- Email: [email protected]
Magnetoresistance in silicon based ferrite magnetic tunnel junction
Current Applied Physics 14 (2014) 259e263
Contents lists available at ScienceDirect
Current Applied Physics journal homepage: www.elsevier.com/locate/cap
Magnetoresistance in silicon based ferrite magnetic tunnel junction S. Ravi a, *, A. Karthikeyan a, N. Angel Nesakumari b, K.S. Pugazhvadivu c, K. Tamilarasan c a
Department of Physics, Mepco Schlenk Engineering College, Sivakasi, India Department of Nanoscience and Technology, Mepco Schlenk Engineering College, Sivakasi, India c Department of Physics, Kongu Engineering College, Perundurai, India b
a r t i c l e i n f o
a b s t r a c t
Article history: Received 31 May 2013 Received in revised form 20 November 2013 Accepted 21 November 2013 Available online 3 December 2013
We report magnetoresistance for silicon based magnetic tunnel junction. We used cobalt ferrite & cobalt nickel ferrite as free layer and pinned layer. The magnetoresistance measured at room temperature through silicon by fabricating FM/Si/FM magnetic tunnel junction. Magnetoresistance shows a loop type behavior with 3.7%. We have successfully demonstrated spin tunneling through silicon with ferrite junction that opens the door for potential candidate for spintronics devices. The spin-filtering effect for this double spin-filter junction is also discussed. Ó 2013 Elsevier B.V. All rights reserved.
Keywords: Magnetic materials Spintronics Magnetoresistance Ferrite based magnetic tunnel junction Double spin-filter junction
1. Introduction Spintronics is now an essential field of research with major application in several technologies. The basic concept of spintronics is the manipulation of spin-polarized currents, in contrast to conventional electronics in which spin of an electron is ignored. Mott [1] experimentally demonstrated this influence followed by some important theoretical works [2e5] more than 10 years before the discovery of GMR. Nowadays spintronics is expanding in so many directions that aim at combining the potential semiconductors [6] and through DNA [7]. Excellent results have been obtained with GaAs doped with Mn [8] but with low Curie temperature of 170 K. Fabrication of hybrid structures associated with ferromagnetic metal and conventional semiconductors are done, but with only modest advances [9e18]. Recently the usage of silicon has attracted much attention due to its plenty of advantages. The main thing is that it can readily replace conventional electronics as we are living in the silicon world. The other things are due to its low atomic weight, lattice inversion symmetry, low isotopic abundance of species having nuclear spin, etc. [7,19,20]. It also possesses long spin life time and spin coherence lengths, which are very significant parameters in spintronics, enabling spin based silicon integrated circuits [21,22]. The main problem is the fundamental impedance
* Corresponding author. Tel.: þ91 4562235690; fax: þ91 4562235111. E-mail address: [email protected] (S. Ravi). 1567-1739/$ e see front matter Ó 2013 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.cap.2013.11.024
mismatch between ferromagnetic metal and silicon [23]. Appelbaum [24] demonstrates spin transport across 10 mm undoped silicon in a device that operates by spin dependent ballistic hot electron injection. The usage ferromagnetic (FM) metal has an advantage of easy spin injection and detection (ignoring the impedance mismatch with indirect bandgap semiconductor), but the problem is its difficulty in device based applications as the most of FM metals are strong ferromagnetic in nature, which leads to complications in read/write process for future devices that run on spin. We used low-conductive ferromagnets (ferrites) and studied spin tunneling phenomenon through silicon. We designed a cobalt ferrite/Si/cobalt nickel ferrite based magnetic junction over copper coated silicon wafer and studied its magnetoresistive phenomenon. The other advantage of this type of junction is that it can act as a spin-filter junction leading to high MR. The schematic representation of the magnetic junction is shown in Fig. 1 with its IeV characteristic showing the expected non-linear behavior. The results are promising to use it in the future devices that run on spin. 2. Experimental section Cobalt ferrite (CoFe2O4) and cobalt nickel ferrite (CoNiFe2O4) were prepared similar to the method reported by Maaz et al. [25]. Highly purified (99.99%) iron chloride, cobalt chloride and nickel chloride were used as precursors. Oleic acid (HPLC grade) is used as a surfactant. NaOH was used to maintain the pH level. Cobalt ferrite was prepared by dissolving iron chloride (0.4 M) and cobalt
260
S. Ravi et al. / Current Applied Physics 14 (2014) 259e263
Fig. 1. Schematic representation of magnetic junction. IeV plot of this junction shows non-linear behavior.
chloride (0.2 M) in 50 ml double distilled water. The resultant mixture was stirred continuously. NaOH (3 M) solution was added slowly until the pH reaches the value of 12. Appropriate amount of oleic acid was added as surfactant. This was heated to 80 C for 1 h and stirred at the rate of 1200 rpm. The resultant solution was filtered and washed several times in the double distilled water. The resultant product was maintained at 100 C overnight and calcined to 800 C for 10 h to obtain precise structure and size. For cobalt nickel ferrite, iron chloride (0.4 M), cobalt chloride (0.1 M) and nickel chloride (0.1 M) were dissolved in 50 ml of double distilled water. Same procedure was used to obtain fine cobalt nickel ferrite powder. The synthesized cobalt ferrite and cobalt nickel ferrite was used to prepare sputtering targets. The sputtering targets were prepared by using Die casting mold to produce 2 inch diameter targets. 25 g of cobalt ferrite and cobalt nickel ferrite powder were taken separately and grinded finely in the mortar with 3% of poly vinyl alcohol (PVA) as a binder. A pressure of about 450 KN was applied to the mixture kept in the Die, using a Universal testing machine to make the sputtering targets. In this process, strong two inch diameter targets of cobalt ferrite and cobalt nickel ferrite were obtained. These targets were sintered for 900 C for 10 h and allowed to cool at room temperature. In the similar way, silicon targets were also prepared by the same method using silicon nanopowder (99.99%, Sigma Aldrich) as starting material. These pellets were preserved under vacuum condition and taken at the time of sputtering.
