Growth and characterisation of spintronics materials

GROWTH AND CHARACTERIZATION OF SPINTRONIC MATERIALS LIU WEI NATIONAL UNIVERSITY OF SINGAPORE 2004 GROWTH AND CHARACTERISATION OF SPINTRONICS MATERIALS LIU WEI (B. Eng., Tianjin Univ., P. R. China) (M. Eng., Tianjin Univ., P. R. China) A THESIS SUBMITTED FOR THE DEGREE OF MASTER OF ENGINEERING DEPARTMENT OF ELECTRICAL AND COMPUTER ENGINEERING NATIONAL UNIVERSITY OF SINGAPORE 2004 Acknowledgements I would like to make a grateful acknowledgement for my supervisor, Dr. Wu Yihong, for his warm-hearted supervising and kind helps to my researches. I also want to express my gratitude to Dr. William J. McMahon for his valuable advising and cooperation in both my learning and project researches. Meanwhile, I thank to Dr. Han Guchang and Dr. Qiu Jinjun for their teaching and helping in the depositions of ZnO thin films. I am also appreciative to Miss Tay Maureen, Mr. Yong Ming Guang and Mr. Lim Heng Lee for their contributions to this project. Finally, I would like to extend my gratitude to all students and staffs of NSE group, who have given me a lot of helps during last two years. a Table of Contents Acknowledgements a Table of Contents b Summary d List of Tables f List of Figures g Nomenclature j Chapter 1 Introduction 1.1 Background 1.1.1 Spintronics 1.1.2 DMS and half metal 1.2 ZnO based DMS 1.2.1 ZnO based DMS 1.2.2 Fabrication and characterization of ZnO based DMS 1.2.3 The criteria to determine a ferromagnetic semiconductor 1.3 Application of DMS and (Zn, TM)O 1.4 CrO2 1.4.1 CrO2 1.4.2 Growth and characterization of CrO2 1.5 Application of CrO2 1.6 Motivation of researches Reference 1 1 1 1 4 4 4 5 7 7 7 8 8 9 11 Chapter 2 Structures and Properties of (Zn, TM)O and CrO2 2.1 ZnO 2.2 Exchange interactions in DMS 2.2.1 sp-d and d-d exchange coupling 2.2.2 Models of interaction in DMS 2.3 Characterization techniques for DMS 2.4 CrO2 Reference 13 13 14 15 16 18 22 23 Chapter 3 ALCVD growth of ZnCoO 3.1 ALCVD and δ-doping 24 24 b 3.2 Characterization of ZnCoO films 3.2.1 Roughness and thickness 3.2.2 Lattice properties 3.2.3 Transport properties 3.2.4 Magnetic properties 3.3 Summary Reference 27 27 29 31 33 37 39 Chapter 4 Sputtering of Mn and Al doped ZnO 4.1 Sputtering 4.2 Al doped ZnO 4.2.1 Surface roughness and XRD 4.2.2 Resistivity measurements 4.2.3 Effect of annealing 4.3 ZnMnO of homogeneous doping 4.3.1 XRD results 4.3.2 Transport properties 4.3.3 Magnetic properties 4.4 ZnMnO of modulated doping 4.4.1 XRD results 4.4.2 Transport properties 4.4.3 Magnetic properties 4.5 ZnO-CoFe-ZnO: Al structures 4.6 Summary Reference 40 40 40 41 43 45 49 51 55 58 60 60 63 65 65 75 76 Chapter 5 CVD growth of CrO2 and fabrication of Devices 5.1 CVD growth of CrO2 5.2 Characterization of CrO2 thin films 5.2.1 XRD and surface roughness 5.2.2 Magnetic properties 5.2.3 Effect of Ion Milling to the CrO2 thin films 5.3 CrO2 based MTJ, CPP Structures 5.3.1 Fabrication of CrO2 based devices 5.3.2 Magnetic properties 5.4 Summary Reference 77 77 78 78 83 88 91 91 94 96 98 Chapter 6 Conclusions 99 c Summary Spintronics is a newly widely studied area. Within this field, DMS and half metal are two kinds of important materials. In my research, I have tried to grow ZnO based DMS and relative nanostructures with both ALCVD and sputtering methods. In my research of half metallic material, I tried to grow CrO2 thin films with CVD methods. In addition, CrO2 based devices and nanostructures have been fabricated, such as CPP, MTJ and nanowires. In ALCVD growth, I mainly concentrated on the deposition of δ-doping structures. I tried different deposition temperature, different pulse time and different δ-doping structures. Lattice properties, transport properties and magnetic properties have been explored. It seems that the films obtained show no ferromagnetic behavior though hysteresis loops have been observed. From the magnetic properties’ study, ZFC-FC curves show some magnetic phase changes. In sputtering of Mn-doped ZnO DMS, Al is doped as donor to increase the films’ carrier density as in theoretical studies, carrier density is important to achieve sp-d exchange interaction to introduce ferromagnetism to DMS thin films. I grew highly conductive thin films successfully. The resistivity is as low as 10-4 Ω cm. The carrier density is as high as 1021 cm-3. Also the magnetic properties have been studied. There is no ferromagnetic behavior in these films either. As the Al doped ZnO is very conductive, I tried to sputter CoFe discontinuous films between one layer of ZnO and one layer of Al doped ZnO in d order to study possible hopping conduction and localized effects of these CoFe. Hall effect measurement has been taken with finding novel oscillations in carrier density vs. magnetic field curves. MR measurement also has been made to investigate the magnetic effects of these discontinuous films. These oscillations may be due to the large Bs of CoFe nanoparticles. CrO2 is a half metal, which is expected to have the highest spin polarization. This high spin polarization will make it suitable to fabricate magnetic devices and these devices should have good magnetic performances. We tried to grow CrO2 films with CVD method. We changed the pressure, temperature and different set-up to optimize the fabrication process. We got successful depositions of CrO2 thin films. After that, we tried to make devices with CrO2 films. Some results have been obtained, but cannot be repeated due to the relatively large surface roughness. Possible current induced effect has been observed too. e List of Tables TAB. 3.1 Thickness and roughness of (Zn, Co)O films 28 TAB. 3.2 (a) Resistivity of (Zn, Co)O 31 (b) Carrier density of (Zn, Co)O 31 TAB. 4.1 Surface roughness of Al-doped ZnO thin films 42 TAB. 4.2 Resistivity of ZnO thin films 44 TAB. 4.3 Carrier density of Al-doped ZnO thin films after annealing at 300 for 1.5 h 47 TAB. 4.4 Carrier density of homogeneously Mn and Al doped ZnO films (500 ) 57 TAB. 4.5 Structures of