Oblique angle deposition-induced anisotropy in Co2FeAl films

Oblique angle deposition-induced anisotropy in Co2FeAl films

Journal of Magnetism and Magnetic Materials 456 (2018) 353–357 Contents lists available at ScienceDirect Journal of Magnetism and Magnetic Materials...

2MB Sizes 0 Downloads 24 Views

Journal of Magnetism and Magnetic Materials 456 (2018) 353–357

Contents lists available at ScienceDirect

Journal of Magnetism and Magnetic Materials journal homepage: www.elsevier.com/locate/jmmm

Oblique angle deposition-induced anisotropy in Co2FeAl films W. Zhou, J. Brock, M. Khan, K.F. Eid ⇑ Department of Physics, Miami University, Oxford, OH 45056, United States

a r t i c l e

i n f o

Article history: Received 23 November 2017 Received in revised form 14 February 2018 Accepted 17 February 2018 Available online 18 February 2018 Keywords: Oblique angle deposition Anisotropy Heusler alloy

a b s t r a c t A series of Co2FeAl Heusler alloy films, fabricated on Si/SiO2 substrates by magnetron sputtering-oblique angle deposition technique, have been investigated by magnetization and transport measurements. The morphology and magnetic anisotropy of the films strongly depended on the deposition angle. While the film deposited at zero degree (i.e. normal incidence) did not show any anisotropy, the films deposited at higher angles showed unusually strong in-plane anisotropy that increased with deposition angle. The enhanced anisotropy was well-reflected in the direction-dependent magnetization and the coercivity of the films that increased dramatically from 30 Oe to 490 Oe. In a similar vein, the electrical resistivity of the films also increased drastically, especially for deposition angles larger than 60°. These anisotropic effects and their relation to the morphology of the films are discussed. Ó 2018 Elsevier B.V. All rights reserved.

1. Introduction From both fundamental and applied points of view, the Heusler class of intermetallic alloys have been the subject of intense research interest, since their high spin polarization makes them candidate materials for spintronic devices. [1,2] Co2FeAl (CFA) is one of the most attractive Heusler alloys due to its high Curie temperature (Tc = 1000 K) [3], relatively low magnetic damping parameter (around 0.001) [4], and high spin polarization. Because of these traits, CFA has been identified for potential use as an alternative magnetic electrode material in data storage architectures such as magnetic tunnel junctions and current-perpendicular-toplane (CPP) spin valves. CFA/MgO/CoFe magnetic tunnel junctions have been shown to possess an immense tunneling magnetoresistance (up to 360%) at room temperature [5], further fueling research interest in CFA. CFA films (thickness ranging from 10 nm to 100 nm) grown under standard conditions on Si and MgO substrates typically exhibit small in-plane coercivity (Hc < 50 Oe). [6,7] Recently, Niesen et al. showed that TiN buffered ultrathin (1 nm) CFA/MgO bilayer films exhibit Hc of as large as  320 Oe. [8] Li et al. showed that CFA films fabricated by cosputtering with Tb exhibit significantly large Hc of 800 Oe. [9] The large Hc of these ultrathin CFA films are generally associated with their perpendicular-to-plane magnetic anisotropy. These experimental results (along with many others, reported in literature) suggest that the magnetic anisotropy of

⇑ Corresponding author. E-mail address: [email protected] (K.F. Eid). https://doi.org/10.1016/j.jmmm.2018.02.056 0304-8853/Ó 2018 Elsevier B.V. All rights reserved.

