Physica B 459 (2015) 46–51
Contents lists available at ScienceDirect
Physica B journal homepage: www.elsevier.com/locate/physb
Effect of substitution of Co with Fe on the structural, electronic and magnetic properties of Heusler alloy Co2CrAl Jagdish Nehra, N. Lakshmi n, K. Venugopalan Department of Physics, Mohanlal Sukhadia University, Udaipur 313 001, Rajasthan, India
art ic l e i nf o
a b s t r a c t
Article history: Received 16 October 2014 Received in revised form 22 November 2014 Accepted 24 November 2014 Available online 26 November 2014
We investigate the effect of the substitution of Co with Fe on the structural, electronic and magnetic properties of a series of quaternary Co2 xFexCrAl Heusler alloys. The alloy orders in the B2 structure for composition of x Z0.4. The half metallicity in the parent Co2CrAl is retained in the whole series of alloys on partial substitution of Co with Fe atoms. DC magnetization studies evidence a linear increase in saturation magnetization and Curie temperature increase in Fe content. Mössbauer spectra show that along the series the average hyperﬁne ﬁeld and combined relative area under the curve of sextets increase linearly with increase in Fe content corroborating results obtained from DC magnetization studies. & 2014 Elsevier B.V. All rights reserved.
Keywords: Half-metallicity Mössbauer spectroscopy First-principles calculations Heusler alloys
1. Introduction Ever since the possibility of existence of half-metallic ferromagnets (HMF's) had been predicted through band calculations by de Groot et al. , intense investigations using a combination of theoretical and experimental techniques have been undertaken. The interest in this class of materials is due to their potential for use in spin-dependent devices such as those making use of tunneling magnetoresistance (TMR) and magnetic random access memory (MRAM) devices. Among the different types of materials which show promise as HMF, full Heusler alloys with L21 and half Heusler alloys with Clb type structures are special due to attractive magnetic properties like high spin moment, high Curie temperature, ease of tailoring of physical properties through substitution with other elements etc. The L21 and the B2-type Co2CrAl alloys have been reported to exhibit evidence of HMF in their band structures calculations. However, the experimental saturation magnetic moment and spin polarization of the B2-type Co2CrAl alloy are signiﬁcantly lower than the expected theoretical values and due to the unavoidable A2 phase separation in it which signiﬁcantly reduces the saturation moment of Co2CrAl and the halfmetallicity is also destroyed. In comparison to Co-based Heusler alloys, most of the Fe-based Heusler alloys have lower Curie temperatures and thus are not realistic candidates for practical applications for room temperature operating spintronic devices. Studies on the electronic properties of Fe based Fe2YAl (Y¼Ti, V, Cr) n
Corresponding author. E-mail address: [email protected]
http://dx.doi.org/10.1016/j.physb.2014.11.096 0921-4526/& 2014 Elsevier B.V. All rights reserved.
Heusler show that these alloys have low energy band gaps structure . Heusler alloys provide the opportunity of easily tunable electronic and magnetic properties that depend on their number of valence electron concentration. Besides ternary X2YZ compounds, the easiest way to synthesize compose new materials is by the exchange of one or more of the elements X, Y, and Z. For example, the substitutional quaternary alloys of the type X2Y1 xY'xZ or X2YZ1 xZ'x, (X1 xXx)2YZ. Co based alloys of the ﬁrst two types have been investigated by various groups [3,4]. For example, theoretical investigations by Galanakis  show that the Heusler alloys Co2Cr1 xMnxAl and Co2MnAl1 xSnx have are half metallic ferromagnets and follow the Slater Pauling rule. Nakatani et. al.  have made theoretical and experimental investigations of the electronic and magnetic properties of Co2 xFexCrGa Heusler alloy and have shown that spin polarization was enhanced from 0.61 to 0.67 by a small substitution of Fe (x¼0.1) for Co in Co2CrGa. This indicates that a small amount of Fe substitution for Co leads to an enhancement in the half-metallicity of this alloy system while Heusler alloys like Mn2VAl which are ferrimagnetic due to antiparallel spin moments are important due to smaller energy losses [7–10]. This work focuses on quaternary alloys since reducing the valence electrons in the Co2CrAl is a feasible way to tailor the position of the Fermi level within the half-metallic band gap. This decrease in the total number of valence electrons can also be made by the gradual substitution of the Co element by another D-block element, for e.g. Fe, onto the same sub lattice with a smaller number of valence electrons. The previous ab initio calculations motivated us to explore the stability and the half-metallicity of
J. Nehra et al. / Physica B 459 (2015) 46–51
Co2 xFexCrAl alloys and in this paper we study the evolution of magnetic and electronic properties by a systematic substitution of Co with Fe. In an earlier paper we had reported the stability of Co2CrAl by the partial substitution of Fe for Co which eliminates the phase separation while retaining the half metallicity along with high Curie temperature making it attractive for technological applications.
