Effect of Ta underlayer on magnetic properties of ...

3 downloads 0 Views 2MB Size Report
Besides, the change of Heb with various PTa and tTa are related to the crystallinity of FeMn(111) layer, interfacial roughness, and also strain/stress. Correlation ...
Surface & Coatings Technology 303 (2016) 148–153

Contents lists available at ScienceDirect

Surface & Coatings Technology journal homepage: www.elsevier.com/locate/surfcoat

Effect of Ta underlayer on magnetic properties of FeMn/NiFe films H.W. Chang a,⁎, F.T. Yuan b, M.T. Chiang c, M.C. Chan c, S.C. Liou c, D.H. Wei c, S.W. Liao a, P.H. Pan a, C.R. Wang a, Lance Horng d a

Department of Applied Physics, Tunghai University, Taichung 407, Taiwan iSentek Ltd., Advanced Sensor Laboratory, New Taipei City 221, Taiwan Institute of Mechatronic Engineering, National Taipei University of Technology, Taipei 106, Taiwan d Department of Physics, National Changhua University of Education, Changhua 500, Taiwan b c

a r t i c l e

i n f o

Article history: Received 10 November 2015 Revised 26 February 2016 Accepted in revised form 15 March 2016 Available online 16 March 2016 Keywords: Exchange bias FeMn/NiFe system Ta underlayer

a b s t r a c t Effect of Ta underlayer on the magnetic properties of sputter-prepared NiFe(5 nm)/FeMn(20 nm) bilayer films have been studied. The magnetic properties of studied films are optimized by modification of working Ar pressure deposition of Ta (PTa) in the range of 2–12 mTorr and thickness of Ta (tTa) in the range of 0–25 nm. X-ray diffraction results show that the crystallinity of the FeMn(111) strongly depends on the PTa and tTa. All studied films exhibit smooth and flat surface with root-mean-square roughness below 1 nm due to deposition at RT. Large EB field (Heb) of 65–123 Oe with small coercivity (Hc) of 5–16 Oe is obtained. Besides, the change of Heb with various PTa and tTa are related to the crystallinity of FeMn(111) layer, interfacial roughness, and also strain/stress. Correlation between magnetic properties and microstructure is also discussed. This study suggests that proper Ta underlayer is crucial in the exchange bias for NiFe/FeMn bilayer system. © 2016 Elsevier B.V. All rights reserved.

1. Introduction Exchange bias (EB), characterized by a shift of hysteresis loop originated from the interaction between the ferromagnetic (FM) and antiferromagnetic (AFM) layers, has been extensively investigated due to wide applications in advanced spintronic devices and giant magnetoresistance heads [1–2]. Since most of the devices based on EB are in polycrystalline thin film form, the investigations of thin FM and AFM bilayers are important [1–2]. However, the EB field (Heb), the magnetic field strength which M-H loop is shifted, strongly depends on crystallinity, thickness, morphology, and grain size of AFM and FM layers and also the interface between AFM/FM [3–5]. Therefore, controlling the microstructure for both the AFM and FM layers, and the interface become the important issues to overcome. Due to high both exchange anisotropy and blocking temperature, FeMn/NiFe bilayer becomes a commonly used EB system [1]. However, field cooling from Neel temperature of 490 K to room temperature is (RT) required for the formation of AFM state may lead to the interdiffusion between NiFe and FeMn layers and thus degrades the EB field. Therefore, RT deposition is proposed to avoid from the intermixing in this system. In order for the formation of AFM state in FeMn layer, ⁎ Corresponding author at: Department of Applied Physics, Tunghai University, Taichung 407, Taiwan. E-mail address: [email protected] (H.W. Chang).

http://dx.doi.org/10.1016/j.surfcoat.2016.03.044 0257-8972/© 2016 Elsevier B.V. All rights reserved.