CFA films can be controlled by manipulating their thickness and fabrication conditions [10–12]. A fabrication technique by which uniaxial magnetic anisotropy can be introduced in thin films is oblique angle deposition (OAD) [13–16]. The technique is also known as glancing angle deposition (GLAD). [17] In this technique the deposition flux approaches a stationary substrate at an angle a relative to the substrate normal, referred to as the incidence angle. Due to atomic shadowing effects [18], tilted nano-columnar structures are formed which exhibit properties in severe contrast to those of thin films prepared with the deposition normal to the substrate. Besides magnetic anisotropy, these nano-columnar structures possess a higher surfacearea-to-volume ratio, which can have a dramatic impact on the electronic transport [19–21] and optical properties [21,22]. The significant changes in physical properties introduced by OAD are relevant to a variety of data storage and sensing applications. Up-to-date, no report has been made of CFA films deposited by OAD technique. Considering the fact that the magnetic anisotropy of these films can be controlled dramatically by fabrication conditions, it is interesting to study OAD thin films of CFA. Therefore, in this article we report on the magnetic and electrical resistivity properties of CFA thin films prepared by magnetron-sputtering OAD (MS-OAD). A significant enhancement of uniaxial anisotropy has been observed in the films due to OAD. 2. Experiment To allow for the simultaneous growth of several CFA films at different incidence angles, a plastic block (Fig. 1a) consisting of chiseled surfaces with tilt angles ranging from 30° to 85° at 5°

354

W. Zhou et al. / Journal of Magnetism and Magnetic Materials 456 (2018) 353–357

Fig. 1. (a) Diagram of the sample holder design showing various tilted surfaces. Six planes with distinct tilt angles were fabricated on each of the two long sides of the holder. The oblique incidence angle a of the left side surfaces, from top to bottom, was 85°, 30°, 75°, 40°, 60°, and 50°. Similarly, the incidence angle a of the right side surfaces was 35°, 80°, 45°, 70°, 55°, and 65°. (b) Schematic view of MS-OAD (vertical distance not to scale). (c) Cartoon showing the nano-pillar mode of growth in OAD. The coordinate convention employed in the report is also indicated. (d) Image of a microstructure used for resistivity measurement. The length and the width of the thin channels are 400 lm and 27 lm respectively. The small wedge shape on the top rectangular extension pad marks the direction of the in-plane projection of the CFA flux.

intervals was fabricated using a 3-D printer. Fig. 1(b) is a schematic diagram of the MS-OAD setup employed, consisting of a 1 in. sputtering gun with the sample holder 20 cm above. Clean SiO2-coated Si substrates (1000 nm SiO2) with approximate dimensions of 0.5 cm  1 cm were mounted on the angled surfaces depicted in Fig. 1(a). The base pressure of the deposition chamber was 2 10 7 mbar and the CFA films were sputtered at room temperature under a 310 3 mbar partial Ar pressure. CFA was deposited on the substrates at a rate of about 0.17 Å/s, as determined by a quartz crystal monitor. Two in-plane directions were chosen for anisotropy measurements: One direction parallel to the projection of the CFA flux on to the plane of the substrates (henceforth referred to as the parallel direction or x-axis), and the other perpendicular to the projection (henceforth referred to as the y-axis), as shown in Fig. 1(c). Non-patterned thin films were used for magnetization measurements, while photolithographic techniques were utilized to create cross-shaped micro-channels for one batch of substrates to be used in anisotropic measurements of the electrical resistivity, as depicted in Fig. 1(d). These micro-channels were mounted in the holder such that one set of channels was along the previouslydefined x-axis and the other was along the y-axis (i.e. the channel sets were perpendicular to each other). The morphological characteristics of the films were assessed using a Zeiss Supra 35 VP FEG SEM, using the in-lens secondary electron detector. Elemental quantification was provided using a Bruker Quantax energy-dispersive X-ray spectroscopy (EDS) unit housed within the SEM. The static magnetic properties were determined using the vibrating sample magnetometer (VSM) option of a Quantum

Design Physical Property Measurement System (PPMS), in applied fields up to 1 kOe in strength. The room temperature electronic transport properties were probed using the four-probe resistivity module of the PPMS. The thickness of the micro-channels was measured using a Nanosurf easyScan 2 atomic force microscope (AFM).