2. Experimental and theoretical calculations Starting from 3 N Co, Cr, Fe, and Al elements, the samples of Co2 xFexCrAl (x ¼0, 0.2, 0.4, 0.5) were synthesized by arc melting. In order to assure sample homogeneity, the melting process was repeated three times. From this sample, one piece of the ingot was crushed and powdered to a grain size of approximately 100 μm and annealed under conditions of high vacuum (∼10 5 mbar) at 573 K for 15 days in a sealed, evacuated quartz ampoule. The crystal structure was characterized by X-ray diffraction (XRD) using Cu Kα radiation. DC Magnetization measurements and Curie temperature were determined using a Lakeshore 7304 model vibrating-sample magnetometer (VSM). 57Fe Mössbauer spectroscopy at room temperature was performed to gain additional structural and magnetic hyperﬁne ﬁeld information. The spectra were obtained using Mössbauer spectrometer operated in a constant acceleration mode with 57Co source embedded in a rhodium matrix. The spectrometer was calibrated using a standard α-Fe foil. Electronic properties were calculated using the scalar-relativistic full potential linearized augmented plane wave method (FLAPW) provided by Blaha et. al. . The exchange—correlation function was taken within the generalized gradient approximation (GGA) in the parameterization of Perdew et. al. . Spin-orbit coupling was neglected. The energy threshold between the core and valence states was set to 6.0 Ry. Here, the mufﬁn tin sphere radii, R, used were 2.3 au for Co, Fe and Cr and 2.19 au for Al as default values which resulted in nearly touching spheres. The density plane wave cutoff RKmax ¼7.0 and 783 irreducible K points for Brillouin zone integration were used. The convergence criteria of the self-consistent calculations were chosen for charge 0.1 mCoulomb and stability was better than 0.01 mRy for the total energy per cell. Initially, 16 atoms in the L21 unit cell composed of eight Co, four Cr and four Al atoms were taken for the self-consistency calculations for Co2CrAl alloy except for Co1.5Fe0.5CrAl, which is modeled using a tetragonal body central cubic (bct) cell as reported by Alhaj et al. . The space group of this structure is P-4 m2 (115). The sample for x ¼0.25 was calculated with the P-43 m (215) space group, whereas the stoichiometric Co2CrAl full Heusler alloy was calculated with the Fm-3 m (225) space group. For whole study in the series of Co2 xFexCrAl, Fe addition was carried out by gradual replacement of those Co atoms by Fe atoms. In this way, replacement of one of the eight Co atoms by a Fe atom results in x¼ 0.25, while the addition of two Fe atoms in the place of eight Co atoms results in x ¼0.50 etc. in the formula for Co2 xFexCrAl.
