WorksPhysics and Astronomy
Strain control of magneto-optical properties of ultra-thin Ni-Mn-Ga films grown by magnetron sputtering
Abstract
Ultra-thin epitaxial Ni-Mn-Ga films, 10-15 nm thick, were grown by magnetron sputtering on MgO(001), either directly on the substrate or on a Cr buffer, and studied by X-ray diffraction, magnetometry, and magneto-optical Kerr effect (MOKE) spectroscopy. The films were austenitic, with the cubic cell tetragonally distorted in opposite directions. The film on bare MgO was stretched in-plane () and the film on Cr was compressed in-plane (), while their unit-cell volumes differed by less than . As the thermal expansion mismatch was expected to stretch the films in-plane in both cases, the opposite distortions were attributed to the lattice misfit. The room-temperature magnetization loops of the two films did not show significant differences, but their polar and longitudinal MOKE spectra differed in the amplitude, sign, and energy of the spectral structures over the whole 0.76-5 eV range. Magneto-optical simulations showed that the absence of the Cr buffer accounted only for the stronger low-energy polar MOKE rotation extremum of the film on MgO. We attributed the other changes to the strain-induced change of the electronic structure, consistent with theoretical predictions in which the and sub-bands of the minority-spin Ni states exchange their order when the distortion changes sign. In parts of the spectrum, the two films showed Kerr rotation or ellipticity of opposite sign, so the choice of the growth template or a dynamically induced distortion can switch the sign of the magneto-optical response at a given photon energy.
Introduction
The magnetic shape memory alloy Ni-Mn-Ga is a ferromagnetic Heusler alloy whose cubic austenite transforms into lower-symmetry martensitic phases. Since its discovery in a single crystal [1], the magnetic-field-induced strain caused by twin-boundary motion has reached, respectively, 5.1 %, 9.5 % and 12 % in bulk single crystals of the 10M, 14M, and NM martensites [2, 3, 4]. Band-structure calculations attributed the instability of the cubic lattice to a band Jahn–Teller mechanism [5], supported by polarized neutron scattering [6]. In cubic , the states closest to the Fermi level () form a sharp minority-spin peak of degenerate Ni (, ) character, calculated about 0.2 eV below [7] or directly at [8] and observed by photoemission 0.3–0.35 eV below [9, 10].
The total energy of depends only weakly on , consistent with the wide range of observed tetragonal distortions [11]. A tetragonal distortion lifts the degeneracy of the Ni peak, and the energy order of the two sub-bands reverses with the sign of the distortion. For , the states move below and the states above [8], whereas for the peak moves up towards and the -dominated peak moves down [7]. For distortions as small as , i.e., smaller than the distortions and considered in these calculations, the occupation of the split sub-bands depends on whether the cubic peak lies below , as in the generalized gradient approximation (GGA), or at , as in quasiparticle self-consistent GW (QSGW). The effect of such small distortions on the electronic structure has not been tested experimentally.
For integration into microsystems, Ni-Mn-Ga is deposited as thin films, usually hundreds of nanometres or more thick, most commonly on MgO(001) [12, 13, 14, 15, 16]. Thomas et al. [12] showed for a 473 nm thick film that its larger thermal expansion coefficient than that of its substrate, together with a lattice misfit [14] of about 2.5% produce a biaxial tensile stress of up to 105 MPa, which stretches the austenite in-plane () and raises its martensitic transformation temperature by about 63 K. A Cr buffer has low misfit to Ni-Mn-Ga, which enables epitaxial growth without significant interdiffusion and the release of freestanding films by selective etching [17, 18]. Ultra-thin films, in which the substrate constraint acts on the whole film volume, have received much less attention. Films 10 nm and 20 nm thick on MgO(001) remain austenitic over the whole investigated temperature range [15], as do Cr-buffered films up to 38 nm thick [19]. In such films the interface may also contribute, as Cr monolayers on NiMnGa(001) were predicted to carry strongly enhanced magnetic moments [20] and interfacial spin–orbit interaction modifies the polar Kerr effect of ultra-thin Fe/GaAs(001) layers [21].