3. Results and discussions SEM images of films deposited with a = 0° and 85° are shown in Fig. 2(a) and (b), respectively. From these images, it is clear that while the grains form a tight pattern on the normally-deposited (a = 0°) film, the grains on the surface of the OAD-prepared films are more discrete. In the sample fabricated with a = 85°, there exist gap intervals between grains, evidenced by the darker areas in Fig. 2b. These intervals extend in the y-direction, indicating that columns have a greater chance of being in contact with other columns along the y-direction than being in contact with columns along the x-direction. Furthermore, the surface of the normallydeposited film appears smoother than the film deposited with a = 85°. EDS analysis of the CFA layer rendered elemental quantifications matching the target stoichiometry. Fig. 3a–c show the magnetization as function of applied magnetic field [M(H)] curves, collected under the zero-field cooling protocol (ZFC) at 10 K, for the samples deposited at a = 0°, 40°, and 70°, respectively. For each sample, the magnetization measured at each temperature M has been normalized to the saturation magnetization MS. For the normally-deposited sample (a = 0°), M

W. Zhou et al. / Journal of Magnetism and Magnetic Materials 456 (2018) 353–357

355

Fig. 2. (a) The surface of the normal deposited (a = 0°) CFA films, (b) the surface of the oblique deposited CFA films with incident angle 85⁰.

(H) curves collected along the in-plane parallel (x) and in-plane perpendicular (y) directions overlap with each other, demonstrating isotropic magnetic behavior. As the deposition angle increases, interesting behaviors are observed in the M(H) data. For the films deposited with 0° < a  70°, the easy axis of magnetization shifts to the in-plane perpendicular axis (y). As shown in Fig. 3d, the coercivity (Hc) along the in-plane perpendicular (y) direction increases from about 30 Oe to 400 Oe as a increases from 0° to 70° at 10 K. However, at room temperature, Hc along the in-plane perpendicular (y) direction increases from about 25 Oe to 210 Oe as a increases from 0° to 70°. It should be noted that the typically-reported uniaxial coercivity values for CFA films are well below 100 Oe, which is reasonable given that this material exhibits a cubic structure. [7] Therefore, the large coercivity of 400 Oe (200 Oe at room temperature) for the OAD prepared CFA films is

a significant improvement, considering the fact that this large coercivity is achieved by simply depositing on SiO2-coated Si substrates. This demonstrates that OAD could be an efficient method of inducing strong anisotropy in magnetically isotropic systems. An interesting behavior is observed in the M(H) data for the films deposited at a  80° (see Fig. 4). Unlike the films grown at a  70° where the easy-axis is along the in-plane perpendicular (y) direction (see Fig. 3), the easy-axis of the films with a  80° appears to be along the in-plane parallel direction (x). The coercive fields of these films are also larger than that of the films with a  70° (Hc = 470 Oe and 490 Oe for a = 80° and 85°, respectively). This change in the magnetic anisotropy with a has been observed in other OAD systems, and can likely be attributed to the increased structural anisotropy associated with heightened metal-shadowing effects, as the columnar growth extends more towards the deposition flux

Fig. 3. Magnetization as a function of applied field, M(H), obtained at 10 K for the CFA films deposited at oblique angles (a) a = 0°, (b) a = 40°, (c) a = 70°. Fig. 3d shows the coercive field, HC, at 10 K and 298 K, as a function of deposition angle.

356

W. Zhou et al. / Journal of Magnetism and Magnetic Materials 456 (2018) 353–357

Fig. 5. Room temperature electrical resistivity in the x and y directions as a function of a. While both resistivities increase sharply with a above 60° to 70°, the xdirection resistivity is higher due to the shadowing effect in OAD. The film thickness as a function of incidence angle a is shown as an inset.

Fig. 4. Magnetization as a function of applied field, M(H), obtained at 10 K for the CFA films deposited at oblique angles (a) a = 80° and (b) a = 85°.

direction. [7,23,24] The M(H) data collected at 298 K exhibits similar behavior, albeit with different coercivity. The a-dependence of the resistivity at room temperature is shown in Fig. 5. Much like the magnetization, the resistivity is nearly congruent along the x- and y- directions (i.e. isotropic) when a = 0°. When a – 0° a disparity between resistivity in the x and y directions appears. The resistivity in the parallel direction (x) is always larger than in the perpendicular direction (y). More importantly, the resistivity increases along both directions as the incidence angle a increases. This rise in resistivity is highly nonlinear, with a sharp increase for angles above 60°, such that the sample with a = 80° exhibits resistivity values two orders of magnitude higher than those observed at low a. The resistivity becomes infinite for larger angles of deposition, suggesting a lack of connectivity between the now discrete nano-pillars. This poses a significant challenge for experiments that use shadowing effects to grow nanostructures [25], as the samples become insulating, arresting any current flow. This observed rise in resistivity with incidence angle is consistent with behavior reported in the literature for other materials deposited using the OAD technique. [14,26] The variation in resistivity with incidence angle seems to be a global property, with the same trend independent of the chemical identity of the film. This suggests that charge conduction across these films is dominated by the connectedness of the nanopillar structures, as opposed to more intrinsic material properties. The resistivity measurements suggest that as the angle of incidence increases, the nano-pillars gradually become more and more

