This paper was converted on www.awesomepapers.org from LaTeX by an anonymous user.
Want to know more? Visit the Converter page.

Distinct topological Hall responses in CeCu2-type EuZn2 and EuCd2 films

Yuto Watanabe Department of Physics, Tokyo Institute of Technology, Tokyo 152-8551, Japan    Shinichi Nishihaya Department of Physics, Tokyo Institute of Technology, Tokyo 152-8551, Japan    Markus Kriener RIKEN Center for Emergent Matter Science (CEMS), Wako 351-0198, Japan    Ayano Nakamura Department of Physics, Tokyo Institute of Technology, Tokyo 152-8551, Japan    Masaki Uchida [email protected] Department of Physics, Tokyo Institute of Technology, Tokyo 152-8551, Japan
Abstract

Rare earth intermetallic compounds crystallized in AlB2-type and its low-symmetry derivative CeCu2-type structures potentially host diverse frustrated magnetic structures and rich magnetotransport phenomena. We report the film growth of CeCu2-type EuZn2 by molecular beam epitaxy and the observation of topological Hall responses highly contrastive to isostructural EuCd2. While their magnetization curves are rather similar, the topological Hall effect observed in EuZn2 is simpler, with the only one component enhanced at the magnetic transition field. EuZn2 may be a unique system for studying the magnetic domain boundary effect on topological Hall responses among the CeCu2-type rare-earth intermetallic compounds.

Transverse motion of carriers neither proportional to the magnetic field nor the magnetization has attracted broad interest both in fundamental and applied physics AHEreview . Such a nonmonotonic Hall response originates in a change in the real-space or momentum-space Berry curvature in the magnetization process. In particular, the Hall response reflecting a real-space spin texture is termed topological Hall effect (THE), where a noncoplanar spin texture with finite scalar spin chirality, defined by the solid angle spanned by three adjacent spins, generates an emergent magnetic field on charged carriers THE ; Nd2Mo2O7 . A magnetic skyrmion lattice is a typical example of the noncoplanar spin texture. Recently, the range of skyrmion host compounds has been expanded from noncentrosymmetric chiral to centrosymmetric frustrated magnets MnSi ; MnGe ; FeGe ; GdPdSi ; EuAl4 .

As centrosymmetric magnets which can host a skyrmion lattice, AlB2-type compounds represented by Gd2PdSi3 have attracted growing attention GdPdSi . For example in Gd2PdSi3, a magnetic Gd triangular lattice layer and a Pd/Si honeycomb lattice layer are alternately stacked to form a two-dimensionally frustrated system, where a giant topological Hall response is observed with rather monotonic change in magnetization GdPdSi . The AlB2-type structure is also famous for many lower-symmetry derivatives including the CeCu2-type one AlB2_derivative . In rare-earth intermetallic compounds RA2RA_{2} (R=R= La – Lu, A=A= Zn, Cd, or Cu) with the CeCu2-type structure, a distorted RR triangular lattice layer and a three-dimensionally buckled AA honeycomb lattice layer are alternately stacked CeCu2-type . Differences in distortion and buckling as well as magnetic anisotropy of rare-earth magnetic moments can bring about diverse frustrated magnetic structures and magnetotransport in RA2RA_{2}. In addition to helical magnetic ordering with various propagation directions and rotation planes as observed in GdCu2 and PrZn2 GdCu2_mag ; PrZn2_mag ; DyZn2_mag , rich topological Hall responses have been recently found for EuCd2 EuCd2_THE . However, comprehensive research on the magnetotransport in RA2RA_{2} is still lacking.

Figure 1(a) summarizes the ratio of out-of-plane and in-plane rare-earth atomic distances (doutd_{\mathrm{out}} and dind_{\mathrm{in}}) in CeCu2-type intermetallic compounds EuCd2 ; EuZn2 ; CeZn2_mag ; PrZn2_mag ; NdZn2 ; NdZn2_mag ; DyZn2_mag ; HoZn2 ; YbZn2 ; CeCu2 ; GdCu2 ; NdCu2 ; EuCu2 ; DyCu2 ; HoCu2 ; ErCu2 ; TmCu2 ; YbCu2 . In most of them including EuCd2, pairs of the nearest neighbor rare-earth atoms are formed between the adjacent triangular lattice layers (dout<dind_{\mathrm{out}}<d_{\mathrm{in}}). Actually, EuZn2 is the only magnetic compound with such pairs within the triangular lattice layers (dout>dind_{\mathrm{out}}>d_{\mathrm{in}}), since LaZn2 is nonmagnetic. Drastic changes in magnetotransport can be thus expected reflecting the different frustration arrangements of the Eu2+ isotropic magnetic moments EuCd2 ; EuZn2 , as compared in Figs. 1(b) and 1(c). In this paper, we report film growth of EuZn2 by molecular beam epitaxy and observation of topological Hall responses highly contrastive to EuCd2.