Magneto-optical Kerr effect (MOKE) spectroscopy is well suited to probe such changes, as it depends crucially on band-structure details such as spin–orbit coupling and exchange splitting [22]. In a Ni-Mn-Ga single crystal, the polar Kerr rotation reversed its sign between 2 and 3 eV upon the martensitic transformation, without a shift of the main spectral bands [23, 24]. In Cr-buffered films, the polar MOKE spectra evolved with thickness [19], but the tetragonal distortion of the austenite () changed together with the thickness [19]. The effect of a small strain-induced distortion, and in particular of its sign, on the magneto-optical spectra has thus not been isolated, nor have interface optical effects been separated from strain effects. Here we compare 10 nm and 15 nm thick films grown under identical conditions directly on MgO(001) and on a Cr buffer, with opposite distortions of the austenite, and . As the optical penetration depth (below 25 nm [19]) is comparable to the film thickness, MOKE probes the whole strained film including its interface. We show that these small distortions produce markedly different polar and longitudinal MOKE spectra and use optical simulations to separate the optical effect of the buffer from the change of the electronic structure.
Experimental
A series of ultra-thin epitaxial Ni-Mn-Ga films was prepared using magnetron sputtering at 550 °C deposition temperature and Pa vacuum on MgO(100) substrate. Two sets of films with different nominal thicknesses (10 nm and 15 nm) were prepared with and without 20 nm Cr buffer and a reference 20 nm Cr buffer without Ni-Mn-Ga film.
X-ray diffraction (XRD) and X-ray reflectivity (XRR) measurements were carried out using a Rigaku SmartLab diffractometer with a Cu anode ( nm, nm) in parallel-beam geometry. The lattice parameters were established from symmetric and asymmetric - scans of strong fundamental reflections, and the XRR data were analyzed using GenX 3 python package developed by Glavic and Björck [25].
Magnetization loops were measured at room temperature using Quantum Design PPMS up to 9 T. The linear diamagnetic contribution from the MgO substrate was subtracted from the data.
The polar magneto-optical Kerr effect (PMOKE) was measured with a custom-built rotating polarizer ellipsometer (RPE) at near-normal incidence in T applied magnetic field from ultra-violet (UV) to infra-red (IR) range, i.e., in the 0.76–5 eV spectral range. We used Ocean Optics QEPro spectrometer for the ultra-violet (UV) and visible (VIS) regions and spectrometer Ocean Optics Flame NIR for the near-infra-red (NIR) region. To eliminate magnetization-even effects, spectra were averaged over both magnetic field polarities and 10 magnetizing cycles to obtain precision of up to a few millidegrees.
Results and discussion
Nominal compositions, thicknesses, and lattice parameters of the studied films are listed in Table 1. All films grew epitaxially in the austenitic phase with tetragonally distorted cubic lattice. Film lattice grew rotated by in-plane with respect to MgO and parallel to Cr lattice, which is itself slightly tetragonally distorted (). Lattice parameters were determined for the nominally 15 nm thick films, which show that the two templates distort the cubic cell in opposite directions. On bare MgO(001), the film is stretched in-plane and shortened out-of-plane () without fully adopting the substrate lattice with residual misfit . On the Cr buffer, the film has a smaller misfit , is compressed in-plane, and elongated out-of-plane (). Remarkably, the unit-cell volumes of both films (198.3 Å and 198.8 Å) differ by less than 0.3 %, corresponding to the same equivalent cubic lattice parameter of 0.583 nm. The two films thus differ essentially only in the sign of the tetragonal distortion, and .