3. Results and discussion 3.1. Structural characterization Fig. 1 shows the XRD patterns of all the annealed Co2 xFexCrAl samples recorded at room temperature. The Rietveld reﬁnement (using the FULLPROF Suite program) of the XRD patterns conﬁrms that all of these annealed samples maintains the single phase structure. The ternary full Heusler alloys have L21 structure, which contains four atoms that form the base of the fcc primitive cell.
Fig. 1. XRD pattern of Co2 xFexCrAl (x ¼ 0, 0.2, 0.4, 0.5) samples with Rietveld ﬁtting.
This structure exhibits the Fm-3m symmetry. The Wyckoff positions 8c (1/4, 1/4, 1/4), 4a (0, 0, 0) and 4b (1/2, 1/2, 1/2) are occupied by Co, Cr, and Al, respectively in Co2CrAl alloy. In quaternary Co2 xFexCrAl alloy, the symmetry is reduced by center of inversion if one of the Co atoms is replaced by another type of transition metal Fe. This structure exhibits the lower symmetric F-43 m space group consisting of a primitive fcc cell with a basis containing four atoms on the Wyckoff positions 4a, 4b and where 8c is divided into two parts 4c (1/4, 1/4, 1/4), and 4d (3/4, 3/4, 3/4). The Co atoms are situated on the 4a site whereas the Fe atoms occupy the 4b site. The Cr and Al atoms are located on the 4c and 4d site respectively. The diffraction patterns of these samples clearly indicate that the samples for x ¼0.0 and 0.2 have crystallized in the disordered A2 single phase because of the total absence of the (111) and (200) super lattice reﬂections. The other two samples, i.e., corresponding to values of x ¼0.4 and 0.5 have crystallized in the partially ordered B2 type structure as evidenced by the clear presence of the superlattice (200) reﬂection. However the absence of the (111) peak indicates that in this sample, disorder exists between the atoms at the Y and Z sites resulting in the B2-type structure rather than the fully ordered L21. Therefore in samples with x ¼0.4 and 0.5, disorder between the Cr and Al atoms alone exists. In effect, in these two samples, the Co/Fe is still conﬁned to the 4a and 4b sites, while atoms at the Y (Cr) and Z (Al) sites are evenly distributed between the 4c and 4d sites. There is a systematic, linear decrease in the lattice parameter along the series (Fig. 2) with increase in Fe content according to Vegard's law , suggesting that Fe replaces the Co with good structural stability instead of forming a secondary phase. 3.2. Electronic properties of Co2 xFexCrAl alloys The total density of states for Co2 xFexCrAl for x ¼0, 0.25 and 0.5 are shown in Fig. 3 from which it is evident that the majorityspin channel exhibits a metallic behavior, while the minority-spin channel is semiconducting with energy gaps crossed by the Fermi level leading to a complete 100% spin polarization of the conduction electrons. Band structures of the alloys, shown in Fig. 4, have direct band gaps at the Γ point. These calculations show that half-metallic behavior is preserved by alloying the full Heusler alloy Co2CrAl by Fe at the Co sites. Our calculations are in agreement with results of similar calculations done for the quaternary Heusler alloys Co2Fe1 xMnxSi and Co2 xCrxCrAl (x ¼0, 0.25, 0.5, 0.75, 1) which have been reported to be half metallic [15,16]. Thus
J. Nehra et al. / Physica B 459 (2015) 46–51
ferrimagnets on partial substitution of Co with Fe. These results are similar to those obtained recently by Alhaj et al.  who have also predicted Co2 xFexCrAl to be half metallic ferrimagnets using ﬁrst-principles calculations. The minority density of states (DOS), shown in Fig. 3 exhibits a clear gap around the Fermi energy (EF) which is in situated at the conduction band side in the parent Co2CrAl Heusler alloy. Shifting of the Fermi level towards the middle of the gap conﬁrms the more stable half-metallic character of the material even after substitution of Co for Fe at the 4a/4b sites. The high density below EF is dominated by d states being located at Co and Fe sites. The states of 3d metal atoms extend from 4 to þ 2 eV and hybridize with each other. Galanakis et. al. already have reported  that the covalent hybridization between the lower energy D-states of the high-valent transition metal atom Co/Fe and the higher-energy D-states of the lower-valent transition metal Cr can lead to the formation of bonding and antibonding bands. The bonding and antibonding states are mainly localized at the high-valent transition metal Co/Fe and lower valent transition metal Cr sites. So, a d– d band gap is formed near the Fermi level. Fig. 2. Variation of lattice constant with Fe content in Co2 xFexCrAl. The line is a linear ﬁt to experimental points.