Thin films of EuZn2 and EuCd2 were grown on Al2O3 (0001) substrates using an Epiquest RC1100 II-V molecular beam epitaxy chamber EuAs ; EuCd2_THE ; EuCd2Sb2_film ; EuCd2As2_film ; EuCdSb2_film ; EuSb2_film . The growth temperature for EuZn2 films was typically set to 400 C. Eu and Zn were supplied by an effusion cell with a Zn-rich flux ratio (PZn/PEu=6P_{\mathrm{Zn}}/P_{\mathrm{Eu}}=6). This flux ratio is lower than in the growth of EuCd2 films (PCd/PEu=1544P_{\mathrm{Cd}}/P_{\mathrm{Eu}}=15-44) EuCd2_THE , consistent with the less volatile nature of Zn compared to Cd. The film thickness is designed at 65 nm. As confirmed in Figs. 2(a) and 2(b), x-ray diffraction 2θ2\theta-ω\omega scans for both EuZn2 and EuCd2 films exhibit sharp peaks from the (020) lattice plane without any impurity peaks, indicating that the distorted Eu triangular lattice layers are stacked on the Al2O3 (0001) plane.

Figures 2(c) – 2(f) compare fundamental magnetic properties taken for EuZn2 and EuCd2 films. The temperature dependence of the magnetization MM exhibits a cusp at the Néel temperature TNT_{\text{N}} = 35 K for EuZn2 and 37 K for EuCd2, almost consistent with previous reports in polycrystalline bulks EuZn2 ; EuCd2 . Both in the EuZn2 and EuCd2 films, MM keeps increasing below TNT_{\text{N}}, suggesting that their magnetic ground states are not a simple collinear antiferromagnetic ordering. Magnetization curves, taken with sweeping the out-of-plane field BoutB_{\text{out}} at 2 K, are rather monotonic and also quite similar between EuZn2 and EuCd2. A transition from phase I to II occurs at the metamagnetic transition field BmB_{\text{m}} = 2.3 T for EuZn2 and 1.3 T for EuCd2 with tiny hysteresis, and then the Eu2+ magnetic moments are forcedly aligned along the out-of-plane direction above the saturation field BsB_{\text{s}} = 5.1 T for EuZn2 and 3.8 T for EuCd2. Another small hysteresis loop is confirmed centered at 0 T, indicating that there is a weak ferromagnetic component at the ground state of EuZn2 as well as EuCd2.

Figure 3 summarizes magnetotransport observed for the EuZn2 film. As shown in Fig. 3(a), the temperature dependence of the longitudinal resistivity ρxx\rho_{xx} is metallic with a residual resistivity ratio (RRR) of 7. It exhibits a clear kink at TNT_{\text{N}}, reflecting strong coupling between the itinerant carriers and the localized Eu2+ magnetic moments. In the out-of-plane field dependence shown in Fig. 3(b), ρxx\rho_{xx} exhibits a peak at BmB_{\text{m}} and then negative magnetoresistance up to BsB_{\text{s}}. Figure 3(c) shows the Hall resistivity ρyx\rho_{yx} measured with sweeping the out-of-plane field BoutB_{\text{out}} at 2 K. The hole density estimated by fitting the ρyx\rho_{yx} curve above the saturation field at 2 K is 1.7×10221.7\times 10^{22} cm-3, which is similar to that of EuCd2EuCd2_THE . It is clear that there is a nonmonotnic Hall component in addition to the ordinary Hall resistivity ρyxO\rho_{yx}^{\text{O}} and the anomalous Hall resistivity ρyxA\rho_{yx}^{\text{A}}. Here ρyxO\rho_{yx}^{\text{O}} is expressed by ρyxO\rho_{yx}^{\text{O}} = R0R_{0}BoutB_{\text{out}} with the Hall coefficient R0R_{0}, and ρyxA\rho_{yx}^{\text{A}} is extracted by ρyxA\rho_{yx}^{\text{A}} = rsr_{s}ρxx\rho_{xx}M2{}^{2}M with rsr_{s} determined as a fitting parameter, since the longitudinal conductivity converted from ρxx\rho_{xx} is located in the so-called intrinsic region of the conventional scaling plot AHE_scaling . Figure 3(d) displays the additional component obtained by subtracting ρyxO\rho_{yx}^{\text{O}} and ρyxA\rho_{yx}^{\text{A}} from the measured Hall resistivity. Hereafter we call this ρyxT\rho_{yx}^{\text{T}}, which is ascribed to the topological Hall component originating in frustrated spin configuration, as discussed below.