Table 1: Nominal composition and thickness (in nm—found in brackets following composition of each layer) of Ni-Mn-Ga epitaxial films deposited on MgO substrate were derived, respectively, from sputtering target composition and deposition times. Thicknesses , were determined by XRR, tetragonal lattice parameters , , , , and cubic were determined by XRD.
| Nominal composition | (nm) | (nm) | (nm) | (nm) | (nm) | (nm) | ||
| Ni₄₈Mn₃₀Ga₂₂(10) | 8.7(2) | – | – | – | – | – | – | – |
| Cr(20)/Ni₄₈Mn₃₀Ga₂₂(10) | 9.0(2) | 22.3(2) | – | – | – | – | – | – |
| Ni₄₈Mn₃₀Ga₂₂(15) | 12.3(2) | – | 0.5789 | 0.5853 | 0.989 | – | – | – |
| Cr(20)/Ni₄₈Mn₃₀Ga₂₂(15) | 13.2(2) | 22.3(2) | 0.5917 | 0.5797 | 1.021 | 0.2884 | 0.2892 | 0.997 |
| Cr(20) | – | 25.1(3) | – | – | – | 0.2885 | 0.2891 | 0.998 |
| Note: |
Table 1.
The distortions are set by the template during growth at 550 °C and modified by the subsequent cooling. As described by Chernenko et al. [26], a film clamped to a much thicker substrate acquires on cooling an in-plane strain , which is tensile when the linear thermal expansion coefficient of the film, , exceeds that of the substrate, . The coefficient of Ni-Mn-Ga films [26], K, exceeds those of MgO [27], K, and of Cr near room temperature [28], K below and K above its Néel temperature (found at 43.3 °C in that study). Cooling to room temperature ( K) thus stretches a Ni-Mn-Ga film clamped to MgO in-plane by up to about 0.16 %. The Cr buffer, which contracts less than MgO, is thermally compressed in-plane by about 0.2 %, but its in-plane lattice parameter still exceeds the out-of-plane one (–0.998), so the misfit to the larger MgO lattice outweighs the thermal contribution. Since the 22 nm thick buffer is expected to follow the thermal contraction of the much thicker MgO substrate, the thermal strain of the Ni-Mn-Ga film is tensile on both templates, and the opposite sign of the distortion must originate in the lattice misfit. On Cr, the doubled in-plane lattice parameter of the buffer, nm, is smaller than the equivalent cubic parameter of the film, 0.583 nm, and the film is compressed in-plane by about 0.6 %. On MgO, nm is larger, and the film, with most of the misfit relaxed, is stretched in-plane by about 0.4 %. The thermal strain accounts for less than half of this stretch, similar to the observation of Thomas et al. [12] that the measured tensile strain is about three times the calculated thermal strain.
Comparison of magnetization loops for films with and without Cr buffer did not reveal expected re-orientation of easy axis due to the substrate constraint. As seen from Figure 1, the magnetization signal is mainly dominated by diamagnetic contribution of MgO substrate. After the subtraction of this contribution the loops of both samples (with and without Cr layer) exhibit the same magnetization behavior within the given experimental accuracy.
Figure 1. Room-temperature magnetization loop of 15 nm-thick Ni-Mn-Ga film with Cr-buffer (black) decomposed into diamagnetic (red) and ferromagnetic (blue) contributions.
Although the magnetization response was effectively indistinguishable in the films, the PMOKE spectra differ strongly (Figure 2). While both PMOKE rotation spectra share a negative extremum at 0.85 eV, in most of the VIS-to-UV region, they have different spectral structures present. For , the

Figure 2. Strain controlled normal incidence PMOKE rotation (top) and ellipticity (bottom) spectra in 15-nm-thick Ni-Mn-Ga films with (solid black lines) and without (dotted black lines) Cr buffer grown with elongated or shortened the originally cubic unit cell in the out-of-plane direction (-axis) as evident from the ratio.
PMOKE rotation is dominated by a strong positive peak at 1.45 eV, followed by a weak feature near 2.4 eV and a negative extremum at 3.4 eV. For , the 1.45 eV peak collapses into a plateau between 1.5 eV and 2.3 eV, whereas the higher-energy positive extremum is enhanced and found 0.5 eV higher, at 2.9 eV, and the UV extremum moves to 4.25 eV. The PMOKE ellipticity spectra show similar trends as they are bound by Kramers-Kronig relations [29]. The nominally 10 nm and 15 nm thick films differed only in overall amplitude, which can be attributed to surface quality and probed volume, not in spectral shape.