Fig. 3. Spin-resolved density of states for ordered Co2CrAl, Co1.75Fe0.25CrAl and Co1.5Fe0.5CrAl samples.
the reduction in the total number of valence electrons by partial replacement of the high valence element Co by the low valence element Fe is found to be an effective way to achieve a shift of Fermi energy within the half-metallic band gap [17–20]. Similarly, in another study it was found that the initial increase in spin polarization with x in the Co2Fe1 xCrxSi alloy series can be attributed to the same effect . The results of our study show that, the majority spin electrons are metallic whereas there is an energy gap varying from 0.68 eV to 0.21 eV around the Fermi level in the bands for the minority spin electrons in samples with x ¼0–0.5. The Al moment stays small and is negligible. On substitution, the Fe moment varies from 1.12 to 0.96 μB and from 0.82 to 0.92 μB for Co. However, as is evident from Table 1, the total atomic spin moment decreases slightly with the substitution of Co by Fe since the spin magnetic moment of Fe atoms is antiferromagnetically coupled to Co and Cr. Results of theoretical calculations thus show that while Co2CrAl is a half metallic ferromagnet, the series becomes half metallic
3.3. Mössbauer study The RT Mössbauer spectra of all the samples shown in Fig. 5 and reported in this study exhibit a paramagnetic singlet along with three sextets. The sextets corresponding to the subspectra have been labeled as S1, S2 and S3 respectively in decreasing order of hyperﬁne ﬁeld values. Since these spectra are not well resolved, they have ﬁrst been ﬁtted for a distribution of hyperﬁne ﬁelds with the constraint that the width of an individual lorentzian corresponds to that of metallic Fe. The probability distribution clearly shows the presence of three distinct hyperﬁne ﬁelds in all the samples respectively, with variation in the probabilities of individual hyperﬁne ﬁeld values. To get more detailed information on the relative areas, intensities, widths and isomer shifts, the spectra have been re-ﬁtted for discrete sextets using the WinNormos site ﬁtting program. Average hyperﬁne ﬁeld values obtained from the hyperﬁne ﬁeld distribution were used for initial site assignment and then reﬁned. In the fully ordered XX'YZ type of quaternary Heusler alloys structure, X' (Fe) atoms has 8 atoms as ﬁrst nearest neighbor (nn) consisting of 4X (Co) at the 4a site and 4Y (Cr) at the 4c sites. The next nearest neighbor (nnn) consists of 6 (3X and 3Y) atoms . In the B2 type of structure, the Cr and Al atomic sites 4c and 4d become equivalent due to disorder (i.e, Y–Z disorder alone). However, when the disordering is of the A2 type, some of the Cr and Al can also be at the 4a, 4b sites and Co/Fe at the 4c, 4d sites. The Mössbauer spectra of the present samples all possess two distinct sextets, one relaxation sextet and one singlet. Values of the different parameters obtained by ﬁtting the spectra are given in Table 2. Along the series, the combined relative area under the curve of sextets S1 and S2 and that of S3 increase linearly with increase in Fe content (Fig. 6). Also, the value of the hyperﬁne ﬁeld of S1 and S2 consistently increase with increase in Fe content while the average hyperﬁne ﬁeld value of the relaxation sextet S3 remains nearly the same throughout. Hyperﬁne ﬁeld values of S1 and S2 in samples with x¼ 0.4 and 0.5 are nearly the same except that the area under the curve due to the singlet which can be attributed to a nearest neighbor conﬁguration of Cr/Al alone is signiﬁcantly lower for x¼ 0.5. The Mössbauer spectra of the samples with x ¼0.4 and 0.5 are consistent with a partially ordered, B2 like structure as evidenced by XRD spectra of these samples. When X' is Fe as in the present samples, even without X–Y or X–Z disorder, variation in the actual number of X/X' as nn can give rise to different hyperﬁne ﬁelds. Thus in samples with x ¼0.4 and 0.5, S1 and S2 which have close lying values is due to slightly different number of Co/Fe as nn to a