ρyxT\rho_{yx}^{\text{T}} appears below TNT_{\mathrm{N}}, indicating that it is related to the magnetic ordering. As reported for some magnetic compounds, ρyx\rho_{yx} may be affected by the change in ρxx\rho_{xx} through a field dependence of the scattering time. In ErB4, for example, a nonmonotonic Hall component has been explained by a modified two-band model incorporating this effect ErB4 . In the present case, however, the fitting curve of ρyxO\rho_{yx}^{\text{O}}+ρyxA\rho_{yx}^{\text{A}} swells rather in the direction opposite to the hump around 2.5 T, and thus the change in the scattering time would not be the cause.

Here comparison with EuCd2 is helpful for discussion. Figures 4(a) and 4(b) compare ρyxT\rho_{yx}^{\text{T}} taken for EuZn2 and EuCd2 at 2 K. In EuZn2, only the peak P1 appears centered at BsB_{\text{s}} between phase I and II. In EuCd2, on the other hand, ρyxT\rho_{yx}^{\text{T}} is characterized by three peaks P1-P3, and among them, peak P1 similarly appears centered at BsB_{\text{s}} as in EuZn2. Another difference is that a large hysteresis loop is observed in EuCd2 accompanied with a sharp enhancement of P1 at low temperatures. The origin of THE in EuCd2 has been understood the real-space spin Berry curvature, or noncoplanar spin textures realized on the distorted Eu2+Eu^{2+} triangular lattice layers EuCd2_THE . It has been clarified based on the temperature dependence measurements that the sign of the THE peaks P1-P3 originating from the real-space spin Berry curvature remain unchanged, while ρyxA\rho_{yx}^{\text{A}} reflecting the momentum-space Berry curvature exhibits a sign change at 20 K. Although the sign change of ρyxA\rho_{yx}^{\text{A}} does not occur in EuZn2 at elevated temperatures, it is reasonable to suggest that THE in EuZn2 also originates in noncoplanar spin textures in the magnetization process, considering the similarity of the magnetization properties between the two materials. In fact, the distorted Eu triangular lattice in the acac-plane of CeCu2-type structure can host uncanceled local Dzyaloshinskii-Moriya interaction which favors spin configuration canted from the acac-plane to the out-of-plane bb-axis, possibly leading to a noncoplanar ordering.

Figures 4(c) and 4(d) compare a color-map of ρyxT\rho_{yx}^{\text{T}} taken in the downward field sweep on the BoutB_{\text{out}}-TT phase diagram. In EuZn2, ρyxT\rho_{yx}^{\text{T}} consisting only of the peak P1 appears antisymmetric to the magnetic field and continues to reach a maximum value at the magnetic transition field between phases I and II even upon increasing temperature. In EuCd2, the appearance of ρyxT\rho_{yx}^{\text{T}} is more complex and not antisymmetric to the field due to the large hysteresis at low temperatures, but the enhancement of P1 at the magnetic transition field is commonly confirmed in the negative field region. As the peak P1 is enhanced not within a specific phase but at a phase boundary, it is naturally ascribed to the magnetic domain boundary formed between the two phases. In evidence, it has been demonstrated that P1 in EuCd2 strongly depends on the field cooling processes, which can effectively modulate the domain boundary density EuCd2_THE . On the other hand, the differences in the appearance of THE between EuZn2 and EuCd2 are probably derived from different pairs of the nearest neighbor Eu atoms and dominant magnetic interactions as illustrated in Fig. 4(c). While spin modulation propagating along the cc-axis has been reported for other intermetallic compounds RRZn2 with dout<dind_{\mathrm{out}}<d_{\mathrm{in}} CeZn2_mag ; PrZn2_mag ; NdZn2_mag ; DyZn2_mag , distinctive magnetic structures are realized in EuZn2 with dout>dind_{\mathrm{out}}>d_{\mathrm{in}}, which deserve clarification in future studies. The appearance of a simple single-component THE also suggests that EuZn2 is potentially a unique system for further studying the magnetic domain boundary effect on THE among the CeCu2-type rare-earth intermetallic compounds.