Silicon wafer were scribed on the backside with a diamond pen in to 1 1 cm squares and cleaned using the ultrasonic cleaner with ethanol and acetone. To the polished side, copper was sputtered using commercially available copper target using RF magnetron sputtering at the pressure of 4 105 mbar. Copper film was deposited uniformly over the silicon wafer to the thickness of 10 nm, which was monitored by the thickness monitor of the sputtering unit. This 10 nm copper was used as conductive electrode to the magnetic junction. The prepared cobalt nickel ferrite targets were now placed in the target holder and sputtered to form 5 nm thickness cobalt nickel ferrite, which serves as pinned layer of the junction. Silicon layer was sputtered over this from the prepared silicon target and sputtered to form 2 nm thick silicon layer. Finally, the cobalt ferrite target was used to sputtered to form a 5 nm thick over this. Cobalt ferrite was used as free layer in the magnetic junction. Thus, we obtain Si/Cu(10 nm)/CoNiFe2O4(5 nm)/ Si (2 nm)/CoFe2O4 (5 nm) based magnetic junction. All the layer thickness was controlled by the RF/DC sputtering unit. 3. Results and discussion XRD measurements of cobalt ferrite and cobalt nickel ferrite pellets were presented in Fig. 2. In Fig 2a, calcined (800 C) cobalt ferrite shows all the peaks for CoFe2O4 with indices quite agrees well with standard data (JCPDS: 22-1086). No other peaks and impurities were detected. Therefore, it confirmed the formation of phase pure CoFe2O4 particles. The particle size was found to be
Fig. 2. XRD measurement for cobalt ferrite (a) and cobalt nickel ferrite (b) showing the formation phase pure product with particle size 20 nm (a) and 25 nm (b).
S. Ravi et al. / Current Applied Physics 14 (2014) 259e263
261
20 nm from Scherer’s formula. Fig 2b shows the XRD for cobalt nickel ferrite indicating the presence of spinel cubic structure [26] and all the peaks corresponding to cobalt nickel ferrite [27]. It almost matches with cobalt ferrite exhibiting pure magnetite structure. Again no other peaks were found and hence the synthesized particles exhibit pure phase with no impurity. Generally cobalt (Co2þ) and nickel (Ni2þ) have almost same ionic radii with (0.78 A for cobalt and 0.74 A for nickel) and hence possess same structure in the host matrix. Thus, the peaks for cobalt ferrite and nickel ferrite overlaps and exist in the mixed state (JCPDS card no. 10-0325 (NiFe2O4) and 22-1086 (CoFe2O4)), leading to the difficulty in differentiating and resembles for perfect magnetite structure [28]. For cobalt ferrite the lattice parameter are high compared to cobalt nickel ferrite. The larger value of the lattice parameter is due to cobalt domination in the structure [29]. XRD pattern for the thin film is presented in Fig. 3. Fig. 3a and b are the XRD pattern for cobalt nickel ferrite and cobalt ferrite film respectively. The films shows the peak at (5 1 1), (4 4 0) and (6 2 2) spinel structure with no other impurities (JCPDS number # 22-1086). The high intense peak near 70 and 31 corresponds to Si (4 4 0) and (2 2 0). Thus cobalt ferrite and cobalt nickel ferrite prefers in (6 2 2) orientation texture with (4 4 0) plane. AFM image for cobalt nickel ferrite film is presented in Fig. 4. It shows uniform distribution of cobalt nickel ferrite nanoparticles allowing the silicon to grow over this. The roughness factor for this film is found to be less than 4 nm, which clearly states the smoothness and quality of the film. The MeH measurements for the ferrites are reported in our previous report with coercive fields of 1250 Oe and 427 Oe for cobalt nickel ferrite and cobalt ferrite respectively [34]. The resistance of the sample was measured prior to magnetoresistance measurement. It was found that the resistance vary from
Fig. 4. AFM image of cobalt nickel ferrite film grown on Si substrate (a) 3 dimensional view (b) Topographical view. This shows uniform formation for cobalt nickel ferrite with roughness factor Rq < 4 nm.
13 to 15 KU. Fig. 5 shows the characteristic MR curve obtained at room temperature. This was recorded for two field sweep directions. The black curve and arrow indicates that field sweep from 4000 to þ4000 Oe. The red curve and arrow indicates the reverse field sweep from þ4000 Oe to 4000 Oe. Since the MR behavior is not like single peak or single curve like structure with maximum at zero field, it is clear that this not granular MR between ferrite and Si, rather it is MR through Si between two ferrites MR is obtained using the general formula:
TMR ¼
Fig. 3. XRD Pattern of (a) cobalt nickel ferrite and (b) cobalt ferrite films showing that the film preferably crystallizes at (440) and (622) planes.
Rp RAP RP
100;
where RP and RAP are resistance with minimum and maximum in the field sweep range. The MR curve was characteristically showing zero base line MR and hysteretic behavior with largest MR change of 3.7%. The hysteretic curves are indicative of spin valve behavior. By sweeping the field from 4000 Oe, the resistance was almost constant at 1660 Oe it increases rapidly and reaches a maximum at 110 Oe. Further sweep results in a sudden drop at þ64 Oe and almost remain constant up to þ4000 Oe. In the reverse sweep
S. Ravi et al. / Current Applied Physics 14 (2014) 259e263
TMR %
262
Bias Voltage (V) Fig. 5. Magnetoresistance curve for our junction. Black arrow for forward field sweep and red arrow indicates reverse field sweep. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)
direction, same type pattern is obtained resulting in hysteretic behavior. It could be noted that the jump in the resistance for positive and negative field occurs at different field. As scattering occurs in the interface of Si, the gradual decrease in the resistance was noticed. Similar pattern was observed in the reverse field direction. A marginal swift in the resistance may be due to relaxation mechanism of free and pinned layer. This may be due the magnetization saturation through Si does not respond rapidly when the field is reversed (again this is only speculation and the exact mechanism for this type needs to be studied). David J Smith et al. [30] reported that about 2% (i.e. 300 U) deviation of parallel and antiparallel merger in high field. They attribute this result due to the interfacial effect of oxide layer, although they could not predict the exact mechanism of this anomalous result. In our work, although for negative field there is a merger at 1389 Oe, positive higher field has the deviation of 100 U. Some other reports are also available similar to this type without detailed understanding. The difference jump occurred is comparable with the reports for Si based junction [31e33]. Infact this difference is very low compared to our earlier report on SiO2 based ferrite magnetic junction [34]. Similar type of magnetoresistance behavior was reported by Lüders et al. [35] for LSMO/NFO/Au junction with an abrupt jump at 1000 Oe with no merger for higher field. One of the reason they cited is due the domain structure of ferrite based electrode which itself act as spin filtering agent. The same is applicable for our study, where we are using two ferrite electrodes. The difference is very higher for their report (w0.18 MU). The MR curve for ferrite/Si was presented in Fig. 6, which shows the different pattern similar to
Fig. 7. TMR % Vs bias voltage for cobalt nickel ferrite/Si/cobalt ferrite magnetic tunnel junction.