modulated doped ZnO thin films 62 TAB. 4.6 Carrier density of modulated doped ZnO thin films 62 TAB. 5.1 Surface roughness of CrO2 thin films 79 f List of Figures FIG. 2.1 Lattice structure of ZnO 14 FIG. 2.2 Hall effect measurement setup 19 FIG. 2.3 Principle of XPS measurement 21 FIG. 2.4 Lattice structure of CrO2 22 FIG. 2.5 Illustrative band structure of CrO2 22 FIG. 3.1 Schematic illustration of ALCVD process 24 FIG. 3.2 (a) XRD pattern of sample No. 1 29 (b) XRD pattern of sample No. 2 29 (c) XRD pattern of sample No. 3 29 (d) XRD pattern of sample No. 4 29 (e) XRD pattern of sample No. 5 29 (f) XRD pattern of sample No. 6 29 FIG. 3.3 (a) M-H and ZFC-FC curves of high temperature deposited ZnCoO film 35 (b) M-H and ZFC-FC curves of low temperature deposited ZnCoO film 36 FIG. 4.1 (a) XRD pattern of Al-doped ZnO thin films No. 1 42 (b) XRD pattern of Al-doped ZnO thin films No. 2 42 (c) XRD pattern of Al-doped ZnO thin films No. 3 42 (d) XRD pattern of Al-doped ZnO thin films No. 4 42 (e) XRD pattern of Al-doped ZnO thin films No. 5 42 FIG. 4.2 Effect of vacuum and H2 annealing to the resistivity of Al-doped ZnO thin films 46 g FIG. 4.3 (a) (b) Illustration of homogeneous doped ZnO thin films 50 Illustration of modulated doped ZnO thin films 50 FIG. 4.4 (a) XRD patterns of homogeneous Mn-doped ZnO thin films grown at room temperature: ZnO 100 W, Al2O3 20 W 52 (b) XRD patterns of homogeneous Mn-doped ZnO thin films grown at room temperature: ZnO 200 W, Al2O3 40 W 52 FIG. 4.5 (a) XRD patterns of homogeneous Mn-doped ZnO thin films grown at 500 ℃: ZnO 100 W, Mn 30 W, Al2O3 40 W 54 (b) XRD patterns of homogeneous Mn-doped ZnO thin films grown at 500 ℃: ZnO 100 W, Mn 50 W, Al2O3 40 W 54 FIG. 4.6 resistivity of homogeneous Mn-doped ZnO thin films grown at room temperature and 500 ℃ 56 FIG. 4.7 (a) M-H curves of homogeneous Mn-doped thin film grown at 500 ℃: ZnO 100 W, Mn 30 W, Al2O3 40 W 59 (b) M-H curves of homogeneous Mn-doped thin film grown at 500 ℃: ZnO 100 W, Mn 50 W, Al2O3 40 W 59 FIG. 4.8 XRD patterns of modulated doped ZnO thin films with different structures 61 FIG. 4.9 Resistivity of modulated doped ZnO thin films FIG. 4.10 M-H curves of ZnO-CoFe-ZnO: Al films 64 67 FIG. 4.11 (a) XRD patterns of ZnO-CoFe-ZnO: Al films: without annealing 68 (b) XRD patterns of ZnO-CoFe-ZnO: Al films: with annealing at 300 ℃, 1.5 h 69 FIG. 4.12 (a) Resistivity, carrier density and Hall resistance of ZnO-CoFe (4 nm)-ZnO: Al 70 (b) Resistivity, carrier density and Hall resistance of ZnO-CoFe (2 nm)-ZnO: Al 71 h (c) Resistivity, carrier density and Hall resistance of ZnO-CoFe (1 nm)-ZnO: Al 72 (d) Resistivity, carrier density and Hall resistance of ZnO-CoFe (0.5 nm)-ZnO: Al 73 FIG. 5.1 Schematic illustration of CVD process to deposit CrO2 film 77 FIG. 5.2 XRD patterns of CrO2 epitaxial films 79 FIG. 5.3 AFM data of CrO2 thin films 81 M-H curve: parallel to plane of CrO2 thin film 84 (b) M-H curve: perpendicular to plane of CrO2 thin film 84 (c) MR curve of CrO2 thin film 85 M-H curve of double layer of CrO2 films 87 MR curve of double layer of CrO2 films 87 FIG. 5.6 (a) MR curves of effect of Ion Milling to the CrO2 films: before Ion Milling 89 (b) MR curves of effect of Ion Milling to the CrO2 films: Ion Milling for 3 mins 89 (c) MR curves of effect of Ion Milling to the CrO2 films: Ion Milling for 8 mins 90 FIG. 5.4 (a) FIG. 5.5 (a) (b) FIG. 5.7 Process to make CrO2 based MTJ, CPP and nanowires 92 FIG. 5.8 Microscope image of CrO2 MTJ structure 93 FIG. 5.9 (a) (b) M-H curve of CrO2 based MTJ sample 95 MR curve of CrO2 based MTJ sample 95 FIG. 5.10 Four Probe MR measurement 96 FIG. 5.11 V-I curve of CrO2 nanopillar 98 i Nomenclature DMS diluted magnetic semiconductor ALCVD atomic layer chemical vapor deposition XRD X-ray diffraction SQUID superconducting quantum interference devices VSM vibrating-sample magnetometer MTJ magnetic tunneling junction CPP current-perpendicular-to plance TM transition metal GMR giant magnetoresistance MR magnetoresistance AHE anomalous Hall effect CVD chemical vapor deposition RKKY Rudermann, Kittel, Kasuya, Yoshida AFM atomic force microscopy EM electromagnetic EXAFS extended X-ray absorption fine structure XPS X-ray photoelectron spectroscopy SIMS secondary ion mass spectrometry AMR anisotropic magnetoresistance RIE reactive ion etching VO oxygen vacancy Zni Zn interstials j TCO transparent conductive oxide Ra arithmetic average roughness k Chapter 1 Introduction 1.1 Background 1.1.1 Spintronics Spintronics uses both carrier and its spin, which gives us new choices and functionality in electronic devices[1] [2] [3]. In past years, the traditional electronic devices use only carrier to transport and process electric signal. But the development of electrical industry and technology requires higher performance devices: smaller, faster, cheaper and less energy consuming. The smaller dimension is one of the basic and main requirements. Using both spin and charge of carriers can effectively shrink the size of devices, lower device power consumption and enhance running speed. In addition, we can integrate computation and storage components together, which can be used in high-density data storage application. 1.1.2 DMS and half metal DMSs and half metals will play important roles in spintronics. 