In summary, we have grown EuZn2 thin films by molecular beam epitaxy and revealed topological Hall responses which are in sharp contrast to EuCd2. While their magnetization curves in the magnetization process are rather similar, the topological Hall effect observed in EuZn2 is simpler, with the only one component enhanced at the metamagnetic transition field. Hence, we propose EuZn2 as a unique system for further studying the magnetic domain boundary effect on THE among the CeCu2-type rare-earth intermetallic compounds.

Acknowledgements.
We thank M. Kawasaki for help in a part of the magnetization measurements. We also thank H. Ishizuka, Y. Yamasaki, T. Nakajima, F. Kagawa, and K. Matsuura for fruitful discussions. This work was supported by JST FOREST Program Grant No. JPMJFR202N, Japan, by Grant-in-Aids for Scientific Research JP22H04471, JP22H04501, JP22K18967, JP22K20353, JP23K13666, JP24H01614, and JP24H01654 from MEXT, Japan, and by STAR Award funded by the Tokyo Tech Fund, Japan.

Conflict of Interest

The authors have no conflicts to disclose.

Author Contributions

Yuto Watanabe: Data curation (lead); Formal analysis (equal); Investigation (lead); Methodology (equal); Visualization (lead); Writing - original draft preparation (lead). Shinichi Nishihaya: Formal analysis (equal), Investigation (equal), Methodology (equal), Writing-review & editing (equal). Markus Kriener: Formal analysis (supporting), Investigation (equal), Writing - review & editing (equal). Ayano Nakamura: Formal analysis (supporting), Investigation (supporting), Writing - review & editing (equal). Masaki Uchida: Conceptualization (lead); Funding acquisition (lead); Methodology (lead); Project administration (lead); Supervision (lead); Writing - review & editing (lead).