Cr-buffered Ni-Mn-Ga films 8–160 nm thick show the same sequence of extrema in the PMOKE rotation, with negative extrema around 1 eV and 3.8 eV and two smaller positive ones around 1.7 eV and 2.6 eV, of which the former is higher in the martensitic and the latter in the austenitic films [19]. Such a phase dependence was first observed in a bulk single crystal, in which the PMOKE rotation changes its sign between 2 and 2.5 eV upon the transformation between 10M martensite and austenite [24]. In the films, however, the ratio of the two positive extrema does not depend on the phase alone, as the spectra also evolve with thickness. The films up to 38 nm thick are austenitic and compressed in-plane by the Cr buffer, with a tetragonal distortion that increases with thickness, and their PMOKE extrema change in amplitude and shift in energy by up to 0.7 eV with thickness. The 80 nm and 160 nm thick films are martensitic, but with a complex twinned microstructure in which the short axis of the martensitic cell lies in-plane and out-of-plane in roughly equal fractions. The film on Cr studied here () thus belongs to the same class of in-plane compressed austenite as the thinner films of Ref. [19], although it was grown from a different target, and its positive extremum at 2.9 eV and UV extremum at 4.25 eV lie close to those calculated for Cr-buffered films of similar thickness (solid lines in Figure 4). The film on bare MgO () is distorted in the opposite sense, which was not subject of the mentioned study, and its comparison with the twinned martensitic films, which contain both orientations of the short axis, is not straightforward. The longitudinal MOKE (LMOKE) spectra of the 10 nm thick films differ between the two samples in a similar way (Figure 3). The LMOKE rotation peak near 2 eV has opposite signs in the two films, and the higher-energy peak lies at 3.2 eV in the film on Cr and at 2.9 eV in the film on bare MgO.
Figure 3. Influence of Cr buffer on 45-deg incidence LMOKE of 10 nm Ni-Mn-Ga film with 193 mT field applied along [100].
Figure 4. Normal incidence PMOKE rotation (top) and ellipticity (bottom) spectra of 16-nm thick Ni-Mn-Ga film on 28-nm thick Cr buffer (solid lines; data were taken from Ref. 19) and a simulated PMOKE spectra (dashed lines) of the same film with the Cr buffer absent.
Since the optical penetration depth is comparable to the film thickness, the layer beneath the film also affects the measured spectra by optical means, through the reflection at the film/substrate interface. To estimate this effect, we simulated the removal of the Cr buffer from 16 nm Ni-Mn-Ga film from Ref. 19 using the Yeh matrix formalism [30]. The model consisted of a magnetized Ni-Mn-Ga layer, described by the diagonal () and off-diagonal () elements of its permittivity tensor, an isotropic Cr layer, and a semi-infinite MgO substrate, at normal incidence in the 0.76–4.6 eV photon energy range. We used the permittivity tensors of the 16 nm Ni-Mn-Ga film with 28 nm thick Cr buffer from our previous study [19], whose was determined from PMOKE with the same layer model [19], together with the permittivities of a Cr layer and MgO obtained on calibration samples. Results of the simulations (Figure 4) suggest that removing the buffer increases the amplitudes and sign changes for some photon ranges, but the overall spectral shape is somewhat conserved. The optical effect of the buffer could explain the stronger low-energy rotation extremum of the film on bare MgO, but not its other spectral changes, which we therefore attribute mainly to the change of the electronic structure of the film caused by the opposite strain.
Near , the tetragonal distortion acts mainly on the minority-spin Ni states of symmetry, which in the tetragonal cell separate into () and () sub-bands [8]. GGA calculations describe the qualitative changes of these sub-bands with . Ayuela et al. [31] showed that the splitting is mainly due to Ni and proportional to the deformation, with little change of the Mn bands near for , and that at their calculated minima, and , the lower sub-band has character for and character for . The same order was found by Barman et al. [7] at and by Obata et al. [8] at . The position of these sub-bands relative to , however, is not reliable in GGA, which places the cubic peak about eV below its position at obtained with QSGW [8]. In the GGA calculations of Barman et al., the upper sub-band therefore reaches only at or , whereas with the peak at , as in QSGW, even the small splitting in our films is expected to shift the upper sub-band partly above and the lower one further below. The empty minority-spin Ni states just above , which can act as final states of optical transitions, would then have character in the film with and character in the film with . A change of the unit-cell volume, which in the same GGA calculations shifts the minority-spin Ni peak relative to , plays no role here, as the volumes of the two films are practically equal. The positive rotation peak at eV, the most prominent feature of the film with , lies in the region below eV, where the magneto-optical activity of Ni-Mn-Ga has been tentatively attributed to transitions involving Ni states [23]. The spectral changes, however, extend over the whole measured range. Since the PMOKE spectra depend on both the diagonal and the off-diagonal permittivity, assigning individual spectral features to electronic transitions would require the permittivity tensors of both films and magneto-optical calculations for small distortions of both signs, which have not been reported.