Fe50Mn50/Ni81Fe19 films are prepared on SiO2/Si(100) substrates at room temperature by sputtering at the external magnetic field of 1 kOe induced by NdFeB sintered magnets in this work. Since exchange bias is considered as an interface phenomenon, and therefore, the morphology and roughness may affect EB. The proper underlayer was reported helpful in promotion of interface quality and crystalline of AFM and FM layers [6–7]. Accordingly, Ta is adopted as an underlayer in this study in order to obtain better crystallinity FeMn and NiFe layers, and also interface. In this study, effect of Ta underlayer on the magnetic properties of FeMn(20 nm)/NiFe(5 nm) films prepared on SiO2/Si(100) substrates at room temperature (RT) by sputtering at the external magnetic field of 1 kOe induced NdFeB sintered magnets are reported.

2. Experiment FeMn(20 nm)/NiFe(5 nm)/Ta(tta nm) films with various thickness of Ta underlayer (tTa) in the range of 0–25 nm and working pressures (PAr) in the range of 2–12 mTorr were deposited on SiO2/Si(100) substrate by DC magnetron sputtering system. The adopted SiO2/Si(100) substrates have very flat surface with very low root-mean-square surface roughness (R) of below 0.2 nm, measured by an atomic force microscopy (AFM) (MS-838, Force Precision Instrument, Taiwan). The base pressure was better than 5 × 10− 7 Torr. In order to induce unidirectional

H.W. Chang et al. / Surface & Coatings Technology 303 (2016) 148–153

anisotropy, an external magnetic field of about 1 kOe, induced by high performance NdFeB sintered magnet, was applied during deposition. The structural characterization was carried out by X-ray diffractometer (XRD) (PHILIPS X'PERT Pro MPD, Netherlands) using Cu Kα radiation. Magnetic properties were measured by an alternating gradient magnetometer (AGM) (MicroMag™2900, USA). The thickness and surface morphology of the sample were measured by both AFM and scanning electron microscopy (SEM) (JEOL JSM-6500F, Japan). The microstructure was directly observed by a transmission electron microscopy (TEM) (JEOL JEM-2100, Japan). 3. Results and discussion Fig. 1(a)–(d) shows the in-plane hysteresis loops of FeMn(20 nm)/ NiFe(5 nm)/Ta (20 nm) films at various PAr of 4, 6, 8, 10 mTorr,

149

respectively. Clearly, the hysteresis loop shifts along the negative direction of the applied magnetic field and this indicates an EB for this series of samples. The exchange bias field (Heb) and coercivity (Hc) as function of PAr for Ta deposition is summarized in Fig. 2(a). Obviously, large Heb of 65–109 Oe is obtained for PAr = 2–12 mTorr. With the increase of PAr, Heb increases from 65.3 Oe for PAr = 2 mTorr to 109 Oe for PAr = 8 mTorr at first, and then decreases to 79 Oe for PAr = 12 mTorr. On the other hand, low Hc of 5–15 Oe is found for low PAr in the region of 2–8 mTorr, but Hc is largely increased to 31–39 Oe for high PAr = 9– 12 mTorr. The effect of Ta underlayer is further studied in FeMn(20 nm)/NiFe (5 nm) films with various tta. The results are shown in Fig. 1 (e)–(h). The sample without Ta underlayer shows no exchange bias. When Ta layer is induced and increased in thickness, biasing field as well as coercivity of the NiFe layer increases. The magnetic properties are summarized in

Fig. 1. Magnetic hysteresis loops of FeMn/NiFe films with working pressure of (a) 4 mTorr, (b) 6 mTorr, (c) 8 mTorr, and (d) 10 mTorr, and with Ta thickness of (e) 0 nm, (f) 5 nm, (g) 10 nm, (h) 20 nm.

150

H.W. Chang et al. / Surface & Coatings Technology 303 (2016) 148–153

Fig. 2. Dependence of Heb and Hc on working pressure (a) and thickness of Ta underlayer (b).

Fig. 4. TEM images of NiFe(5 nm)/FeMn(20 nm)/Ta(20 nm) films at Ar pressure of (a) 8 mTorr and (b) 10 mTorr.

Fig. 3. XRD patterns of NiFe(5 nm)/FeMn(20 nm) bilayer deposited on Ta underlayer with (a) various Ar pressures during Ta deposition and (b) various Ta thicknesses.