disconnected until such a point where they are completely separated, unable to facilitate conduction across the film. It is also apparent that the resistivity in the parallel direction increases faster than the resistivity in the perpendicular direction. The inset of Fig. 5 shows the film thickness t versus oblique angle a, as measured by AFM. The thickness obviously decreases with the oblique angle a, but the dependence does not exactly follow the Law of Cosines, which is expected for thermal evaporation deposition techniques. [27] This deviation is due to the Ar partial pressure that was present during the sputtering process, which leads to scattering of the deposited atoms as they move between the target and substrate. This scattering causes the atoms to strike the substrate surface at a distribution of angles, rather than one well-defined angle (i.e. the material flux formed by sputtering is not collimated). Two possible conditions that will create magnetic anisotropy in a thin film are an anisotropic crystalline structure or an anisotropic morphology. While previous studies have shown that the crystalline structure assumed by CFA films on MgO substrates can produce magnetic anisotropy, [25] similar reasoning cannot explain the magnetic anisotropy of CFA films deposited at oblique angles on a SiO2 substrate, since the in-plane static magnetic properties are isotropic. The magnetic anisotropy observed in the OADfabricated CFA films discussed here must arise from the asymmetric morphology produced by the shadowing effect, with respect to the in-plane parallel and in-plane perpendicular directions. It can be deduced from the SEM images that the major axis of columnar growth tilts toward the direction of the incident flux and does not incline in the direction perpendicular to the incident flux direction [28]. This asymmetry in growth also gives rise to anisotropic behavior in the electrical resistivity, as the shadowing effect produces more voids (i.e. less connectivity) in the x-direction than in the y-direction; this explains why the in-plane resistivity across the parallel direction is larger than the in-plane resistivity across the perpendicular direction. As the oblique angle a increases, the increased prevalence of voids and the decreased width of the tilted columns explain the increased in-plane resistivity along both directions. It is worth noting that the thickness of the MS-OAD CFA films varies with the oblique angle a, but that all films are

W. Zhou et al. / Journal of Magnetism and Magnetic Materials 456 (2018) 353–357

thicker than 50 nm. Previous reports have shown that the thickness of OAD-prepared films has a minimal impact on coercivity (Hc) but a significant impact on the saturation magnetization Ms. [29] Additionally, the extant literature has also shown that the in-plane resistivity is inversely proportional to film thickness, since shorter tilted columns are less likely to form contacts with neighboring columns. In light of this discussion, the resistivity behavior demonstrated in Fig. 5 is most likely a function of both incident flux angle and film thickness. 4. Conclusions In summary, we have studied the magnetization and electrical resistivity of CFA films prepared by MS-OAD. The experimental results showed that due to OAD angle-dependent porosity and anisotropy were introduced in the films and enhanced with increasing incidence flux angle. The anisotropy of the films resulted in a significant increase of their coercivity, which was observed in the magnetization data. A largest Hc of 490 Oe was observed at 10 K for the film deposited at an OAD angle of 85°. This is quite large considering the thickness (greater than 50 nm) of the films and that no buffer layer was used. The in-plane resistivity of these films increased dramatically as the incident angle exceeded 60°. Furthermore, this in-plane resistivity is generally larger in the parallel direction (to the flux of incident atoms during growth) than the perpendicular direction for all incident flux angles. The results show that OAD is a promising technique to introduce enhanced anisotropy in thin films. Acknowledgement The SEM photos were collected by Matt L. Duley at the Miami University Center for Advanced Microscopy & Imaging, Oxford, OH 45056 USA.