granular GMR. Thus, it clearly shows that the spin tunneling through silicon occurred which is shown by characteristic MR curve presented in Fig. 5. Fig. 7 displays the TMR Vs bias voltage for the junction at room temperature (RT). TMR% decreases with increasing bias voltage and the bias voltage dependence on TMR obtained is compatible with the tunneling theory [36e39]. The V1/2 of maximum for both positive and negative bias voltage is very high (in the range of þ0.64 V and 0.72 V) and hence be desirable for magnetoresistance device. Spin-filtering effect is an important aspect in the spintronics field to achieve high TMR [39]. We have reported the spin-filtering effect of DNA in FeCo/DNA/FeCo based magnetic junction with spinfiltering efficiency of 30% [7]. It has been shown that cobalt ferrite is an excellent example of spin-filter in room temperature due to its high Curie temperature and insulating property [40]. In our experiment, CoFe2O4/Si/CoNiFe2O4 based tunnel junction should as a double spin-filter junction. Since several parameters like exchange splitting, average barrier height, growth technique, barrier thickness and so on affect the practical reliability of observing the double spin-filter effect [41], the experimental effect was not observed and the spin-filtering efficiency was found to be negligible. The bias dependence on TMR also shows that TMR decreases as bias voltage increases opposite to that of spin-filtering effect. The possible reason may be due to the impedance mismatch between ferrites with silicon, which can be improved by the fabricating technique. The other reason may be due to the cobalt ferrite and cobalt nickel ferrites have different magnetic properties (or structures) that are not suitable for observing the double spin-filter effect [41]. The only report available for double spin-filter effect was by Miao et al. [42] for EuS/Al2O3/EuS (with EuS being a superconductor) junction with uncharacteristic bias dependence. We are working extensively to show the double spin-filter effect for our junction and hope that it would be made in the coming days.
4. Conclusions We have successfully fabricated ferrite based magnetic junction and studied spin tunneling phenomenon through silicon for spintronics devices. The exchange bias between cobalt ferrite and cobalt nickel ferrite supports good magnetoresistive behavior with 3.7%.
Acknowledgments Fig. 6. MR behavior of ferrite/Si granular junction showing single peak (The inset picture shows the schematic diagram for this junction).
The corresponding author (Dr. S. Ravi) is very much grateful to Department of Science and Technology (DST), Government of India
S. Ravi et al. / Current Applied Physics 14 (2014) 259e263
for providing the financial assistance through the project (Fast track scheme for young scientist) 100/IFD/5581/2010-2011. References [1] N.F. Mott, The electrical conductivity of transition metals, Proc. R. Soc. A 153 (1936) 699e718. [2] A. Fert, I.A. Campbel, Two current conduction in nickel, Phys. Rev. Lett. 21 (1968) 1190e1192. [3] A. Fert, I.A. Campbel, Electrical resistivity of ferromagnetic nickel and iron based alloys, J. Phys. F. Met. Phys. 6 (1976) 849e871. [4] A. Fert, I.A. Campbel, Transport properties of ferromagnetic transition metals, J. Phys. 32 (1971). C1-46eC1-50. [5] B. Loegel, F. Gautier, Origine de la resistivite dans le cobalt et ses alliages dilues, J. Phys. Chem. Solids 32 (1971) 2723e2735. [6] I. Zutic, J. Fabian, S.D. Sarma, Spintronics fundamental and applications, Rev. Mod. Phys. 76 (2004) 323e410. [7] S. Ravi, P. Sowmya, A. Karthikeyan, Magnetoresistance and spin-filtering efficiency of DNA-sandwiched ferromagnetic nanostructures, SPIN 3 (2013) 1350003 (4 pp.). [8] H. Ohno, A. Shen, F. Matsukura, et al., A new diluted magnetic semiconductor based on GaAs, Appl. Phys. Lett. 69 (1996) 363e365. [9] V.F. Motsnyi, P. Van Dorpe, W. Wan Roy, et al., Optical investigation of electrical spin injection into semiconductor, Phys. Rev. B 68 (2003) 245319e 245331. [10] A.T. Hanbicki, O.M.J. van’t Erve, R. Magno, et al., Analysis of the transport process providing spin injection through Fe/GaAs Schottky barrier, Appl. Phys. Lett. 82 (2003) 4092e4094. [11] Y. Lu, V.G. Truong, P. Renucci, et al., MgO thickness dependence of spin injection efficiency in spin-light emitting diodes, Appl. Phys. Lett. 93 (2008) 152102-1e152102-3. [12] X. Lou, C. Adelmann, S.A. Crooker, et al., Electrical detection of spin transport in lateral ferromagnet-semiconductor devices, Nat. Phys. 3 (2007) 197e202. [13] O.M.J. van’t Erve, A.T. Hanbicki, M. Holub, et al., Electrical injection and detection of spin-polarised carriers in silicon in a lateral transport geometry, Appl. Phys. Lett. 91 (2007) 212109-1e212109-13. [14] M. Tran, H. Jaffres, C. Deranlot, et al., Enhancement of spin accumulation at the interface between a spin-polarized tunnel injection and a semiconductor, Phys. Rev. Lett. 102 (2009) 036601-1e036601-14. [15] S.P. Dash, S. Sharma, R.S. Patel, et al., Electrical creation of spin polarization in silicon at room temperature, Nature 462 (2009) 491e494. [16] S. Datta, B. Das, Electrical analog of electric optic modulator, Appl. Phys. Lett. 56 (1990) 665e667. [17] H.-C. Koo, J.-H. Hwoon, J. Eom, et al., Control of spin precession in a spin injected field effect transistor, Science 325 (2009) 1515e1518. [18] M. Konig, S. Wiedmann, C. Brune, et al., Quantum spin hall insulator state in HgTe quantum wells, Science 318 (2007) 766e770. [19] I. Zutic, J. Fabian, S.C. Erwin, Spin injection and detection in silicon, Phys. Rev. Lett. 97 (2006) 026602e026605. [20] A.M. Tyryshkin, S.A. Lyon, A.V. Astashkin, A.M. Raitsimring, Electron spin relaxation times of phosphorus donors in silicon, Phys. Rev. B 68 (2003) 193207e193210. [21] I. Zutic, J. Fabian, Spintronics: silicon twists, Nature 447 (2007) 268e270. [22] M.A. Nidsen, I.L. Chuang, Quantum Computation and Quantum Information, Cambridge University Press, 2000.