Fig. 2(b). Large Heb of 93–123 Oe is obtained for tTa = 7.5–25 nm. With the increase of tTa, Heb slightly increases from 0 Oe for tTa = 0 to 15 Oe for tTa = 5 nm, rapidly rises to 93 Oe for tta = 7.5 nm, and finally almost saturates to 110–123 Oe for tTa = 12.5–25 nm. On the other hand, low Hc of 1–19 Oe is found for tTa = 0–25 nm. Fig. 3(a) shows XRD patterns of FeMn(20 nm)/NiFe(5 nm)/Ta (20 nm) films at various PAr. Clearly, strong FeMn(111) diffraction peak found for PAr = 2–8 mTorr reveals a good crystallinity, but it is wakened for high PAr = 9–12 mTorr. It has been reported that at high Ar pressure, the reduction of mean free path and the deterioration of the ability for the surface atoms adsorbed on the substrate due to the collision of excessive background gas particles lead to the formation of more void and waviness fibrous structure [8]. The crystallinity of Ta layer is thus degenerated. Additionally, a transition in orientation from (330) to (200) + (330) is also observed with increasing working pressure. It is found that the crystal structure and preferred orientation of Ta directly influences the texture of following deposited FeMn layer. Similar results have been reported [9]. Degraded crystallinity of FeMn in the samples with Ta layer deposited at higher pressures (N8 mTorr) results in lower Heb. Besides, the peak shift of Ta(330) and FeMn(111) toward high angle with increasing PAr from 2 mTorr to 8 mTorr indicates tensile stress. This tensile stress might play important role in improving Heb. Effect of stress/strain of the underlayer on the NiFe/FeMn films needs to further study and will be studied in near future. Fig. 3(b) shows XRD patterns of FeMn(20 nm)/NiFe(5 nm)/Ta (tta) films with various tTa. For Ta-free sample, no FeMn(111) diffraction peak is found and results in the absence of EB. When 5–25 nm thick

H.W. Chang et al. / Surface & Coatings Technology 303 (2016) 148–153

Ta is inserted, FeMn(111) texture appears, and thus EB appears. The improved FeMn crystallinity, signaled by strengthened FeMn(111) peak, with increasing tta leads to the improved Heb and almost saturates to 110–123 Oe for tTa = 12.5–25 nm. Microstructure of those samples has been also studied. The continuous microstructure with fine grains of 5–15 nm is observed as shown in Fig. 4 due to RT deposition. This fine microstructure hints a flat and smooth interface between NiFe and FeMn layers. Furthermore, finer grains found for the sample at PAr = 10 mTorr than that at PAr = 8 mTorr is consistent with the XRD result. Fig. 5 (a)–(d) show SEM and AFM images of the samples at PAr = 8 mTorr and PAr = 10 mTorr. Clearly, the flat and smooth surface with

151

R well below 0.7 nm is found for this series of samples. The R as a function of PAr is summarized in Fig. 5 (e). With the increase of PAr, R is increased, especially for PAr N 8 mTorr. As mentioned earlier, at high Ar pressure, more void and waviness fibrous structure is easier to form and thus the deterioration of FeMn(111) texture and the roughened interface and surface [8–10]. R largely increases in the range of PAr from 8 to 12 mTorr. The results suggest that the decrease of Heb and the increase of Hc may be related to localized shape magnetic anisotropy in NiFe layer due to roughened interface. The broadened magnetic anisotropy increases the impedance of magnetic domain wall motion, results in higher coercivity; on the other hand, poor interfacial crystallinity degrades exchange coupling, resulting in decreased biasing field.

Fig. 5. SEM and AFM images of NiFe(5 nm)/FeMn(20 nm)/Ta(20 nm) films at Ar pressure of 8 mTorr (a) and (c) and 10 mTorr (b) and (d). (e) The roughness of NiFe(5 nm)/FeMn(20 nm)/ Ta(20 nm) films with various Ar pressure during Ta deposition.