J. Nehra et al. / Physica B 459 (2015) 46–51
Fig. 4. Band structure of Co2CrAl Heusler alloy in (a) the minority and (b) the majority spin channels. Band structure of Co1.5Fe0.5CrAl alloy in (c) the minority and (d) the majority spin channels. Table 1 Calculated atom-resolved and total spin magnetic moments per unit cell in μB for the Co2 xFexCrAl alloys (each cell contains 16 atoms). The atom-resolved spin moments have been scaled to one atom. Sample name
Co2CrAl Co1.75Fe0.25CrAl Co1.5Fe0.5CrAl
12.00 11.03 10.04
0.82 0.86 0.92
– 1.12 0.96
1.46 1.57 1.64
0.05 0.04 0.02
particular Fe throughout the sample. The relaxation spectrum is due to the superposition of sextets with low, but slightly differing hyperﬁne ﬁelds which can be attributed to Fe with varying number of Cr and Al along with some Fe/Co as nearest neighbors. The severe disordering due to randomization of Co, Fe, Cr and Al among the available sites (A2 type of ordering) in sample with x ¼0.2 is evident in the lowered ﬁeld values of sextets S1 and S2 due to occupation of Fe/Co sites by Cr/Al and large area occupied by the singlet in this sample. 3.4. DC magnetization DC magnetization measurements were recorded as a function of applied magnetic ﬁeld (M–H) and temperature (M–T) (Fig. 7).
The M–H curves at room temperature in DC magnetization studies conﬁrm that whole series is predominantly ferromagnetic in nature. At room temperature, the saturation magnetization (Ms) increases linearly with increase Fe substitution showing an increase of nearly 9 emu/g with every 5 at% increase in Fe [Fig. 8]. Ms at 20 K is lower than that at RT and as is apparent from Fig. 8, it remains nearly the same at 35 emu/g for upto x ¼0.4 after which there is an abrupt increase to 55 emu/g. Theoretically Co2CrAl is a half-metallic ferromagnetic with a moment of 3 mB/f.u. from band structure calculations. However, as pointed out by us in our earlier paper , due to the unavoidable A2 type phase separation because very low swap energy between Co–Al and Cr–Al leading to antisite atomic disorder, experimentally it is observed that the saturation moment of Co2CrAl is only 1.7 mB/f.u. and the half-metallicity is destroyed. In our earlier paper we had shown that substitution of Co with Fe followed by a proper heat treatment leads to structural stability along with preservation of the half-metallicity. In the present series, the sample with x¼ 0.2 has an Ms of 34 emu/g at RT. This value is also less than the predicted value obtained by applying Slater Pauling rules due to atomic A2 disordering as conﬁrmed by XRD studies. The saturation magnetization is signiﬁcantly increased at RT upto 55 emu/g for x¼ 0.4 (B2 ordered) sample. The x¼ 0.5 sample has saturation
J. Nehra et al. / Physica B 459 (2015) 46–51
Fig. 5. Mössbauer spectra at room temperature of Co2 xFexCrAl for x ¼0.2, 0.4, 0.5 with site ﬁtting and hyperﬁne ﬁeld (Hf) distributions of samples. Contributions to the probability distribution is given in terms of the hyperﬁne ﬁeld components. Table 2 Parameters evaluated from ﬁtted Mössbauer spectra-Isomer shift (IS) quoted with respect to Fe, width of individual lorentzians, hyperﬁne magnetic ﬁeld (HF) and relative areas occupied by subspectral components. Sample Name
IS ( 7 0.02 mm/s)