263
[23] G. Schmidt, D. Ferrand, L.W. Molenkamp, A.T. Filip, B.J. van Wees, Fundamental obstacle for electrical spin injection from ferromagnetic metal into a diffusive semiconductor, Phys. Rev. B 62 (2000) R4790eR4793. [24] I. Appelbaum, B. Hunang, D.J. Monsma, Electronic measurement and control of spin transport in silicon, Nature 447 (2007) 295e298. [25] K. Maaz, S. Karim, A. Mumtaz, S.K. Hasanain, J. Liu, J.L. Duan, Synthesis and magnetic characterization of nickel ferrite nanoparticles prepared by coprecipitation route, J. Magn. Magn. Mater. 321 (2009) 1838e1842. [26] A. Baykal, N. Kasapoglu, Z. Durmus, H. Kavas, M.S. Toprak, Y. Koseoglu, CTABassisted hydrothermal synthesis and magnetic characterization of NixCo1xFe2O4 nanoparticle (x¼0.0, 0.6, 1.0), Turk. J. Chem. 33 (2009) 33e45. [27] A.M. El-Sayed, Influence of zinc content on some properties of NieZn ferrites, Ceram. Int. 28 (2002) 363e367. [28] A. Cabanas, M. Poliakoff, The continuous hydrothermal synthesis of nanoparticulate ferrites in near critical and supercritical water, J. Mater. Chem. 11 (2001) 1408e1416. [29] R. Mukherjee, T. Sahu, S. Sen, P. Sahu, Structural and microstructural evolution due to increasing Co substitution in Ni1xCoxFe2O4: an X-ray diffraction study using the Rietveld method, Mater. Chem. Phys. 128 (2011) 364e370. [30] D.J. Smith, M.R. McCartney, C.L. Platt, A.E. Berkowitz, Structural characterization of thin film ferromagnetic tunnel junctions, J. Appl. Phys. 83 (1998) 5154e5158. [31] S. Sankar, A.E. Berkowitz, D.J. Smith, Spin-dependent tunneling in discontinuous CoeSiO2 magnetic tunnel junctions, Appl. Phys. Lett. 73 (1998) 535e537. [32] C.H. Li, O.M.J. van ’t Erve, B.T. Jonker, Electrical injection and detection of spin accumulation in silicon at 500 K with magnetic metal/silicon dioxide contacts, Nat. Commun. 2 (2011). Article No. 245. [33] J.C. Denardin, M. Knobel, L.S. Dorneles, L.F. Schelp, Structural and magnetotransport properties of discontinuous Co/SiO2 multilayers, Mater. Sci. Eng. B 112 (2004) 120e122. [34] S. Ravi, A. Karthikeyan, V. Aravindan, K.S. Pugazhvadivu, K. Tamilarasan, Experimental investigations of SiO2 based ferrite magnetic tunnel junction, Mater. Sci. Eng. B 178 (2013) 937e941. [35] U. Lüders, M. Bibes, K. Bouzehouane, E. Jacquet, J.-P. Contour, S. Fusil, J.F. Bobo, J. Fontcuberta, A. Barthélémy, A. Fert, Spin filtering through ferrimagnetic NiFe2O4 tunnel barriers, Appl. Phys. Lett. 88 (2006) 082505e082507. [36] S. Zhang, P.M. Levy, A.C. Marley, S.S.P. Parkin, Quenching of magnetoresistance by hot electrons in magnetic tunnel junctions, Phys. Rev. Lett. 79 (1997) 3744e3747. [37] C. Lü, M.W. Wu, X.F. Han, Magnon- and phonon-assisted tunneling in a highmagnetoresistance tunnel junction using Co75Fe25 ferromagnetic electrodes, Phys. Lett. A 319 (2003) 205e210. [38] X.F. Han, W.Y. Lai, J.-Q. Wang, C.J. O’Connor, T. Miyazaki, A self-consistent calculation of intrinsic magnetoelectric properties in magnetic tunnel junctions, J. Magn. Magn. Mater. 239 (2002) 167e169. [39] K. Ray, S.P. Ananthavel, D.H. Waldeck, R. Naaman, Asymmetric scattering of polarized electrons by organized organic films of chiral molecules, Science 283 (1999) 814e816. [40] A.V. Ramos, M.-J. Guittet, J.-B. Moussy, R. Mattana, C. Deranlot, F. Petroff, C. Gatel, Room temperature spin filtering in epitaxial cobalt-ferrite tunnel barriers, Appl. Phys. Lett. 91 (2007) 122107 (3 pp.). [41] E.Y. Tsymbal, I. Zutic, Handbook of Spin Transport and Magnetism, CRC Press, Taylor & Francis Group, 2012, p. 261. [42] G.X. Miao, M. Muller, J.S. Moodera, Magnetoresistance in double spin filter tunnel junctions with nonmagnetic electrodes and its unconventional bias dependence, Phys. Rev. Lett. 102 (2009) 076601 (4 pp.).
