially , are poorly determined by this approach, with extreme results from differing calculations differing by an factor of nearly four for TbAlGe and TmAlGe. The wide range of possible values may reflect either unaccounted for uncertainties in the determination of the Weiss temperatures and/or the limitations of using a Hamiltonian truncated at second-order terms.
Despite providing a relatively poor estimate of the absolute values of and , we can still gain useful information from the inferred crystal field parameters. The sign of determines whether the system displays an easy -axis (if ) or easy -plane (if ), and leads to further asymmetry between the - and -axes. In the reference frame with the - crystal field axis along the -axis of the RAlGe structure, the crystal field's - and -axis are along the respective - and - crystallographic directions. In this basis, a negative will therefore favor an easy -axis, whereas if is positive, a positive(negative) will favor an easy -axis. Examining Fig. A.14 and using any set of the estimated parameters, we find positive and for –Ho samples, indicating an easy -axis and for , Tm, a negative is found, indicating an easy -axis. These qualitative observations support that the gross features of the magnetic anisotropy in the RAlGe (–Tm) series are dictated by the crystal field. In particular, the change from having an easy - to - axis on crossing from HoAlGe to ErAlGe is due to the changing sign of the leading parameter.
Figure A.14. Second-order crystal field parameters of the RAlGe – series estimated using the experimental Weiss temperatures and Eqs. A.2. (a) and (b) .
Crystallographic tables
| Atom | Site | Occ. | (Å) | |||
| Ho | 4c | 1 | 0.18991(2) | 0.75 | 1 | 0.00643(13) |
| Ge | 4c | 0.5 | 0.39318(6) | 0.75 | 1 | 0.00687(15) |
| Al | 4a | 0 | 0.5 | 0.5 | 1 | 0.0082(3) |
Table B.2. Atomic positions, site occupancies, and isotropic thermal displacement parameters for HoAlGe determined from single crystal X-ray diffraction refinement.
| Atom | ||||||
| Ho | 0.00528(16) | 0.00652(17) | 0.00749(18) | 0 | 0 | 0 |
| Ge | 0.0049(3) | 0.0070(3) | 0.0086(3) | 0 | 0 | 0 |
| Al | 0.0059(8) | 0.0103(8) | 0.0083(8) | 0.0000(6) | 0 | 0 |
Table B.3. Anisotropic thermal displacement parameters (in Å) for HoAlGe determined from single crystal X-ray diffraction refinement.
| Chemical formula | HoAlGe |
| Formula weight (g/mol) | 264.5 |
| Temperature | 297.49(10) |
| Wavelength (Å, Ag K) | 0.56087 |
| Crystal system | orthorhombic |
| Space group | (#63) |
| Unit cell dimensions | |
| 4.0344(3) Å alpha = 90 | |
| 10.3634(7) Å beta = 90 | |
| 5.7290(4) Å gamma = 90 | |
| Volume (cm) | 239.53(3) |
| 4 | |
| Calculated density (g/cm) | 7.335 |
| Absorption coefficient (mm) | 24.024 |
| Absorption Correction | multi-scan |
| 448 | |
| Crystal size (mm) | 0.28 x 0.19 x 0.1 |
| theta range for data collection | 3.0840-31.537 |
| Index ranges (min/max, , , ) | [-7/7, -18/19, -10/7] |
| Reflections collected | 2645 |
| Independent reflections | 436 () |
| Completeness | 100 |
| Refinement method | Full-matrix least-squares on |
| Data / restrains / parameters | 436 / 0 / 14 |
| GOF | 1.043 |
| Final indices [] | , |
| indices [all data] | , |
| Extinction coefficient | 0.078(3) |
| Largest diff. peak and hole (e-/Å) | 3.627 and -3.309 |
Table B.4. Single crystal X-ray diffraction data and structural refinement information for HoAlGe.
References
- [1]S.-Y. Xu, N. Alidoust, G. Chang, H. Lu, B. Singh, I. Belopolski, D. S. Sanchez, X. Zhang, G. Bian, H. Zheng, et al., Discovery of Lorentz-violating type II Weyl fermions in LaAlGe, Science advances 3 (6) (2017) e1603266.
- [2]G. Chang, B. Singh, S.-Y. Xu, G. Bian, S.-M. Huang, C.-H. Hsu, I. Belopolski, N. Alidoust, D. S. Sanchez, H. Zheng, et al., Magnetic and noncentrosymmetric Weyl fermion semimetals in the RAlGe family of compounds (R = rare earth), Physical Review B 97 (4) (2018) 041104.DOI
- [3]H. Flandorfer, D. Kaczorowski, J. Gröbner, P. Rogl, R. Wouters, C. Godart, A. Kostikas, The systems Ce–Al–Si, Ge: phase equilibria and physical properties, Journal of Solid State Chemistry 137 (2) (1998) 191–205.DOI
- [26]P. C. Canfield, New materials physics, Reports on Progress in Physics 83 (1) (2019) 016501.
- [27]P. C. Canfield, T. Kong, U. S. Kaluarachchi, N. H. Jo, Use of frit-disc crucibles for routine and exploratory solution growth of single crystalline samples, Philosophical Magazine 96 (1) (2016) 84–92.
- [28]Canfield crucible sets, https://lspceramics.com/canfield-crucible-sets-2/, accessed: 2022-03-23.
- [29]B. H. Toby, R. B. Von Dreele, GSAS-II: the genesis of a modern open-source all purpose crystallography software package, Journal of Applied Crystallography 46 (2) (2013) 544–549.