HF ( 7 1 kOe)
Area ( 71%)
S1 S2 S3 Singlet
0.01 0.01 0.08 0.01
234 152 116
11 9 58 22
S1 S2 S3 Singlet
0.01 0.01 0.08 0.02
290 263 125
10 9 65 16
S1 S2 S3 Singlet
0.02 0.04 0.09 0.03
296 274 120
10 13 70 7
magnetization of 63 emu/g (2.2 mB/f.u.) at RT, which nearly matches the Slater Pauling values. The zero ﬁeld cooled (ZFC) and ﬁeld cooled (FC) magnetization curves have been obtained as a function of temperature M(T) for all the samples in a ﬁeld of 45 Oe and are shown in Fig. 9. From Fig. 9 it is evident that the zero ﬁeld cooled (ZFC) and ﬁeld cooled (FC) curves are irreversible with the irreversible at temperatures lying above room temperature for annealed samples. Also, the presence of an anti ferromagnetic component is evident from the ZFC curve in the whole series which also show that the Neel temperature is close to room temperature because of which the saturation magnetization at 20 K is almost constant upto x¼ 0.4. The FC–ZFC curve for sample with x¼ 0.5 shows only a very small separation between the curves indicative of the weakly antiferromagnetic nature of the sample which also accounts for the fact that this sample has a much higher value of Ms. Thus the relaxation component in the Mössbauer spectra can be attributed to the antiferromagnetic component. The negative moment of Fe atoms obtained by theoretical calculations (Table 1) also indicate antiferromagnetic interactions with other atoms. As discussed
Fig. 6. Variation of the relative area under the curve with Fe substitution of sub spectral components S1, S2 and S3 obtained from RT Mössbauer measurements.
earlier, Co2CrAl is crystallized in low degree of ordering due to low swap energy in Cr–Al, Co–Cr swap. In this condition antisite Cr atom is antiferromagnetic coupled with Cr atoms at ordinary site and Co atom. The overall contribution of antiferromagnetic interactions is decreased with increase in Fe substitution at RT. These results are also corroborated by results obtained by Mössbauer studies from which we observe a steady increase in both the value and combined relative area under the curve of sextets S1 and S2. While the area under the curve for the relaxation sextet attributable to the antiferromagnetic component also increases, there is not much change in its actual value (Fig. 6). Together these results show that inter-grain ferromagnetic interactions are strengthened with an increase in Fe content. This is also reﬂected in the linear increase in the Curie temperature (Fig. 7). Also, the Curie temperature is well above RT for all samples making it useful from the application point of view.
J. Nehra et al. / Physica B 459 (2015) 46–51
Fig. 9. Zero ﬁeld cooled (ZFC) and ﬁeld cooled (FC) curves of the samples. Fig. 7. Room temperature magnetization curves of the Co2 xFexCrAl (x ¼ 0, 0.2, 0.4, 0.5) samples. Inset shows M–T curve to determine the Curie temperature.
Udaipur, India. Jagdish Nehra gratefully acknowledges ﬁnancial support of SRF from CSIR New Delhi.
Fig. 8. Variation of saturation magnetization as a function of Fe in Co2 xFexCrAl. The scattered points (black for RT values and red for values obtained at 20 K) are experimental values of Ms and the solid green line is a linear ﬁt.
4. Conclusion Investigations on the structural, electronic and magnetic properties of Co2 xFexCrAl quaternary Heusler alloy for x¼0, 0.2, 0.4, 0.5 show that the half metallicity in the parent Co2CrAl is retained in the whole series of alloys on partial substitution of Co with Fe atoms. Experimentally the saturation magnetization and Curie temperature is found to increase linearly with an increase in Fe substitution.