Contents lists available at ScienceDirect
Current Applied Physics journal homepage: www.elsevier.com/locate/cap
Magnetoresistance in silicon based ferrite magnetic tunnel junction S. Ravi a, *, A. Karthikeyan a, N. Angel Nesakumari b, K.S. Pugazhvadivu c, K. Tamilarasan c a
Department of Physics, Mepco Schlenk Engineering College, Sivakasi, India Department of Nanoscience and Technology, Mepco Schlenk Engineering College, Sivakasi, India c Department of Physics, Kongu Engineering College, Perundurai, India b
a r t i c l e i n f o
a b s t r a c t
Article history: Received 31 May 2013 Received in revised form 20 November 2013 Accepted 21 November 2013 Available online 3 December 2013
We report magnetoresistance for silicon based magnetic tunnel junction. We used cobalt ferrite & cobalt nickel ferrite as free layer and pinned layer. The magnetoresistance measured at room temperature through silicon by fabricating FM/Si/FM magnetic tunnel junction. Magnetoresistance shows a loop type behavior with 3.7%. We have successfully demonstrated spin tunneling through silicon with ferrite junction that opens the door for potential candidate for spintronics devices. The spin-filtering effect for this double spin-filter junction is also discussed. Ó 2013 Elsevier B.V. All rights reserved.
Keywords: Magnetic materials Spintronics Magnetoresistance Ferrite based magnetic tunnel junction Double spin-filter junction
1. Introduction Spintronics is now an essential field of research with major application in several technologies. The basic concept of spintronics is the manipulation of spin-polarized currents, in contrast to conventional electronics in which spin of an electron is ignored. Mott [1] experimentally demonstrated this influence followed by some important theoretical works [2e5] more than 10 years before the discovery of GMR. Nowadays spintronics is expanding in so many directions that aim at combining the potential semiconductors [6] and through DNA [7]. Excellent results have been obtained with GaAs doped with Mn [8] but with low Curie temperature of 170 K. Fabrication of hybrid structures associated with ferromagnetic metal and conventional semiconductors are done, but with only modest advances [9e18]. Recently the usage of silicon has attracted much attention due to its plenty of advantages. The main thing is that it can readily replace conventional electronics as we are living in the silicon world. The other things are due to its low atomic weight, lattice inversion symmetry, low isotopic abundance of species having nuclear spin, etc. [7,19,20]. It also possesses long spin life time and spin coherence lengths, which are very significant parameters in spintronics, enabling spin based silicon integrated circuits [21,22]. The main problem is the fundamental impedance
* Corresponding author. Tel.: þ91 4562235690; fax: þ91 4562235111. E-mail address: [email protected] (S. Ravi). 1567-1739/$ e see front matter Ó 2013 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.cap.2013.11.024
mismatch between ferromagnetic metal and silicon [23]. Appelbaum [24] demonstrates spin transport across 10 mm undoped silicon in a device that operates by spin dependent ballistic hot electron injection. The usage ferromagnetic (FM) metal has an advantage of easy spin injection and detection (ignoring the impedance mismatch with indirect bandgap semiconductor), but the problem is its difficulty in device based applications as the most of FM metals are strong ferromagnetic in nature, which leads to complications in read/write process for future devices that run on spin. We used low-conductive ferromagnets (ferrites) and studied spin tunneling phenomenon through silicon. We designed a cobalt ferrite/Si/cobalt nickel ferrite based magnetic junction over copper coated silicon wafer and studied its magnetoresistive phenomenon. The other advantage of this type of junction is that it can act as a spin-filter junction leading to high MR. The schematic representation of the magnetic junction is shown in Fig. 1 with its IeV characteristic showing the expected non-linear behavior. The results are promising to use it in the future devices that run on spin. 2. Experimental section Cobalt ferrite (CoFe2O4) and cobalt nickel ferrite (CoNiFe2O4) were prepared similar to the method reported by Maaz et al. [25]. Highly purified (99.99%) iron chloride, cobalt chloride and nickel chloride were used as precursors. Oleic acid (HPLC grade) is used as a surfactant. NaOH was used to maintain the pH level. Cobalt ferrite was prepared by dissolving iron chloride (0.4 M) and cobalt
260
S. Ravi et al. / Current Applied Physics 14 (2014) 259e263
Fig. 1. Schematic representation of magnetic junction. IeV plot of this junction shows non-linear behavior.
chloride (0.2 M) in 50 ml double distilled water. The resultant mixture was stirred continuously. NaOH (3 M) solution was added slowly until the pH reaches the value of 12. Appropriate amount of oleic acid was added as surfactant. This was heated to 80 C for 1 h and stirred at the rate of 1200 rpm. The resultant solution was filtered and washed several times in the double distilled water. The resultant product was maintained at 100 C overnight and calcined to 800 C for 10 h to obtain precise structure and size. For cobalt nickel ferrite, iron chloride (0.4 M), cobalt chloride (0.1 M) and nickel chloride (0.1 M) were dissolved in 50 ml of double distilled water. Same procedure was used to obtain fine cobalt nickel ferrite powder. The synthesized cobalt ferrite and cobalt nickel ferrite was used to prepare sputtering targets. The sputtering targets were prepared by using Die casting mold to produce 2 inch diameter targets. 25 g of cobalt ferrite and cobalt nickel ferrite powder were taken separately and grinded finely in the mortar with 3% of poly vinyl alcohol (PVA) as a binder. A pressure of about 450 KN was applied to the mixture kept in the Die, using a Universal testing machine to make the sputtering targets. In this process, strong two inch diameter targets of cobalt ferrite and cobalt nickel ferrite were obtained. These targets were sintered for 900 C for 10 h and allowed to cool at room temperature. In the similar way, silicon targets were also prepared by the same method using silicon nanopowder (99.99%, Sigma Aldrich) as starting material. These pellets were preserved under vacuum condition and taken at the time of sputtering.