357

References [1] R.A. de Groot, F.M. Mueller, P.G. van Engen, K.H.J. Buschow, Phys. Rev. Lett. 50 (1983) 2024. [2] T. Graf, C. Felser, S.P. Parkin, Prog. Solid State Chem. 39 (2011) 1. [3] S. Trudel, O. Gaier, J. Hamrle, B. Hillebrands, J. Phys. D43 (2010) 193001. [4] S. Mizukami, D. Watanabe, M. Oogane, Y. Ando, Y. Miura, M. Shirai, T. Miyazaki, J. Appl. Phys. 105 (2009) 07306. [5] W.H. Wang, H. Sukegawa, K. Inomata, Phys. Rev. B 82 (2010) 092402. [6] M. Belmeguenai, H. Tuzcuoglu, M.S. Gabor, T. Petrisor, C. Tiusan, D. Berling, F. Zighem, S.M. Mourad, Chérif, J. Magn. Magn. Mater. 373 (2015) 140. [7] M. Belmeguenai, H. Tuzcuoglu, M.S. Gabor, T. Petrisor Jr., C. Tiusan, D. Berling, F. Zighem, T. Chauveau, S.M. Chérif, P. Moch, Phys. Rev. B 87 (2013) 184431. [8] A. Niesen, J. Ludwig, M. Glas, R. Silber, Jan-Michael Schmalhorst, E. Arenholz, and G. Reiss, J. Appl. Phys. 121 (2017) 223902. [9] X.Q. Li, X.G. Xu, D.L. Zhang, J. Miao, Q. Zhan, M.B.A. Jalil, G.H. Yu, Y. Jiang, Appl. Phys. Lett. 96 (2010) 142505. [10] A. Hirohata, W. Frost, M. Samiepour, J.Y. Kim, Materials 11 (2018) 105. [11] Z. Wen, H. Sukegawa, T. Seki, T. Kubota, K. Takanashi, S. Mitani, Scient. Rep. 7 (2017) 45026. [12] X. Li, S. Yin, Y. Liu, D. Zhang, X. Xu, J. Miao, Y. Jiang, Appl. Phys. Express 4 (2011) 043006. [13] T.G. Knorr, R.W. Hoffman, Phys. Rev. 113 (1959) 1039. [14] D.O. Smith, M.S. Cohen, G.P. Weiss, J. Appl. Phys. 31 (1960) 1755. [15] A. Lisfi, J.C. Lodder, Phys. Rev. B 63 (2001) 1744. [16] M.J. Hadley, R.J. Pollard, J. Appl. Phys. 92 (2002) 7389. [17] K. Robbie et al., J. Vac. Sci. Technol. A 15 (1997) 1460. [18] A.G. Dirks, H.J. Leamy, Thin Solid Films 47 (1977) 219. [19] K. Kuwahara, Japan, J. Appl. Phys. 13 (1974) 7. [20] K. Kuwahara, Thin Solid Films 164 (1988) 165. [21] H. Savaloni, N. Abbaszadeh, J. Electron. Mater. 45 (2016) 3343. [22] D. Schmidt, A.C. Kjerstad, T. Hofmann, R. Skomski, E. Schubert, M. Schubert, J. Appl. Phys. 105 (2009) 113508. [23] C. Quirós, L. Peverini, J. Diaz, A. Alija, C. Blanco, M. Velez, O. Robach, E. Ziegler, J. M. Alameda, Nanotechnology 25 (2014) 335704. [24] C. Li, G. Chai, C. Yang, W. Wang, D. Xue, Sci. Rep. 5 (2015) 17023. [25] J. Chu, X. Peng, M. Sajjad, B. Yang, P.X. Feng, Thin Solid Films 520 (2012) 3493. [26] Y. Zhong, Y.C. Shin, C.M. Kim, B.G. Lee, E.H. Kim, Y.J. Park, K.M.A. Sobahan, C.K. Hwangbo, Y.P. Lee, T.G. Kim, J. Mater. Res. 23 (2008) 2500. [27] A. Siad, A. Besnard, C. Nouveau, P. Jacquet, Vacuum 131 (2016) 305. [28] A. Barranco, A. Borras, A.R. Gonzalez-Elipe, A. Palmero, Prog. Mat. Sci. 76 (2016) 59. [29] L. Shan-Dong et al., Chin. Phys. B 23 (2014) 10.