Acknowledgments This work is supported by DST-FIST and UGC-DSA programs of the Department of Physics, Mohanlal Sukhadia University,
 R.A. de Groot, F.M. Mueller, P.G. van Engen, K.H.J. Buschow, Phys. Rev. Lett. 50 (1983) 2024.  E. Shreder, S.V. Streltsov, A. Svyazhin, A. Makhnev, V.V. Marchenkov, A. Lukoyanov, H.W. Weber, J. Phys.: Condens. Matter 20 (2008) 045212.  V. Alijani, J. Winterlik, G.H. Fecher, S.S. Naghavi, C. Felser, Phys. Rev. B 83 (2011) 184428.  V. Alijani, S. Ouardi, G.H. Fecher, J. Winterlik, S.S. Naghavi, X. Kozina, G. Stryganyuk, C. Felser, E. Ikenaga, Y. Yamashita, S. Ueda, K. Kobayashi, Phys. Rev. B 84 (2011) 224416.  I. Galanakis, J. Phys.: Condens. Matter 16 (2004) 3089.  T.M. Nakatani, Z. Gercsi, A. Rajanikanth, Y.K. Takahashi, K. Hono, J. Phys. D: Appl. Phys. 41 (2008) 225002.  R. Weht, W.E. Pickett, Phys. Rev. B 60 (1999) 13006.  S. Ishida, S. Asano, J. Ishida, J. Phys. Soc. Jpn. 53 (1984) 2718.  K. Özdoğan, I. Galanakis, E. Şaşıoğlu, B. Aktaş, J. Phys.: Condens. Matter 18 (2006) 2905.  E. Şaşıoğlu, L.M. Sandratskii, P. Bruno, J. Phys.: Condens. Matter 17 (2005) 995.  P. Blaha, K. Schwarz, P. Sorantin, S.B. Trickey, Comput. Phys. Commun. 59 (1990) 399.  P. Perdew, S. Burke, M. Ernzerhof, Phys. Rev. Lett. 77 (1996) 3865.  Bedor Abu Alhaj, Bothina Hamad, Phys. Status Solidi B 251 (2014) 184–189. http://dx.doi.org/10.1002/pssb.201349215.  L. Vegard, Z. Cryst. 67 (1928) 239.  Z. Gercsi, K. Hono, J. Phys.: Condens. Matter 19 (2007) 326216.  H. Luo, L. Ma, Z. Zhu, G. Wu, H. Liu, J. Qu, Y. Li, Physica B 403 (2008) 1797.  B. Balke, G.H. Fecher, H.C. Kandpal, C. Felser, K. Kobayashi, E. Ikenaga, J.-J. Kim, S. Ueda, Phys. Rev. B 74 (2006) 104405.  G.H. Fecher, C. Felser, J. Phys. D: Appl. Phys. 40 (2007) 1582.  T.M. Nakatani, A. Rajanikanth, Z. Gercsi, Y.K. Takahashi, K. Inomata, K. Hono, J. Appl. Phys 102 (2007) 033916.  S.V. Karthik, A. Rajanikanth, T.M. Nakatani, Z. Gercsi, Y.K. Takahashi, K. Inomata, K. Hono, Appl. Phys. Lett. 102 (2007) 043903.  I. Galanakis, P.H. Dederichsand, N. Papanikolaou, Phys. Rev. B 66 (2002) 174429.  Tanja Graf, Frederick Casper, J.ürgen Winterlik, Benjamin Balke, Gerhard H. Fecher, Claudia Felser, Z. Anorg. Allg. Chem. 635 (2009) 976.  Jagdish Nehra, V.D. Sudheesh, N. Lakshmi, K. Venugopalan, Phys. Status Solidi RRL 7 (2013) 289–292. http://dx.doi.org/10.1002/pssr.201307057.