Silicon wafer were scribed on the backside with a diamond pen in to 1 1 cm squares and cleaned using the ultrasonic cleaner with ethanol and acetone. To the polished side, copper was sputtered using commercially available copper target using RF magnetron sputtering at the pressure of 4 105 mbar. Copper film was deposited uniformly over the silicon wafer to the thickness of 10 nm, which was monitored by the thickness monitor of the sputtering unit. This 10 nm copper was used as conductive electrode to the magnetic junction. The prepared cobalt nickel ferrite targets were now placed in the target holder and sputtered to form 5 nm thickness cobalt nickel ferrite, which serves as pinned layer of the junction. Silicon layer was sputtered over this from the prepared silicon target and sputtered to form 2 nm thick silicon layer. Finally, the cobalt ferrite target was used to sputtered to form a 5 nm thick over this. Cobalt ferrite was used as free layer in the magnetic junction. Thus, we obtain Si/Cu(10 nm)/CoNiFe2O4(5 nm)/ Si (2 nm)/CoFe2O4 (5 nm) based magnetic junction. All the layer thickness was controlled by the RF/DC sputtering unit. 3. Results and discussion XRD measurements of cobalt ferrite and cobalt nickel ferrite pellets were presented in Fig. 2. In Fig 2a, calcined (800 C) cobalt ferrite shows all the peaks for CoFe2O4 with indices quite agrees well with standard data (JCPDS: 22-1086). No other peaks and impurities were detected. Therefore, it confirmed the formation of phase pure CoFe2O4 particles. The particle size was found to be
Fig. 2. XRD measurement for cobalt ferrite (a) and cobalt nickel ferrite (b) showing the formation phase pure product with particle size 20 nm (a) and 25 nm (b).
S. Ravi et al. / Current Applied Physics 14 (2014) 259e263
261
20 nm from Scherer’s formula. Fig 2b shows the XRD for cobalt nickel ferrite indicating the presence of spinel cubic structure [26] and all the peaks corresponding to cobalt nickel ferrite [27]. It almost matches with cobalt ferrite exhibiting pure magnetite structure. Again no other peaks were found and hence the synthesized particles exhibit pure phase with no impurity. Generally cobalt (Co2þ) and nickel (Ni2þ) have almost same ionic radii with (0.78 A for cobalt and 0.74 A for nickel) and hence possess same structure in the host matrix. Thus, the peaks for cobalt ferrite and nickel ferrite overlaps and exist in the mixed state (JCPDS card no. 10-0325 (NiFe2O4) and 22-1086 (CoFe2O4)), leading to the difficulty in differentiating and resembles for perfect magnetite structure [28]. For cobalt ferrite the lattice parameter are high compared to cobalt nickel ferrite. The larger value of the lattice parameter is due to cobalt domination in the structure [29]. XRD pattern for the thin film is presented in Fig. 3. Fig. 3a and b are the XRD pattern for cobalt nickel ferrite and cobalt ferrite film respectively. The films shows the peak at (5 1 1), (4 4 0) and (6 2 2) spinel structure with no other impurities (JCPDS number # 22-1086). The high intense peak near 70 and 31 corresponds to Si (4 4 0) and (2 2 0). Thus cobalt ferrite and cobalt nickel ferrite prefers in (6 2 2) orientation texture with (4 4 0) plane. AFM image for cobalt nickel ferrite film is presented in Fig. 4. It shows uniform distribution of cobalt nickel ferrite nanoparticles allowing the silicon to grow over this. The roughness factor for this film is found to be less than 4 nm, which clearly states the smoothness and quality of the film. The MeH measurements for the ferrites are reported in our previous report with coercive fields of 1250 Oe and 427 Oe for cobalt nickel ferrite and cobalt ferrite respectively [34]. The resistance of the sample was measured prior to magnetoresistance measurement. It was found that the resistance vary from
Fig. 4. AFM image of cobalt nickel ferrite film grown on Si substrate (a) 3 dimensional view (b) Topographical view. This shows uniform formation for cobalt nickel ferrite with roughness factor Rq < 4 nm.
13 to 15 KU. Fig. 5 shows the characteristic MR curve obtained at room temperature. This was recorded for two field sweep directions. The black curve and arrow indicates that field sweep from 4000 to þ4000 Oe. The red curve and arrow indicates the reverse field sweep from þ4000 Oe to 4000 Oe. Since the MR behavior is not like single peak or single curve like structure with maximum at zero field, it is clear that this not granular MR between ferrite and Si, rather it is MR through Si between two ferrites MR is obtained using the general formula:
TMR ¼
Fig. 3. XRD Pattern of (a) cobalt nickel ferrite and (b) cobalt ferrite films showing that the film preferably crystallizes at (440) and (622) planes.
Rp RAP RP
100;
where RP and RAP are resistance with minimum and maximum in the field sweep range. The MR curve was characteristically showing zero base line MR and hysteretic behavior with largest MR change of 3.7%. The hysteretic curves are indicative of spin valve behavior. By sweeping the field from 4000 Oe, the resistance was almost constant at 1660 Oe it increases rapidly and reaches a maximum at 110 Oe. Further sweep results in a sudden drop at þ64 Oe and almost remain constant up to þ4000 Oe. In the reverse sweep
S. Ravi et al. / Current Applied Physics 14 (2014) 259e263
TMR %
262
Bias Voltage (V) Fig. 5. Magnetoresistance curve for our junction. Black arrow for forward field sweep and red arrow indicates reverse field sweep. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)
direction, same type pattern is obtained resulting in hysteretic behavior. It could be noted that the jump in the resistance for positive and negative field occurs at different field. As scattering occurs in the interface of Si, the gradual decrease in the resistance was noticed. Similar pattern was observed in the reverse field direction. A marginal swift in the resistance may be due to relaxation mechanism of free and pinned layer. This may be due the magnetization saturation through Si does not respond rapidly when the field is reversed (again this is only speculation and the exact mechanism for this type needs to be studied). David J Smith et al. [30] reported that about 2% (i.e. 300 U) deviation of parallel and antiparallel merger in high field. They attribute this result due to the interfacial effect of oxide layer, although they could not predict the exact mechanism of this anomalous result. In our work, although for negative field there is a merger at 1389 Oe, positive higher field has the deviation of 100 U. Some other reports are also available similar to this type without detailed understanding. The difference jump occurred is comparable with the reports for Si based junction [31e33]. Infact this difference is very low compared to our earlier report on SiO2 based ferrite magnetic junction [34]. Similar type of magnetoresistance behavior was reported by Lüders et al. [35] for LSMO/NFO/Au junction with an abrupt jump at 1000 Oe with no merger for higher field. One of the reason they cited is due the domain structure of ferrite based electrode which itself act as spin filtering agent. The same is applicable for our study, where we are using two ferrite electrodes. The difference is very higher for their report (w0.18 MU). The MR curve for ferrite/Si was presented in Fig. 6, which shows the different pattern similar to
Fig. 7. TMR % Vs bias voltage for cobalt nickel ferrite/Si/cobalt ferrite magnetic tunnel junction.