In summary, 10–15 nm thick epitaxial Ni-Mn-Ga films grown on bare MgO(001) and on a Cr buffer are tetragonally distorted in opposite directions, with and 1.021, while their unit-cell volumes and room-temperature magnetization loops are practically the same. Despite these small distortions, their PMOKE spectra differ significantly over the whole 0.76–5 eV range in the amplitude, sign, and energy of the spectral features. Our optical simulations showed that the absence of the Cr buffer explains the stronger low-energy rotation extremum of the film on MgO but not the other changes, which we attribute to the strain-induced change in the electronic structure of the film. These changes are consistent with the calculated splitting of the minority-spin Ni states, in which the and sub-bands exchange their order when the sign of the distortion changes [31]. Magneto-optical spectroscopy thus detects strain-induced changes of the electronic structure in films only about 10 nm thick, even where magnetometry shows no difference, and the choice of the buffer layer may be used to modify the magneto-optical response of such ultra-thin films. In the rotation between about 2.5 and 3.3 eV and in the ellipticity between about 3.4 and 4.6 eV, the two films show MOKE of opposite sign, so that the choice of the template can switch the sign of the magneto-optical response at a given photon energy, which may be of interest for magneto-optical devices.
Declaration of competing interest
The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.
Data availability
The data that support the findings of this study are openly available in Zenodo, reference number 32.
Declaration of generative AI and AI-assisted technologies in the manuscript preparation process
During the preparation of this work, the authors used Claude Opus 5.5 (Anthropic) to improve the readability and structure of the text, to check the cited statements against the original publications, and for literature research. The authors reviewed and edited all output of this tool and take full responsibility for the content of the publication.
Acknowledgements
This work was supported by Czech Science Foundation Grant No 24-11361S, OPJAK - Ferroic multifunctionalities (FerrMion) CZ.02.01.01/00/22_008/0004591, and the Charles University grant SVV–2025–260836.
References
- [1]K. Ullakko, J. K. Huang, C. Kantner, R. C. O’Handley, V. V. Kokorin, Large magnetic-field-induced strains in Ni₂MnGa single crystals, Applied Physics Letters 69 (1996) 1966–1968. doi:10.1063/1.117637.DOI
- [2]O. Heczko, A. Sozinov, K. Ullakko, Giant field-induced reversible strain in magnetic shape memory NiMnGa alloy, IEEE Transactions on Magnetics 36 (2000) 3266–3268. doi:10.1109/20.908764.DOI
- [3]A. Sozinov, A. A. Likhachev, N. Lanska, K. Ullakko, Giant magnetic-field-induced strain in NiMnGa seven-layered martensitic phase, Applied Physics Letters 80 (2002) 1746–1748. doi:10.1063/1.1458075.DOI
- [4]A. Sozinov, N. Lanska, A. Soroka, W. Zou, 12% magnetic field-induced strain in Ni-Mn-Ga-based non-modulated martensite, Applied Physics Letters 102 (2013). doi:10.1063/1.4775677.DOI