Data Availability

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

References

  • (1) N. Nagaosa, J. Sinova, S. Onoda, A. H. MacDonald, and N. P. Ong, Rev. Mod. Phys. 82, 1539–1592 (2010).
  • (2) K. Ohgushi, S. Murakami, and N. Nagaosa, Phys. Rev. B 62, R6065(R) (2000).
  • (3) Y. Taguchi, Y. Oohara, H. Yoshizawa, N. Nagaosa, and Y. Tokura, Science 291, 2573-2576 (2001).
  • (4) S. Mühlbauer, B. Binz, F. Jonietz, C. Pfleiderer, A. Rosch, A. Neubauer, R. Georgii, and P. Böni, Science 323, 915-919 (2009).
  • (5) N. Kanazawa, Y. Nii, X.-X. Zhang, A. S. Mishchenko, G. De Filippis, F. Kagawa, Y. Iwasa, N. Nagaosa, and Y. Tokura, Nat. Commun. 7, 11622 (2016).
  • (6) X. Z. Yu, N. Kanazawa, Y. Onose, K. Kimoto, W. Z. Zhang, S. Ishiwata, Y. Matsui, and Y. Tokura, Nat. Mater. 10, 106-109 (2011).
  • (7) T. Kurumaji, T. Nakajima, M. Hirschberger, A. Kikkawa, Y. Yamasaki, H. Sagayama, H. Nakao, Y. Taguchi, T. Arima, and Y. Tokura, Science 365, 914-918 (2019).
  • (8) R. Takagi, N. Matsuyama, V. Ukleev, L. Yu, J. S. White, S. Francoual, J. R. L. Mardegan, S. Hayami, H. Saito, K. Kaneko, K. Ohishi, Y. O¯\bar{\mathrm{O}}nuki, T. Arima, Y. Tokura, T. Nakajima, and S. Seki, Nat. Commun. 13, 1472 (2022).
  • (9) R.-D. Hoffman and R. Pöttgen, Z. Kristallogr. 216, 127-145 (2001).
  • (10) D. Debray, J. Less-Common Met. 30, 237-248 (1973).
  • (11) K. H. J. Buschow, and F. J. van Steenwijk, Physica B 85, 122-126 (1977).
  • (12) K. H. J. Buschow, W. J. Huiskamp, H. Th LeFever, F. J. v Steenwijk, and R. C. Thiel, J. Phys. F: Met. Phys. 5, 1625 (1975).
  • (13) D. Debray, M. Sougi, and P. Meriel, J. Chem. Phys. 56, 4325-4328 (1972).
  • (14) M. Ohashi, T. Kitai, T. Kaneko, S. Abe, S. Funahashi, and Y. Yamaguchi, J. Magn. Magn. Mater. 140, 1119-1120 (1995).
  • (15) G. Borzone, G. Cacciamani, R. Ferro, J. Charles, and J. Hertz, J. Less-Common Met. 128, 297-312 (1987).
  • (16) D. Debray and M. Sougi, J. Chem. Phys. 58, 1783-1786 (1972).
  • (17) M. Ohashi, T. Kitai, T. Kaneko, H. Yoshida, Y. Yamaguchi, and S. Abe, J. Magn. Magn. Mater. 90, 585-586 (1990).
  • (18) D. Debray and M. Sougi, J. Chem. Phys. 57, 2156-2189 (1972).
  • (19) D. J. Michel and E. Ryba, J. Less-Common Met. 18, 159-165 (1969).
  • (20) R. Trump, S. Thierfeldt, M. Loewenhaupt, and T. Chattopadhyay, J. Appl. Phys. 69, 4699-4701 (1991).
  • (21) C. S. Garde, J.Ray, and G. Chandra J. Phys.: Condens. Matter 5, 6737-6744 (1993).
  • (22) M. Rotter, A. Lindbaum, E. Gratz, G. Hilscher, H. Sassik, H. E. Fischer, M. T. Fernandez-Diaz, R Arons, and E. Seidl, J. Mag. Mag. Mater. 214, 281-290 (2000).
  • (23) Z. Kanghou, and H. Chunxiao, and C. Lili, J. Alloys Compd. 189, L31-L33 (1992).
  • (24) G. A. Costa and E. A. Franceschi, J. Less-Common Met. 106, 175-182 (1985).
  • (25) E. Franceschi, J. Less-Common Met. 87 249-256 (1982).
  • (26) Z. Smetana and V. Šíma, J. Magn. Magn. Mater. 59, 145-152 (1986).
  • (27) A. R. Storm and K. E. Benson, Acta Cryst. 16, 701-702 (1963).
  • (28) E. Gratz, A. Lindbaum, and M. Rotter, J. Phys.: Condens. Matter 5, 7955-7958 (1993).
  • (29) A. Iandelli and A. Palenzona, J. Less-Common Met. 25, 333-335 (1971).
  • (30) S. Nishihaya, Y. Watanabe, M. Kriener, A. Nakamura, and M. Uchida, Phys. Rev. B 110, 035159 (2024).
  • (31) M. Uchida, S. Sato, H. Ishizuka, R. Kurihara, T. Nakajima, Y. Nakazawa, M. Ohno, M. Kriener, A. Miyake, K. Ohishi, T. Morikawa, M. S. Bahramy, T. Arima, M. Tokunaga, N. Nagaosa, and M. Kawasaki, Sci. Adv. 7, eabl5381 (2021).
  • (32) M. Ohno, S. Minami, Y. Nakazawa, S. Sato, M. Kriener, R. Arita, M. Kawasaki, and M. Uchida, Phys. Rev. B 105, L201101 (2022).