granular GMR. Thus, it clearly shows that the spin tunneling through silicon occurred which is shown by characteristic MR curve presented in Fig. 5. Fig. 7 displays the TMR Vs bias voltage for the junction at room temperature (RT). TMR% decreases with increasing bias voltage and the bias voltage dependence on TMR obtained is compatible with the tunneling theory [36e39]. The V1/2 of maximum for both positive and negative bias voltage is very high (in the range of þ0.64 V and 0.72 V) and hence be desirable for magnetoresistance device. Spin-filtering effect is an important aspect in the spintronics field to achieve high TMR [39]. We have reported the spin-filtering effect of DNA in FeCo/DNA/FeCo based magnetic junction with spinfiltering efficiency of 30% [7]. It has been shown that cobalt ferrite is an excellent example of spin-filter in room temperature due to its high Curie temperature and insulating property [40]. In our experiment, CoFe2O4/Si/CoNiFe2O4 based tunnel junction should as a double spin-filter junction. Since several parameters like exchange splitting, average barrier height, growth technique, barrier thickness and so on affect the practical reliability of observing the double spin-filter effect [41], the experimental effect was not observed and the spin-filtering efficiency was found to be negligible. The bias dependence on TMR also shows that TMR decreases as bias voltage increases opposite to that of spin-filtering effect. The possible reason may be due to the impedance mismatch between ferrites with silicon, which can be improved by the fabricating technique. The other reason may be due to the cobalt ferrite and cobalt nickel ferrites have different magnetic properties (or structures) that are not suitable for observing the double spin-filter effect [41]. The only report available for double spin-filter effect was by Miao et al. [42] for EuS/Al2O3/EuS (with EuS being a superconductor) junction with uncharacteristic bias dependence. We are working extensively to show the double spin-filter effect for our junction and hope that it would be made in the coming days.
4. Conclusions We have successfully fabricated ferrite based magnetic junction and studied spin tunneling phenomenon through silicon for spintronics devices. The exchange bias between cobalt ferrite and cobalt nickel ferrite supports good magnetoresistive behavior with 3.7%.
Acknowledgments Fig. 6. MR behavior of ferrite/Si granular junction showing single peak (The inset picture shows the schematic diagram for this junction).
The corresponding author (Dr. S. Ravi) is very much grateful to Department of Science and Technology (DST), Government of India
S. Ravi et al. / Current Applied Physics 14 (2014) 259e263
for providing the financial assistance through the project (Fast track scheme for young scientist) 100/IFD/5581/2010-2011. References [1] N.F. Mott, The electrical conductivity of transition metals, Proc. R. Soc. A 153 (1936) 699e718. [2] A. Fert, I.A. Campbel, Two current conduction in nickel, Phys. Rev. Lett. 21 (1968) 1190e1192. [3] A. Fert, I.A. Campbel, Electrical resistivity of ferromagnetic nickel and iron based alloys, J. Phys. F. Met. Phys. 6 (1976) 849e871. [4] A. Fert, I.A. Campbel, Transport properties of ferromagnetic transition metals, J. Phys. 32 (1971). C1-46eC1-50. [5] B. Loegel, F. Gautier, Origine de la resistivite dans le cobalt et ses alliages dilues, J. Phys. Chem. Solids 32 (1971) 2723e2735. [6] I. Zutic, J. Fabian, S.D. Sarma, Spintronics fundamental and applications, Rev. Mod. Phys. 76 (2004) 323e410. [7] S. Ravi, P. Sowmya, A. Karthikeyan, Magnetoresistance and spin-filtering efficiency of DNA-sandwiched ferromagnetic nanostructures, SPIN 3 (2013) 1350003 (4 pp.). [8] H. Ohno, A. Shen, F. Matsukura, et al., A new diluted magnetic semiconductor based on GaAs, Appl. Phys. Lett. 69 (1996) 363e365. [9] V.F. Motsnyi, P. Van Dorpe, W. Wan Roy, et al., Optical investigation of electrical spin injection into semiconductor, Phys. Rev. B 68 (2003) 245319e 245331. [10] A.T. Hanbicki, O.M.J. van’t Erve, R. Magno, et al., Analysis of the transport process providing spin injection through Fe/GaAs Schottky barrier, Appl. Phys. Lett. 82 (2003) 4092e4094. [11] Y. Lu, V.G. Truong, P. Renucci, et al., MgO thickness dependence of spin injection efficiency in spin-light emitting diodes, Appl. Phys. Lett. 93 (2008) 152102-1e152102-3. [12] X. Lou, C. Adelmann, S.A. Crooker, et al., Electrical detection of spin transport in lateral ferromagnet-semiconductor devices, Nat. Phys. 3 (2007) 197e202. [13] O.M.J. van’t Erve, A.T. Hanbicki, M. Holub, et al., Electrical injection and detection of spin-polarised carriers in silicon in a lateral transport geometry, Appl. Phys. Lett. 91 (2007) 212109-1e212109-13. [14] M. Tran, H. Jaffres, C. Deranlot, et al., Enhancement of spin accumulation at the interface between a spin-polarized tunnel injection and a semiconductor, Phys. Rev. Lett. 102 (2009) 036601-1e036601-14. [15] S.P. Dash, S. Sharma, R.S. Patel, et al., Electrical creation of spin polarization in silicon at room temperature, Nature 462 (2009) 491e494. [16] S. Datta, B. Das, Electrical analog of electric optic modulator, Appl. Phys. Lett. 56 (1990) 665e667. [17] H.-C. Koo, J.-H. Hwoon, J. Eom, et al., Control of spin precession in a spin injected field effect transistor, Science 325 (2009) 1515e1518. [18] M. Konig, S. Wiedmann, C. Brune, et al., Quantum spin hall insulator state in HgTe quantum wells, Science 318 (2007) 766e770. [19] I. Zutic, J. Fabian, S.C. Erwin, Spin injection and detection in silicon, Phys. Rev. Lett. 97 (2006) 026602e026605. [20] A.M. Tyryshkin, S.A. Lyon, A.V. Astashkin, A.M. Raitsimring, Electron spin relaxation times of phosphorus donors in silicon, Phys. Rev. B 68 (2003) 193207e193210. [21] I. Zutic, J. Fabian, Spintronics: silicon twists, Nature 447 (2007) 268e270. [22] M.A. Nidsen, I.L. Chuang, Quantum Computation and Quantum Information, Cambridge University Press, 2000.