1 In DMS, transition metal atoms substitute the lattice sites of host semiconductor. The first DMSs appeared in 1960s to combine both electronic materials and magnetic materials, such as EuSe, EuS [4]. In early 1980’s, - compounds based DMS (CdTe:Mn, ZnSe:Mn etc.) appeared. But most of above materials are not ferromagnetic. Most groups cannot get ferromagnetic semiconductors but other kinds of magnetic phases, i.e. spin glass, paramagnetic phase, which have been obtained. Nevertheless, in recent years, DMS has become a hotly studied field and several more kinds of magnetic semiconductor (GaAs: Mn, ZnO: Co, TiO2: Co, Ge: Mn etc.) have been produced. However, although many groups claimed that they had successfully grown ferromagnetic semiconductor films, only GaAs: Mn [5] [6] is widely believed to be ferromagnetic. Another major problem of DMS research is the relatively low Curie temperatures (Tc), for example Tc of (Ga, Mn)As is only around 120K, which is much lower than room temperature. But room temperature Curie temperature is a necessity for applications of DMS in industry. Nowadays, - and - group semiconductors are chosen to fabricate DMS as some theoretical calculations predict that possible 2 ferromagnetic semiconductors with Tc higher than room temperature can be obtained from these materials [7] [8]. Furthermore, other types of materials are used too, e.g. GaN [9], TiO2 [10] [11] and Ge [12] are used as candidates of host semiconductor, whose Tc is calculated as higher than room temperature too [7]. Half metal is an interesting and important material in spintronics, in which the carriers of one spin direction show semiconductor properties, while carriers of the other spin direction show metallic properties. Different half metallic materials have different mechanisms behind. Totally, there are about four kinds of half metal [13]. For example, Fe3O4 and CrO2 are two kinds of half metal. The conduction of Fe3O4 is by the hopping from one Fe site to another with the same spin, which causes the half metallic properties. On the other hand, the hybridization between s and p electrons makes CrO2 to be a 100% polarized half metal. A third class of half metals have both localized spin up and delocalized spin down carriers (a larger effective mass), e.g. (La0.7Sr0.3)MnO3. In addition, the forth class of half metals are semimetals, which have a great disparity in effective mass between electrons and holes, e.g. Tl2Mn2O7. A semimetal has a 3 small overlap between valence and conduction bands and it has equal numbers of electrons and holes. 1.2 ZnO based DMS 1.2.1 ZnO based DMS ZnO is a wide band gap (3.37 eV) oxide semiconductor, which has applications in electro-optic fields because of its high exciton binding energy (60 meV). According to theoretical calculations [14] [15] [7], it was reported that (Zn, TM)O could be a room temperature DMS. A lot of groups have taken this research [16] [17] [18] [19] [20]. Some of them claimed that they have obtained ferromagnetic semiconductor, but some of them reported that they only got some materials of other magnetic phase [21]. 1.2.2 Fabrication and Characterization of ZnO based DMS In growth of ZnO based DMS, Co and Mn were used as magnetic dopants [21] [22] [23] [24] [25] [16] [26] [27]. Also other magnetic materials were used as dopants [18]. Transport and magnetic properties were studied [28] [29]. Furthermore, different structures were made with ZnO based DMS [30]. ZnO doped with Co has been reported as showing ferromagnetism at room temperature. Other 4 properties of (Zn, TM)O are also thoroughly studied, e.g. the transport properties, the annealing effects and so on. However, there are only some successful cases of room temperature ferromagnetic semiconductor up to now. Other magnetic results come from spin glass state or ferromagnetism, which is due to Co clusters. The inconsistent magnetic properties of ZnO based DMS are due to the different growth methods or different conditions. Several methods have been used to fabricate ZnO based DMS, i.e. PLD, MBE, and sputtering. We used ALCVD and sputter technologies to fabricate ZnO based DMS. ALCVD is a technology to obtain high quality film, which will be explained in detail in later section, whose main usage is to fabricate high-K gate material in new CMOS devices. One main advantage of this technology for making DMS is the dopant quantity can be precisely controlled. Magnetic properties, lattice structures and transport properties of our samples were studied. XRD, Hall effect, SQUID measurements are three main methods to characterize these samples. 1.2.3 The criteria to determine a ferromagnetic semiconductor 5 Up to now, the determination of ferromagnetic semiconductor remains as a problem. The criteria to decide whether a piece of magnetically doped semiconductor is ferromagnetic are not quite clear. Mainly, this difficulty comes from the fact that normally the M-H and MR measurement cannot distinguish the magnetization from ferromagnetic impurity clusters. The XRD patterns and TEM cannot exclude the nanoscale impurity segregation because of the limited resolutions of either method. Moreover, the observed hysteresis loop may come from spin glass too. However, AHE should be a clearer sign of ferromagnetic semiconductor [31] [32]. The Hall resistivity changing with field can indicate the magnetization dependent properties, which will be saturated at high field. While, the spin glass state can be excluded by ac-susceptibility measurement. Also, the spin glass DMS has unique M-T curves. In fact, the magnetic properties of a ferromagnetic semiconductor should be relative to lattice structure and should be variable with carrier densities and thus, change with dopant quantity or an electric field. All these properties need to be examined in an attempt to fabricate a ferromagnetic semiconductor. 6 1.3 Applications of DMS and (Zn, TM)O DMS has attracted many researchers’ interests because it is promising in future electronic devices, mainly in the spintronic field. With DMS, it is possible the spin degree freedom being used. Because of the carrier density-related magnetism, the ferromagnetism of the device can be controlled by an electric field. Spin-FET and newly designed transistor are two most promising applications of DMS. MRAM and other magnetic data storage devices can be other applications of DMS too. Spin injection is another research area in recent years as 100% polarized current transportation is needed in a spintronic device. DMS can be a candidate as source injecting spin-polarized carriers to a nonmagnetic semiconductor. The magneto-optical effect of the ZnO based DMS can be used in a variety of applications such as magneto-optic disk for memories and optical isolators and circulators for optical communication. The advantage of this material is the magnitude of magneto-optical effect can be two orders larger than that of non-magnetic semiconductors. 