  • (33) S. Nishihaya, A. Nakamura, M. Ohno, M. Kriener, Y. Watanabe, M. Kawasaki, and M. Uchida, Appl. Phys. Lett. 124, 023103 (2024).
  • (34) M. Ohno, M. Uchida, Y. Nakazawa, S. Sato, M. Kriener, A. Miyake, M. Tokunaga, Y. Taguchi, and M. Kawasaki, APL Mater. 9, 051107 (2021).
  • (35) M. Ohno, M. Uchida, R. Kurihara, S. Minami, Y. Nakazawa, S. Sato, M. Kriener, M. Hirayama, A. Miyake, Y. Taguchi, R. Arita, M. Tokunaga, and M. Kawasaki, Phys. Rev. B 103, 165144 (2021).
  • (36) S. Onoda, N. Sugimoto, and N. Nagaosa, Phys. Rev. B 77, 165103 (2008).
  • (37) L. Ye, T. Suzuki, and J. G. Checkelsky, Phys. Rev. B 95, 174405 (2017).
Refer to caption
Figure 1: Crystal structure and atomic distances of CeCu2-type rare-earth intermetallic compounds. (a) Ratio of out-of-plane and in-plane rare-earth atomic distances (dout/dind_{\text{out}}/d_{\text{in}}) in CeCu2-type RA2RA_{2} (R=R= La – Lu, A=A= Zn, Cd, or Cu) EuCd2 ; EuZn2 ; CeZn2_mag ; PrZn2_mag ; NdZn2 ; NdZn2_mag ; DyZn2_mag ; HoZn2 ; YbZn2 ; CeCu2 ; GdCu2 ; NdCu2 ; EuCu2 ; DyCu2 ; HoCu2 ; ErCu2 ; TmCu2 ; YbCu2 . Crystal structure of (b) EuZn2 and (c) EuCd2, composed of alternate stacking of a distorted Eu triangular lattice layer and a buckled Zn or Cd honeycomb lattice layer along the bb-axis. While pairs of the nearest neighbor Eu atoms are formed within the triangular lattice layers in EuZn2 (dout>dind_{\text{out}}>d_{\text{in}}), they are between the adjacent layers in EuCd2 (dout<dind_{\text{out}}<d_{\text{in}}).
Refer to caption
Figure 2: Magnetic properties of EuZn2 and EuCd2 thin films. X-ray diffraction 2θ\theta-ω\omega scans of (010)-oriented (a) EuZn2 and (b) EuCd2 films grown on the Al2O3 (0001) substrate. An asterisk denotes the diffraction peaks from the Al2O3 substrate. Temperature dependence of the magnetization in (c) EuZn2 and (d) EuCd2 films, measured in an out-of-plane magnetic field of Bout=0.1B_{\mathrm{out}}=0.1 T. Magnetization curves of (e) EuZn2 and (f) EuCd2 films taken with sweeping BoutB_{\mathrm{out}} at 2 K. Data are normalized by the saturation magnetization MsM_{\mathrm{s}}.
Refer to caption
Figure 3: Magnetotransport of the EuZn2 film. (a) Temperature dependence of longitudinal resistivity ρxx\rho_{xx}. Out-of-plane field dependence of (b) ρxx\rho_{xx} and (c) Hall resistivity ρyx\rho_{yx} at 2 K. ρyx\rho_{yx} data above the saturation field is fitted by a green curve consisting of the ordinary Hall component (ρyxO=R0\rho_{yx}^{\mathrm{O}}=R_{0}BoutB_{\text{out}}) and the anomalous Hall component (ρyxA=rsρxx2M)\rho_{yx}^{\mathrm{A}}=r_{\mathrm{s}}\rho_{xx}^{2}M) with field-independent constants R0R_{0} and rsr_{\mathrm{s}}. (d) Topological Hall component ρyxT\rho_{yx}^{\text{T}} at 2 K, obtained by subtracting ρyxO+ρyxA\rho_{yx}^{\mathrm{O}}+\rho_{yx}^{\mathrm{A}} from ρyx\rho_{yx} in (c).
Refer to caption
Figure 4: Comparison of the topological Hall effect observed in EuZn2 and EuCd2. Topological Hall resistivity ρyxT\rho_{yx}^{\text{T}} plotted together with the magnetization MM for (a) EuZn2 and (b) EuCd2 films at 2 K. A solid (dashed) curve represents the data taken upon decreasing (increasing) the field from +7 (-7) T to -7 (+7) T. Mapping of ρyxT\rho_{yx}^{\text{T}} on the BoutB_{\text{out}}-TT phase diagram for (c) EuZn2 and (d) EuCd2 films. Data taken in the downward field sweep is shown on the map. Magnetic transition fields determined from dM/dBdM/dB are denoted by a diamond, and magnetic phase boundaries are represented by a dashed curve for a guide to the eye.