263
[23] G. Schmidt, D. Ferrand, L.W. Molenkamp, A.T. Filip, B.J. van Wees, Fundamental obstacle for electrical spin injection from ferromagnetic metal into a diffusive semiconductor, Phys. Rev. B 62 (2000) R4790eR4793. [24] I. Appelbaum, B. Hunang, D.J. Monsma, Electronic measurement and control of spin transport in silicon, Nature 447 (2007) 295e298. [25] K. Maaz, S. Karim, A. Mumtaz, S.K. Hasanain, J. Liu, J.L. Duan, Synthesis and magnetic characterization of nickel ferrite nanoparticles prepared by coprecipitation route, J. Magn. Magn. Mater. 321 (2009) 1838e1842. [26] A. Baykal, N. Kasapoglu, Z. Durmus, H. Kavas, M.S. Toprak, Y. Koseoglu, CTABassisted hydrothermal synthesis and magnetic characterization of NixCo1xFe2O4 nanoparticle (x¼0.0, 0.6, 1.0), Turk. J. Chem. 33 (2009) 33e45. [27] A.M. El-Sayed, Influence of zinc content on some properties of NieZn ferrites, Ceram. Int. 28 (2002) 363e367. [28] A. Cabanas, M. Poliakoff, The continuous hydrothermal synthesis of nanoparticulate ferrites in near critical and supercritical water, J. Mater. Chem. 11 (2001) 1408e1416. [29] R. Mukherjee, T. Sahu, S. Sen, P. Sahu, Structural and microstructural evolution due to increasing Co substitution in Ni1xCoxFe2O4: an X-ray diffraction study using the Rietveld method, Mater. Chem. Phys. 128 (2011) 364e370. [30] D.J. Smith, M.R. McCartney, C.L. Platt, A.E. Berkowitz, Structural characterization of thin film ferromagnetic tunnel junctions, J. Appl. Phys. 83 (1998) 5154e5158. [31] S. Sankar, A.E. Berkowitz, D.J. Smith, Spin-dependent tunneling in discontinuous CoeSiO2 magnetic tunnel junctions, Appl. Phys. Lett. 73 (1998) 535e537. [32] C.H. Li, O.M.J. van ’t Erve, B.T. Jonker, Electrical injection and detection of spin accumulation in silicon at 500 K with magnetic metal/silicon dioxide contacts, Nat. Commun. 2 (2011). Article No. 245. [33] J.C. Denardin, M. Knobel, L.S. Dorneles, L.F. Schelp, Structural and magnetotransport properties of discontinuous Co/SiO2 multilayers, Mater. Sci. Eng. B 112 (2004) 120e122. [34] S. Ravi, A. Karthikeyan, V. Aravindan, K.S. Pugazhvadivu, K. Tamilarasan, Experimental investigations of SiO2 based ferrite magnetic tunnel junction, Mater. Sci. Eng. B 178 (2013) 937e941. [35] U. Lüders, M. Bibes, K. Bouzehouane, E. Jacquet, J.-P. Contour, S. Fusil, J.F. Bobo, J. Fontcuberta, A. Barthélémy, A. Fert, Spin filtering through ferrimagnetic NiFe2O4 tunnel barriers, Appl. Phys. Lett. 88 (2006) 082505e082507. [36] S. Zhang, P.M. Levy, A.C. Marley, S.S.P. Parkin, Quenching of magnetoresistance by hot electrons in magnetic tunnel junctions, Phys. Rev. Lett. 79 (1997) 3744e3747. [37] C. Lü, M.W. Wu, X.F. Han, Magnon- and phonon-assisted tunneling in a highmagnetoresistance tunnel junction using Co75Fe25 ferromagnetic electrodes, Phys. Lett. A 319 (2003) 205e210. [38] X.F. Han, W.Y. Lai, J.-Q. Wang, C.J. O’Connor, T. Miyazaki, A self-consistent calculation of intrinsic magnetoelectric properties in magnetic tunnel junctions, J. Magn. Magn. Mater. 239 (2002) 167e169. [39] K. Ray, S.P. Ananthavel, D.H. Waldeck, R. Naaman, Asymmetric scattering of polarized electrons by organized organic films of chiral molecules, Science 283 (1999) 814e816. [40] A.V. Ramos, M.-J. Guittet, J.-B. Moussy, R. Mattana, C. Deranlot, F. Petroff, C. Gatel, Room temperature spin filtering in epitaxial cobalt-ferrite tunnel barriers, Appl. Phys. Lett. 91 (2007) 122107 (3 pp.). [41] E.Y. Tsymbal, I. Zutic, Handbook of Spin Transport and Magnetism, CRC Press, Taylor & Francis Group, 2012, p. 261. [42] G.X. Miao, M. Muller, J.S. Moodera, Magnetoresistance in double spin filter tunnel junctions with nonmagnetic electrodes and its unconventional bias dependence, Phys. Rev. Lett. 102 (2009) 076601 (4 pp.).