1.4 CrO2 1.4.1 CrO2 7 CrO2 is the only half-metallic dioxide, which has been used as a magnetic storage material for many years. Half metals have the advantage to fabricate magnetic devices, such as MTJ, CPP structures with respect to their theoretically 100% polarization, which will enhance the GMR effect discovered by Peter Gruenberg from the KFA research institute in Julich, Germany, and Albert Fert from the University of Paris-Sud, 1988. Another advantage of CrO2 thin film for fabricating MTJ structure is that we can use the natural oxide layer (Cr2O3), mentioned in many literature [33] [34] [35], as the insulating oxide layer of a MTJ. 1.4.2 Growth and characterization of CrO2 We used the two-chamber CVD method to deposit CrO2 films epitaxially, which will be described in detail in later section. The methods to characterize our samples include XRD, VSM, in which a gradiometer picks up the moment change by measuring the magnetic induction in space with and without the sample being measured. 1.5 Application of CrO2 CrO2 is the only half-metallic dioxide. Theoretically, the polarization of CrO2 is 100% [36]. However, the highest measured 8 value is near 100% [37]. The transport properties of CrO2 are theoretically simulated and tested by different groups [38] [39] [40] [41]. The most important application of half metals is the source of polarized current. Some researchers tried to use metal to do this, but there are some problems, such as interface scattering and conductivity mismatching. The 100% polarized carriers are desiring advantage to obtain high MR ratio devices, for example, MTJ and CPP structures. If we use half metal as a magnetic layer of a MTJ or a magnetic layer of spin valve, theoretically, we should get higher MR ratio. 1.6 Motivation of researches Although the fabrication of ferromagnetic (Zn, TM)O thin films, up to now, is still a challenge because of the inconsistent results of different groups and the lack of a mature theory to explain the source of ferromagnetism properties, many researchers reported successful growth of both Co and Mn doped ZnO DMS. In addition, besides sputtering, we tried a new technology, ALCVD, which can grow high quality films and precisely control the doping level. Moreover, the new δ-doping and modulated doping structures, which means there is a 9 magnetic impurity layer or a heavy magnetic doped layer between several layers of ZnO, make our DMS fabrication possible to obtain some novel and good results. At present, the deposition methods and electric transport properties of CrO2 have been studied via high and low pressure CVD methods and Hall effect measurement. However, there are only few groups that have taken the research on CrO2 magnetic devices fabrications, i.e. MTJ, CPP structures and nanowires. Our new approach is to use E-beam Lithography method to fabricate CrO2 based nanowires to study the possible new magnetic transport properties due to the domain shrinkage. In summary, our research on ZnO material growth using ALCVD and sputtering should be able to get some good results. At the same time, the strategies of δ-doping and modulated doping should enable us to observe some new physical phenomena. On the other hand, the special magnetic properties of CrO2 make it a good candidate to make new magnetic devices. The reported successful low pressure CVD growth of CrO2 thin films makes our proposed research possible. 10 (1) (2) (3) (4) (5) (6) (7) (8) (9) (10) (11) (12) (13) (14) (15) (16) (17) (18) (19) (20) (21) (22) (23) (24) (25) (26) (27) (28) (29) (30) (31) (32) (33) (34) (35) (36) (37) (38) (39) G. A. Prinz, Science, 282, pp.1660. 1998. R. Fiederling and M. Keim, Nature, 402, pp.787. 1999. Y. Ohno and D. K. Young, Nature, 402, pp.790. 1999. G. A. Medvedkin and T. Ishibarshi, Jpn. J. Appl. Phys., 39, pp.L949. 2000. H. Ohno, J. Magn. & Mater., 200, pp.110. 1999. T. Jungwirth and J. Sinova, Applied Physics Letters, 83, pp.320. 2003. T. Dietl and H. Ohno, Physical Review B, 63, pp.195205. 2001. K. Sato and H. K. Yoshida, Physica B, 308, pp.9040. 2001. G. T. Thaler, M. E. Overberg, and B. Gila, Applied Physics Letters, 80, pp.3964. 2002. Z. Wang and W. Wang, Applied Physics Letters, 83, pp.518. 2003. W. K. Park and R. J. Ortega-Hertogs, Journal of Applied Physics, 91, pp.8093. 2002. W. K. Park and A. T. Hanbicki, Science, 295, pp.652. 2002. J. M. D. Coey and M. Venkatesan, Journal of Applied Physics, 91, pp.8345. 2002. K. Sato and H. Katayama-Yoshida, Phys. Stat. Sol. (b), 229, pp.673. 2002. K. Ueda and H. Tobata, Applied Physics Letters, 79, pp.988. 2001. Z. Jin and T. Fukumura, Applied Physics Letters, 78, pp.958. 2001. J. H. Kim and H. Kim, Journal of Applied Physics, 92, pp.6066. 2002. D. P. Norton and S. J. Pearton, Applied Physics Letters, 82, pp.239. 2003. S.-J. Han and J. W. Song, Applied Physics Letters, 81, pp.4212. 2002. H.-J. Lee and S.-Y. Jeong, Applied Physics Letters, 81, pp.4020. 2002. J. H. Park and M. G. Kim, Applied Physics Letters, 84, pp.1338. 2004. K. Rode and A. Anane, Journal of Applied Physics, 93, pp.7676. 2003. S. G. Yang and A. B. Pakhomov, IEEE Transaction on Magnetics, 38, pp.2877. 2002. S. W. Yoon and S.-B. Cho, Journal of Applied Physics, 93, pp.7879. 2003. W. Prellier and A. Fouchet, Applied Physics Letters, 82, pp.3490. 2003. V. A. L. Roy and A. B. Djurisic, Applied Physics Letters, 84, pp.756. 2004. X. M. Cheng and C. L. Chien, Journal of Applied Physics, 93, pp.7876. 2003. J. Han and M. Shen, Applied Physics Letters, 82, pp.67. 2003. T. Nagata and A. Ashida, Jpn. J. Appl. Phys., 41, pp.6916. 2002. T. Edahiro, N. Fujimura, and T. Ito, Journal of Applied Physics, 93, pp.7673. 2003. H. Toyosaki, T. Fukumura, and Y. Yamada, Natural Materials, 3, pp.221. 2004. N. Manyala, Y. Sidis, and J. F. Ditusa, Natural Materials, 3, pp.225. 