152

H.W. Chang et al. / Surface & Coatings Technology 303 (2016) 148–153

Fig. 6. SEM and AFM images of NiFe(5 nm)/FeMn(20 nm)/Ta(t nm) films with t = 0 nm (a) and (c) and 25 nm (b) and (d). (e) The roughness of NiFe(5 nm)/FeMn(20 nm)/Ta(t nm) films with various Ta thicknesses.

Fig. 6 (a)–(d) show SEM and AFM images of the samples with tta = 0 nm and 25 nm. Similarly, the flat and smooth surface with R well below 0.5 nm is found. As shown in Fig. 6 (e), when Ta is inserted, the surface becomes quite flat and R is reduced. The reduced R with Ta might be related to morphology change [11–12]. The more discontinuous film is formed for no or thinner Ta underlayer (tTa = 0–5 nm), and the continuous film is obtained for thicker underlayer (tTa = 7.5– 25 nm). Accordingly, Ta underlayer is helpful in the crystallinity of FeMn(111) texture, the flattened surface and interface, and thus the improvement of Heb.

According to random field model [13], the compensated interface is changed to the uncompensated interface by roughness, and the interfacial roughness, determined by PTa, plays an important role in affecting the magnetic properties in this studied system. Proper Ta underlayer is effective in the formation of FeMn(111) texture. All studied films exhibit smooth and flat surface with low root-mean-square roughness below 1 nm due to deposition at RT without any field cooling. FeMn/ NiFe/Ta film with large Heb of 122 Oe and small Hc of 16 Oe is optimized by modification of PTa, tTa. This study provides useful information to fabricate exchange-bias NiFe/FeMn system.

4. Conclusions Acknowledgements Magnetic properties of sputter-prepared NiFe(5 nm)/FeMn(20 nm) bilayer films at RT with Ta underlayer were reported. The crystallinity of FeMn(111) depends on the working pressure of Ta underlayer, and the root-mean-square roughness increases with the increment of PTa.

This research was financially supported by the Ministry of Science and Technology of Taiwan, under grant No. MOST-104-2112-M-029003.

H.W. Chang et al. / Surface & Coatings Technology 303 (2016) 148–153

References [1] J. Nogues, I.K. Schuler, J. Magn. Magn. Mater. 192 (1999) 203. [2] S.A. Wolf, D.D. Awschalom, R.A. Buhrman, J.M. Daughton, S. von Molnar, M.L. Roukes, A.Y. Chtchelkanova, D.M. Treger, Science 294 (2001) 1488. [3] Susumu Soeya, Appl. Phys. Lett. 94 (2009) 242507. [4] S.K. Mishra, F. Radu, S. Valencia, D. Schmitz, E. Schierle, H.A. Durr, W. Eberhardt, Phys. Rev. B 81 (2010) 212404. [5] K. O'Grady, L.E. Fernandez-Outon, G. Vallejo-Fernandez, J. Magn., Magn. Mater. 322 (2010) 883. [6] R.M. Oksuzoglu, Christoph Schug, Brian York, J. Magn. Magn. Mater. 280 (2004) 304.

153

[7] V.K. Sankaranarayanan, Yongkang Hu, CheolGi Kim, Chong-Oh Kim, M. Tsunoda, M. Takahashi, Phys. Status Solidi A 201 (2004) 1688. [8] G.S. Bales, A. Zangwill, J. Vac. Sci. Technol. 9 (1991) 145. [9] Z.Q. Lu, W.Y. Lai, C.L. Chai, Thin Solid Films 375 (2000) 224. [10] H. Ouyang, K.W. Lin, C.C. Liu, S.C. Lo, Y.M. Tzeng, Z.Y. Guo, J. van Lierop, Phys. Rev. Lett. 98 (2007) 097204. [11] Y.K. Tang, Y. Sun, Z.H. Cheng, J. Appl. Phys. 100 (2006) 023914. [12] J.H. Hsu, A.C. Sun, P. Sharma, Thin Solid Films 542 (2013) 87. [13] A.P. Malozemoff, Phys. Rev. B 35 (1987) 3679.