- [5]S. Fujii, S. Ishida, S. Asano, Electronic structure and lattice transformation in Ni₂MnGa and Co₂NbSn, Journal of the Physical Society of Japan 58 (1989) 3657–3665. doi:10.1143/JPSJ.58.3657.DOI
- [6]P. J. Brown, A. Y. Bargawi, J. Crangle, K.-U. Neumann, K. R. A. Ziebeck, Direct observation of a band Jahn-Teller effect in the martensitic phase transition of Ni₂MnGa, Journal of Physics: Condensed Matter 11 (1999) 4715–4722. doi:10.1088/0953-8984/11/24/312.DOI
- [7]S. R. Barman, S. Banik, A. Chakrabarti, Structural and electronic properties of Ni₂MnGa, Physical Review B 72 (2005) 184410. doi:10.1103/PhysRevB.72.184410.DOI
- [8]M. Obata, T. Kotani, T. Oda, Intrinsic instability to martensite phases in ferromagnetic shape memory alloy Ni₂MnGa: Quasiparticle self-consistent GW investigation, Physical Review Materials 7 (2023) 024413. doi:10.1103/PhysRevMaterials.7.024413.DOI
- [9]C. P. Opeil, B. Mihaila, R. K. Schulze, L. Mañosa, A. Planes, W. L. Hults, R. A. Fisher, P. S. Riseborough, P. B. Littlewood, J. L. Smith, J. C. Lashley, Combined experimental and theoretical investigation of the premartensitic transition in Ni₂MnGa, Physical Review Letters 100 (2008) 165703. doi:10.1103/PhysRevLett.100.165703.DOI
- [10]A. Kimura, M. Ye, M. Taniguchi, E. Ikenaga, J. M. Barandiarán, V. A. Chernenko, Lattice instability of Ni-Mn-Ga ferromagnetic shape memory alloys probed by hard X-ray photoelectron spectroscopy, Applied Physics Letters 103 (2013). doi:10.1063/1.4818575.DOI
- [11]V. V. Godlevsky, K. M. Rabe, Soft tetragonal distortions in ferromagnetic Ni₂MnGa and related materials from first principles, Physical Review B 63 (2001) 134407. doi:10.1103/PhysRevB.63.134407.DOI
- [12]M. Thomas, O. Heczko, J. Buschbeck, L. Schultz, S. Fähler, Stress induced martensite in epitaxial Ni-Mn-Ga films deposited on MgO(001), Applied Physics Letters 92 (2008). doi:10.1063/1.2931082.DOI
- [13]J. Buschbeck, R. Niemann, O. Heczko, M. Thomas, L. Schultz, S. Fähler, In situ studies of the martensitic transformation in epitaxial Ni–Mn–Ga films, Acta Materialia 57 (2009) 2516–2526. doi:10.1016/j.actamat.2009.02.008.DOI
- [14]Y. Luo, P. Leicht, A. Laptev, M. Fonin, U. Rüdiger, M. Laufenberg, K. Samwer, Effects of film thickness and composition on the structure and martensitic transition of epitaxial off-stoichiometric Ni–Mn–Ga magnetic shape memory films, New Journal of Physics 13 (2011) 013042. doi:10.1088/1367-2630/13/1/013042.DOI
- [15]P. Ranzieri, S. Fabbrici, L. Nasi, L. Righi, F. Casoli, V. A. Chernenko, E. Villa, F. Albertini, Epitaxial Ni–Mn–Ga/MgO(100) thin films ranging in thickness from 10 to 100nm, Acta Materialia 61 (2013) 263–272. doi:10.1016/j.actamat.2012.09.056.DOI
- [16]A. M. Jakob, M. Hennes, M. Müller, D. Spemann, S. G. Mayr, Coupling of micromagnetic and structural properties across the martensite and Curie temperatures in miniaturized Ni-Mn-Ga ferromagnetic shape memory alloys, Advanced Functional Materials 23 (2013) 4694–4702. doi:10.1002/adfm.201300165.DOI
- [17]A. Backen, S. R. Yeduru, M. Kohl, S. Baunack, A. Diestel, B. Holzapfel, L. Schultz, S. Fähler, Comparing properties of substrate-constrained and freestanding epitaxial Ni-Mn-Ga films, Acta Materialia 58 (2010) 3415–3421. doi:10.1016/j.actamat.2010.02.016.DOI