2004. A. Gupta, X. W. Li, and G. Xiao, Applied Physics Letters, 78, pp.1894. 2000. R. H. Cheng, T. Komesu, H.-K. Jeong, L. Yuan, S. H. Liou, B. Doudin, and P. A. Dowben, Phys. Lett. A, 302, pp.211. 2002. J. Dai and J. Tang, Applied Physics Letters, 77, pp.2840. 2000. K. P. Kamper, W. Schmitt, and G. Guntherodt, Physical Review Letters, 59, pp.2788. 1987. R. Wiesendanger, H.-J. Guntherodt, and G. Guntherodt, Physical Review Letters, 65, pp.247. 1990. I. I. Mazin and D. J. Singh, Physical Review B, 59, pp.411. 1999. K. Suzuki and P. M. Tedrow, Physical Review B, 58, pp.597. 1998. 11 (40) (41) S. S. Manoharan, D. Elefant, G. Reiss, and J. B. Goodenough, Applied Physics Letters, 72, pp.984. 1998. L. Ranno, A. Barry, and J. M. D. Coey, Journal of Applied Physics, 81, pp.5774. 1997. 12 Chapter 2 Structures and Properties of (Zn, TM)O and CrO2 2.1 ZnO The lattice structure of ZnO is wurtzite. This structure can be considered as the close-packed hexagonal structure with a basis of two atoms. Each Zn ion is bond to four O ions in the tetrahedral bond of sp3 hybridization, vice verse. Because the dopant magnetic ion, i. e. Co, only substitutes the Zn ion, the lattice structure of (Zn, Co)O should be the same as that of ZnO. Figure 2.1 is the lattice structure of ZnO. ZnO is a wide band gap - compound semiconductor, which has a direct band gap at the Γ point in the energy band. The s-orbital of the Zn2+ and the p-orbital of the O2- form the conduction band and the valence band. Normally, the interstitial Zn atoms or O vacancies introduce a donor level of 0.05eV under the conduction band edge. So it is difficult to get p-type ZnO because of the defects in the crystal. Moreover, ZnO has high exciton binding energy (60meV), which results in less trapping of carriers and higher luminescent efficiencies. This should, in principle, favor efficient excitonic emission at room temperature. After magnetic doping, the sp-d interaction, which will be 13 explained in detail below, would change both the band structure of ZnO and the electronic state of carriers, which will spin-polarize the current and bring some effects, i. e. Faraday Rotation, GMR effect, etc. FIG. 2.1 Lattice Structure of ZnO 2.2 Exchange interactions in DMS Theoretically, the magnetic properties of DMS have been explained with sp-d exchange coupling and RKKY interactions. Sp-d coupling describes the direct interaction between d electrons of transition metal ions and s or p electrons in conduction band. This coupling makes the spin of conduction carriers polarized. Meanwhile, the RKKY interaction describes the interaction between spins of two transition metal ions via those conduction carriers. This interaction is supposed to introduce ferromagnetism in DMS. 14 2.2.1 sp-d and d-d exchange coupling The ferromagnetic properties of DMS come from the exchange interactions in materials. The lattice scale coupling between the spins of localized TM ions will bring ferromagnetic behavior. In DMS, there are two kinds of exchange interaction between the localized magnetic ions and carriers: strong and weak exchange interaction. The sp-d exchange interaction is the strong one, which is the magnetic coupling between transition metal ion and the spin of the charge carriers. The second kind of exchange coupling is a weak coupling (d-d). It is directly between two magnetic ions (i. e. Co to Co). There are two mechanisms, which cause the sp-d exchange interaction: the normal exchange coming from the 1/r coulomb interaction potential and the kinetic mixing of the sp band and d electrons due to the hybridization of their wave functions. The first potential tends to align the spins of the band electrons with the spins of transition metal ions. This potential is only related to the interaction between two electrons, i.e. conduction electron and d electron of a transition metal ion and does not depend on the orientation and lattice structure of the host material. 15 In the sp-d exchange interaction, s-d interaction is weaker and more localized than the p-d one. Although the s-orbital of the conduction band does not mix with the d-orbital, it is influenced by the magnetic ion. On the other hand, the p-d exchange constant is much stronger, which is dominated by the kinetic exchange contribution. In DMS, the sp-d exchange coupling induces the giant Zeeman splitting in semiconductor band structure. Zeeman splitting happens when an external magnetic field induces splitting of the semiconductor band structure. In DMS, the effective magnetic field on the sp-band electrons is amplified by the magnetic moment of the transition metal ion through sp-d exchange interaction. Moreover, sp-d exchange coupling also induces the magneto-optical effects, such as, Farady effect and Kerr effect, magnetic field-induced metal-insulator transition and the effects of the bound magnetic polaron. 2.2.2 Models of interaction in DMSs RKKY interaction is a model to describe the interaction between a local magnetic impurity and the surrounding electron gas. RKKY is caused by the superposition of the charge density oscillations of the 16 spin up and spin down electrons giving rise to a spin density oscillation. K. Sato’s model [1] [2] predicted the possibility of ferromagnetic DMS at room temperature by the first principle calculations based on the local density approximation. In T. Dietl’s model [3], the tendency toward ferromagnetism has been explained with a mean-field picture, in which uniform mobile carrier spin polarization mediated a long-range ferromagnetic interaction between the magnetic ions. Tc can be obtained from the competition between the ferromagnetic and anti-ferromagnetic interactions. It assumes two spin subsystems, carrier spins and localized spins at magnetic ions, interacting through the sp-d interaction. Having a nonzero magnetization increases the free energy of the localized spin system, but reduces the energy of the carrier systems via spin-splitting of the bands. And the free energy penalty reduces as temperature is reduced and it balances with the energy gain of the carrier system at T=Tc. This model explains the relationship of Tc as a function of ion concentration and hole concentration. Based on this model, T. Dietl predicted that Tc of GaN and ZnO can be raised to above 300K [3]. 