- [30]P. Hohenberg, W. Kohn, Inhomogeneous electron gas, Physical review 136 (3B) (1964) B864.
- [31]W. Kohn, L. J. Sham, Self-consistent equations including exchange and correlation effects, Physical review 140 (4A) (1965) A1133.
- [32]J. P. Perdew, K. Burke, M. Ernzerhof, Generalized gradient approximation made simple, Physical review letters 77 (18) (1996) 3865.
- [33]P. E. Blöchl, Projector augmented-wave method, Physical review B 50 (24) (1994) 17953.pubmed.ncbi.nlm.nih.gov/9976227
- [34]G. Kresse, J. Furthmüller, Efficiency of ab-initio total energy calculations for metals and semiconductors using a plane-wave basis set, Computational materials science 6 (1) (1996) 15–50.
- [35]G. Kresse, J. Furthmüller, Efficient iterative schemes for ab initio total-energy calculations using a plane-wave basis set, Physical review B 54 (16) (1996) 11169.
- [36]H. J. Monkhorst, J. D. Pack, Special points for brillouin-zone integrations, Physical review B 13 (12) (1976) 5188.
- [37]P. E. Blöchl, O. Jepsen, O. K. Andersen, Improved tetrahedron method for Brillouin-zone integrations, Physical Review B 49 (23) (1994) 16223.
- [38]N. Marzari, D. Vanderbilt, Maximally localized generalized Wannier functions for composite energy bands, Physical review B 56 (20) (1997) 12847.
- [39]I. Souza, N. Marzari, D. Vanderbilt, Maximally localized Wannier functions for entangled energy bands, Physical Review B 65 (3) (2001) 035109.
- [40]M. E. Fisher, Relation between the specific heat and susceptibility of an antiferromagnet, Philosophical Magazine 7 (82) (1962) 1731–1743.
- [41]M. E. Fisher, J. Langer, Resistive anomalies at magnetic critical points, Physical Review Letters 20 (13) (1968) 665.
- [42]S. Bud’ko, Z. Islam, T. Wiener, I. Fisher, A. Lacerda, P. Canfield, Anisotropy and metamagnetism in the RNi₂Ge₂ (R = Y, La–Nd, Sm–Lu) series, Journal of Magnetism and Magnetic Materials 205 (1) (1999) 53–78.
- [43]M. Hutchings, Point-Charge Calculations of Energy Levels of Magnetic Ions in Crystalline Electric Fields, Vol. 16 of Solid State Physics, Academic Press, 1964. doi:https://doi.org/10.1016/S0081-1947(08)60517-2.
- [44]D. Noakes, G. Shenoy, The effect of a crystalline electric field on the magnetic transition temperatures of rare-earth rhodium borides, Physics Letters A 91 (1) (1982) 35–36.
- [45]B. Dunlap, Crystalline electric field effects on the paramagnetic susceptibility of polycrystalline materials, Journal of magnetism and magnetic materials 37 (2) (1983) 211–214.
- [46]A. Scheie, PyCrystalField: software for calculation, analysis and fitting of crystal electric field Hamiltonians, Applied Crystallography 54 (1) (2021) 356–362.
- [47]K. Stevens, Matrix elements and operator equivalents connected with the magnetic properties of rare earth ions, Proceedings of the Physical Society. Section A 65 (3) (1952) 209–215.
- [48]Y. Hashimoto, H. Fujii, H. Fujiwara, T. Okamoto, Magnetic Properties of Rare Earth Copper Intermetallic Compounds, RCu₂. I. Heavy Rare Earth., Journal of the Physical Society of Japan 47 (1) (1979) 67–72.
- [49]E. Morosan, S. L. Bud’ko, P. Canfield, M. Torikachvili, A. Lacerda, Thermodynamic and transport properties of RAgGe (R = Tb–Lu) single crystals, Journal of magnetism and magnetic materials 277 (3) (2004) 298–321.
- [50]P. Boutron, Exact Calculation of the Paramagnetic Susceptibility of a Single Crystal with Arbitrary Crystal Field and Exchange Interactions, Phys. Rev. B 7 (1973) 3226–3238. doi:10.1103/PhysRevB.7.3226.
- [51]X. Zhang, Z. Liu, Q. Cui, Q. Guo, N. Wang, L. Shi, H. Zhang, W. Wang, X. Dong, J. Sun, et al., Electronic and magnetic properties of intermetallic kagome magnets RV₆Sn₆ (R = Tb–Tm), Physical Review Materials 6 (10) (2022) 105001.
- [52]J. Lee, E. Mun, Anisotropic magnetic property of single crystals RV₆Sn₆ (R = Y, Gd-Tm, Lu), Physical Review Materials 6 (8) (2022) 083401.
- [53]Y.-L. Wang, Crystal-field effects of paramagnetic Curie temperature, Physics Letters A 35 (5) (1971) 383–384.
- [54]N. Shohata, Magnetic properties of rare earth gallium intermetallic compounds, Journal of the Physical Society of Japan 42 (6) (1977) 1873–1880.
- [55]K. Lee, G. F. Lange, L.-L. Wang, B. Kuthanazhi, T. V. Trevisan, N. H. Jo, B. Schrunk, P. P. Orth, R.-J. Slager, P. C. Canfield, A. Kaminsky, Discovery of a weak topological insulating state and van Hove singularity in triclinic RhBi₂, Nature communications 12 (1) (2021) 1855.DOI