- [18]T. Eichhorn, R. Hausmanns, G. Jakob, Microstructure of freestanding single-crystalline Ni₂MnGa thin films, Acta Materialia 59 (2011) 5067–5073. doi:10.1016/j.actamat.2011.04.032.DOI
- [19]M. Makeš, J. Zázvorka, M. Hubert, P. Vërtát, M. Rameš, J. Zemen, O. Heczko, M. Veis, Stress-induced martensitic transformation in epitaxial Ni-Mn-Ga thin films and its correlation to optical and magneto-optical properties, Physical Review B 113 (2026) 104420. doi:10.1103/b3yr-1yt8.DOI
- [20]J. Bhattacharya, D. Pandey, S. R. Barman, A. Chakrabarti, Surface termination and thickness dependent magnetic coupling of Cr adlayers on Ni₂MnGa(001) surfaces: An ab initio study, Journal of Magnetism and Magnetic Materials 540 (2021) 168398. doi:10.1016/j.jmmm.2021.168398.DOI
- [21]M. Buchner, P. Högl, S. Putz, M. Gmitra, S. Günther, M. Schoen, M. Kronseder, D. Schuh, D. Bougeard, J. Fabian, C. Back, Anisotropic polar magneto-optic Kerr effect of ultrathin Fe/GaAs(001) layers due to interfacial spin-orbit interaction, Physical Review Letters 117 (2016) 157202. doi:10.1103/PhysRevLett.117.157202.DOI
- [22]P. M. Oppeneer, T. Maurer, J. Sticht, J. Kübler, Ab initio calculated magneto-optical Kerr effect of ferromagnetic metals: Fe and Ni, Physical Review B 45 (1992) 10924–10933. doi:10.1103/PhysRevB.45.10924.DOI
- [23]M. Veis, L. Beran, M. Zahradnik, R. Antos, L. Straka, J. Kopecek, L. Fekete, O. Heczko, Magneto-optical spectroscopy of ferromagnetic shape-memory Ni-Mn-Ga alloy, Journal of Applied Physics 115 (2014). doi:10.1063/1.4867754.DOI
- [24]L. Beran, P. Cejpek, M. Kulda, R. Antos, V. Holy, M. Veis, L. Straka, O. Heczko, Optical and magneto-optical studies of martensitic transformation in Ni-Mn-Ga magnetic shape memory alloys, Journal of Applied Physics 117 (2015). doi:10.1063/1.4917059.DOI
- [25]A. Glavic, M. Björck, GenX 3: the latest generation of an established tool, Journal of Applied Crystallography 55 (2022) 1063–1071. doi:10.1107/S1600576722006653.DOI
- [26]V. A. Chernenko, M. Kohl, M. Ohtsuka, T. Takagi, V. A. L’vov, V. M. Knjazkyi, Thickness dependence of transformation characteristics of Ni–Mn–Ga thin films deposited on alumina: Experiment and modeling, Materials Science and Engineering: A 438–440 (2006) 944–947. doi:10.1016/j.msea.2006.02.055.DOI
- [27]Y. S. Touloukian, R. K. Kirby, R. E. Taylor, T. Y. R. Lee, Thermal Expansion: Nonmetallic Solids, volume 13 of Thermophysical Properties of Matter, IFI/Plenum, New York, 1977. doi:10.1007/978-1-4757-1631-3.DOI
- [28]C. N. Koumelis, The thermal expansion coefficient of chromium in the temperature region of 3 to 80 °C, physica status solidi (a) 19 (1973) K65–K69. doi:10.1002/pssa.2210190157.DOI
- [29]J. Levallois, I. O. Nedoliuk, I. Crassee, A. B. Kuzmenko, Magneto-optical Kramers-Kronig analysis, Review of Scientific Instruments 86 (2015). doi:10.1063/1.4914846.DOI
- [30]P. Yeh, Optics of anisotropic layered media: A new 4 × 4 matrix algebra, Surface Science 96 (1980) 41–53. doi:10.1016/0039-6028(80)90293-9.DOI
- [31]A. Ayuela, J. Enkovaara, R. M. Nieminen, Ab initio study of tetragonal variants in Ni₂MnGa alloy, Journal of Physics: Condensed Matter 14 (2002) 5325–5336. doi:10.1088/0953-8984/14/21/307.DOI