17 2.3 Characterization techniques for DMS XRD, AFM, Hall Effect, SQUID and XPS are all techniques, which are used to characterize DMS. XRD is used to determine lattice structures and orientations. Since an atom can scatter X-ray, and if many atoms are together, the scattered waves from all the atoms can interfere. If the scattered waves are in phase, they interfere in a constructive way and we get diffracted beams in specific directions. These directions are described by Bragg’s law nλ = 2d sin θ (2.1) where λ is the wavelength of the X-ray; n is the order of diffraction; d is the interplanar spacing of analyzed crystal; θ is the angle between crystal surface and incident and diffraction rays. When a scanning is made from, i. e. 20 to 90°, with a fixed wavelength, a peak of diffraction intensity will appear when the interplanar distance and θ fit the Bragg’s law. Thus, the XRD patterns show the lattice structure, which is determined by the specific parameter d. AFM is a technique to display surface morphology of films. Typically, AFM uses a tip to probe the surface of a sample, in which the force between the tip and sample is detected by a sensor. Once the 18 force data is treated by computer software, the surface morphology can be displayed. The force is the van der Waals force. The most important information of DMS that AFM can give is the roughness of thin films. If the film roughness is large, it is difficult to use it in an electronic device. Hall effect is a widely used method to characterize semiconductor thin films. Figure 2.2 is an illustrative picture of Hall effect measurement. In a Hall effect measurement, a current I flows in the sample and a perpendicular magnetic field is applied to the current. The carriers, i.e. electrons, will be influenced by both the electric field, which drives the current and the Lorentz force. Hence, the carriers will move to one side of the film by a velocity of v’. As the accumulation of carriers at the both sides of the sample, a Hall voltage can be detected as in the Figure 2.2. By this measurement, we can get carrier density, resistivity, and mobility of carriers in semiconductors. Lorentz force is the main mechanism of Hall effect in a normal semiconductor. H V’ I v V’ I v V FIG. 2.2 Hall effect measurement setup 19 However, in DMS, the carriers are influenced by both Lorentz force and the EM force originated from spin-orbital coupling. This EM force is attributed to AHE. AHE relates the Hall resistance (RHall) to temperature and amplitude of magnetic field as they affect the magnetization [4]. The existence of AHE is a sign of ferromagnetism. The EM force in AHE comes from the spin-orbit interaction between the conduction electron and the localized moment. Asymmetric scattering can occur due to the coupling between the orbital angular momentum of carrier and the spin angular momentum of the localized scattering center, i.e. the transition metal ion in DMS. SQUID is an important approach to explore the magnetic properties of a magnetic material. Nowadays, SQUID provides the highest resolution to moment measurement. The essential part of SQUID setup consists of a superconducting ring with a small insulating layer known as the “weak link”. The principle is the flux passing through the ring is quantized once the ring has gone superconducting but the weak link enables the flux trapped in the ring to change by discrete amounts. Changes in the pick-up voltage occur as the flux is incremented in amounts of ∆Φ=2.067×10-15 Wb [5]. 20 XPS is used to determine the composition of a thin film material. XPS is based on the photoelectric effect, in which a primary X-rays eject photoelectrons from the material (Figure 2.3). Ejected Electron Conduction band EF Valence band EVacuum EV EC Primary X-rays Core levels EL2, 3 EL1 EK FIG. 2.3 Principle of XPS measurement XPS is used to acquire the information on elemental composition and the chemical bonding states of the DMS. The measured energy of ejected electron (Esp) is related to the binding energy (Eb), which depends on atomic composition and chemical environment. Eb is Eb = hν − E sp − qφ sp (2.2) in which hν is the energy of primary X-rays, φsp is the work function of spectrometer. 21 2.4 CrO2 FIG. 2.4 Lattice structure of CrO2 CrO2 is the only stoichiometric binary oxide that is a ferromagnetic half metal. Figure 2.4 is the lattice structure of CrO2, which is the tetragonal rutile. Each oxygen atom has three chromium neighbors and each chromium is octahedrally coordinated by oxygen with two short apical bonds and four longer equatorial bonds [6]. The band structure of CrO2 (Figure 2.5) Ef shows that the carriers are 100% spinpolarized. This is because of the fact: the 4s states are pushed above Ef by hybridization FIG. 2.5 Illustrative band structure of CrO2 with the O(2p) states; the Cr d levels lie close to the top of the O 2p band; the Fermi level lies in the half-full dyz±dzx band [7]. All these make that the electrons of one spin direction are 22 metallic, but the electrons of other spin direction show the properties of semiconductor. The transport properties of CrO2 can be described by two-band model. From Hall effect measurement [8], it is found that there are two types of conduction mechanisms in CrO2, which means that both hole and electron are carriers in CrO2. In Watts et al. work [8], the lowtemperature Hall effect exhibits a sign reversal from positive to negative as the magnetic field is increased above 1T. This is a normal effect when there are both electrons and hole as carriers, which can be explained by the presence of highly mobile holes as well as a much larger number of less mobile electrons. (1) (2) (3) (4) (5) (6) (7) (8) K. Sato and H. Katayama-Yoshida, Semicond. Sci. Technol., 17, pp.367. 2002. K. Sato and H. K. Yoshida, Physica B, 308, pp.9040. 2001. T. Dietl and H. Ohno, Physical Review B, 63, pp.195205. 2001. H. Ohno, J. Magn. & Mater., 200, pp.110. 1999. D. Jiles, Introduction to Magnetism and Magnetic Materials, pp.60, London; New York: Chapman and Hall, 1991. J. M. D. Coey and M. Venkatesan, Journal of Applied Physics, 91, pp.8345. 2002. I. I. Mazin and D. J. Singh, Physical Review B, 59, pp.411. 1999. S. M. Watts, S. Wirth, and S. v. Molnar, Physical Review B, 61, pp.14. 2000. 23 Chapter 3 ALCVD Growth of ZnCoO 3.1 ALCVD and δ-doping We used ALCVD to deposit (Zn, Co)O thin films. This technology uses layer by layer reaction to form a surface controlled deposition. This layer by layer reaction is made of many cycles. Within FIG. 3.1 Schematic illustration of ALCVD process each cycle, the carrier gas (N2) brings different precursors into reactor one by one. Each of these feedings of precursors is called a pulse. Normally there are two pulses in one cycle. Between two different pulses, there is a purging time, in which purging gas (N2) will clean the reactor to prevent a CVD mode reaction. So each atomic layer formed 24 in this sequential process is a result of saturated surface-controlled reactions. This ALCVD process provides excellent step coverage and dense films with no pinholes. Compared to other methods for deposition of DMS, ALCVD process has another advantage of precisely control of doping levels. Figure 3.1 shows a typical process of ALCVD cycle. In Figure 3.1 (a), the first precursor is carried to the substrate by the carrier gas. In Figure 3.1 (b), by an absorption process, the precursor forms a layer at the surface of the substrate. At the same time, the purging gas will bring away the unabsorbed precursor. After that, in Figure 3.1 (c), another flow of carrier gas will bring the second precursor to the substrate, to which some chemical reaction will happen with the first absorbed precursor (Figure 3.1 d). After the reaction, it forms an atomic-layer accuracy film at the surface of the substrate. Meantime, the purging gas will take the waste material to exhaust system. We use an ASM F120 ALCVD system to fabricate ZnO films. Totally, six different precursors can be fed in this system. Either feeding two sources (for example, one is Zinc, the other one is Co) in one pule within one cycle at the same time or inserting a single 25 magnetic dopant cycle, we can make magnetic doping to the semiconductor. We deposit ZnO film on sapphire (001) substrates. The Zn source is Diethylzinc (DEZ), a liquid source, which flows into the reactor at its own vapor pressure when the reactor is in vacuum conditions (around 1-2 mbar). The Cobalt source is [Cobalt (Ⅱ) Acetylacetonate], which is a dark red powder. This source is heated up to 90 , at which it vaporizes and is carried into the reactor by a flow of N2. For the oxygen source, we use a 300 ms flow of 60% ozone, 40% oxygen in addition to water, which flows on its own vapor pressure at 18 . The flow rate is controled by a manual valve. We grow at two different temperatures, 320 and 150 . In order to get epitaxial growth, we first deposit several hundred cycles of pure ZnO, then run several hundred δ-doping cycles of the form [(Zn-O)m-CoO]n, where m=1-5 and n is several hundred. The doping strategy we tried is δ-doping. It means that between several monolayer of ZnO, we insert a Cobalt cycle to form a full layer of Co. With the different number of ZnO layers between two Co layer, we can get different doping levels and precisely control the doping level. 26 3.2 Characterization of ZnCoO films 3.2.1 Thickness and roughness Thickness is a basic parameter of ZnO films. Because both pure and doped ZnO films are transparent films, we used an Ellipsometer to measure the thickness of ZnO films. In Table 3.1, the thickness of six samples is given as well as the doping structures. Four of them (No. 1-4) were deposited at high temperature (320 ). The other two (No. 5, No. 6) were deposited at low temperature (150 ). The pulsing times in each cycle were changed besides the different deposition temperatures. The pulsing time of the high temperature deposition is 300 ms while that of the low temperature deposition is 2 s. It was necessary to lengthen the pulsing time for the low temperature deposition in order to saturate each layer with adsorbed DEZ. It was visibly apparent that the doping was successful, because the color of the doped films is somewhat bluer than the more transparent pure ZnO film. This color difference may affect the calibration of the Ellipsometer, so the thickness of the doped samples can only be approximated, but it was clear that more cycles resulted in a thicker film. 27 TAB 3.1 Thickness and roughness of (Zn, Co)O films Number Structure Thickness (nm) Ra (nm) 1 ZnO1500 144.3 0.997 2 ZnO500, [ZnO(3ML)/Co(1ML)]600 220.6 0.281 3 ZnO500, [ZnO(4ML)/Co(1ML)]600 787.0 4 ZnO500, [ZnO(5ML)/Co(1ML)]600 800.2 1.883 5 ZnO1000 71.20 2.853 6 ZnO300, [ZnO(3ML)/Co(1ML)]300 93.83 Surface roughness of electronic material films is important. Large roughness will cause extra band variation and scattering at the interface of electronic material films, which will degrade the performance of the devices. Thus we measured the roughness of these films. We used an AFM system to do that. Table 3.1 also lists the AFM scan results of sample No. 1, No. 2, No. 4 and No. 5. From the data, the surface roughness of higher temperature (320 ) deposited thin films is lower. And from the surface morphology, the higher temperature deposited films are more uniform compared with the surface of No. 5, which is caused by the fact of higher energy of atoms in high temperature depositions. 28 3.3.2 Lattice properties Figure 3.2 displays the XRD measurement results of above 6 1 (a) No. 1 Δ2θ=-0.08 ZnO (004) Al2O3 (001) 10 Al2O3 (001) 2 ZnO (002) 3 10 ZnO (002) thin films. 1 Al O (001) 2 3 ZnO (101) 2 10 ZnO (002) 3 10 1 2 10 ZnO (002) 3 10 Al O (001) 2 3 10 1 2 10 ZnO (101) 3 10 Al O (001) 2 3 10 ZnO (002) Intensity (counts/s) 10 1 10 2 10 (c) No. 3 Δ2θ=-0.02 (d) No. 4 Δ2θ=-0.12 (e) No. 5 Δ2θ=-0.08 ZnO (002) Al2O3 (001) ZnO (004) 10 (b) No. 2 Δ2θ=-0.38 ZnO (004) 2 ZnO (004) 3 10 ZnO (004) 10 (f) No. 6 Δ2θ=-0.08 1 10 20 40 60 80 2 Theta (degree) FIG. 3.2 XRD patterns of pure and Co-doped ZnO films 29 The ZnO 34.4˚ (002) peak appears in all six patterns though there are some small shifts in the peak positions due to the lattice mismatch and defects or doping [1] [2]. The shifts in peak position are larger than expected if Co is replacing Zn in the lattice, and this is likely to be caused by one of two things: the Co is going into interstitial positions rather than substituting or there is an increase in defect densities for Co doped samples. The 41.675˚ (001) peak of sapphire can also be observed. In addition, we can see a smaller ZnO 72.514˚ (004) peak in most patterns. In two of the samples (No.3, 5), the (101) peak of Zn can be seen, indicating that those films are not completely epitaxial. This sort of growth is likely due to a diminished flow rate of DEZ between growths. In general it is difficult to maintain a constant flow rate between various runs because the DEZ tends to build up at the valves. Finally, the low temperature doped sample has a considerably lower peak than the undoped low temperature sample, despite the similar thicknesses as measured by the Ellipsometer. This is likely due to a somewhat amorphous film. All the XRD peaks are indication of successful deposition of pure and doped ZnO films. 30 3.2.3 Transport properties The resistivity and carrier density of the films are determined by Hall effect measurements using van der Pauw geometry. TAB. 3.2 (a) Resistivity of (Zn, Co)O Structure Number Resistivity (×10-2 ohm cm) 7 ZnO1500 3.319 8 ZnO500, [ZnO(2ML)/Co(1ML)]600 1.191 9 ZnO1000 0.2447 10 ZnO300, [ZnO(3ML)/Co(1ML)]300 0.2233 11 (Before ZnO500, [ZnO(4ML)/Co(1ML)]600 1.035 Annealing) 11 (After 1.055 Annealing) 12 (Before ZnO300, [ZnO(3ML)/Co(1ML)]300 0.2233 Annealing) 12 (After 7.049 Annealing) (b) Carrier density of (Zn, Co)O Structure Number Carrier Density (×1020 1/cm3) 7 ZnO1500 0.1275 8 ZnO500, [ZnO(2ML)/Co(1ML)]600 0.228 9 ZnO1000 1.504 10 ZnO300, [ZnO(3ML)/Co(1ML)]300 1.637 11 (Before ZnO500, [ZnO(4ML)/Co(1ML)]600 0.1969 Annealing) 11 (After 0.1912 Annealing) 12 (Before ZnO300, [ZnO(3ML)/Co(1ML)]300 1.637 Annealing) 12 (After 0.2144 Annealing) 31 The resistivity and carrier density of four samples are listed in Table 3.2 together with their deposition structures. No. 7 and No. 8 are deposited at 320 . No. 9 and No. 10 are deposited at 150 . We can get the conclusion that samples grown at high temperature have resistivity in the order of 10-2 Ω cm compared to 10-3 Ω cm for low temperature deposited films. The lower resistivity of low temperaturedeposited samples should come from more defects in films, which bring carriers. For low temperature deposition, the resistivity decreases when Co dopant is introduced. Table 3.2 also shows increases in the number of carriers when Co dopant is added. The carrier densities are in the order of 1020 (n type) for low temperature depositions and 1019 (n type) for high temperature ones as low temperature-deposited samples have more defects. We also investigated annealing effect to our ALCVD grown ZnCoO thin films. A low temperature and a high temperature film were annealed at 150 for 5 minutes plus warming and cooling time of the oven (~20 min). Table 3.2 shows the resistivity and carrier density measurements of these samples (No. 11 is a high temperature, 320 , deposited film; No. 12 is a low temperature, 150 , deposited 32 film). From Table 3.2, the high temperature deposition sample initially has a higher resistivity than the low temperature deposited sample, but this reverses after annealing due to healing effect of annealing. Combined with XRD data, the peak widths and intensities of low temperature-deposited sample are wider and lower. This suggests that the doped low temperature samples are more amorphous, and contain a larger number of defects, which bring more carriers. 3.2.4 Magnetic properties The main applications of DMS are in the spintronic or magnetic storage area, so it is very important to understand the magnetic properties of these samples. We use SQUID to test our high and low temperature deposition samples. The structure of the high temperature (320 (Figure 3.3 a) is buffer layer of ZnO ) deposited film 500 cycles, [ZnO(4ML)/Co(1ML)]600; the structure of the low temperature (150 ) deposited film (Figure 3.3 b) is buffer layer of ZnO 300 cycles, [ZnO(3ML)/Co(1ML)]300. Figure 3.3 shows the M-H and M-T curves of these two samples. From the M-H curves in Figure 3.3 (a) and (b), we can see that at 33 lower temperatures, such as 2 K, the moment is larger than that at high temperatures. More importantly, the samples show hysteresis properties over the range of several kiloOested. Because of the low transition temperature (~2K), this hysteresis most likely comes from a spin glass state though it is not definite and more measurements are needed. 34 -5 Moment (emu) 4.0x10 -5 2.0x10 400 K 0.0 300 K -5 -2.0x10 2K -5 -4.0x10 -6000 -4000 -2000 0 2000 4000 6000 Field (Oe) M-H -6 -3.0x10 -6 Moment (emu) -4.0x10 FC 100 Oe -6 -5.0x10 -6 -6.0x10 -6 -7.0x10 ZFC -6 -8.0x10 -6 -9.0x10 0 100 200 300 400 Temperature (K) ZFC-FC (a) 35 -5 4.0x10 -5 3.0x10 -5 Moment (emu) 2.0x10 -5 1.0x10 400 K 0.0 -5 -1.0x10 300 K -5 -2.0x10 2K -5 -3.0x10 -5 -4.0x10 -6000 -4000 -2000 0 2000 4000 6000 Field (Oe) M-H -6 Moment (emu) -3.0x10 -6 -3.5x10 FC 100Oe -6 -4.0x10 ZFC -6 -4.5x10 0 100 200 300 400 Temperature (K) ZFC-FC (b) FIG. 3.3 M-H and ZFC-FC curves of high temperature (a) and low temperature (b) deposited ZnCoO thin films 36 The moment vs. temperature graphs are plotted by firstly decreasing the temperature of the samples to 2 K. During the first cycle, no field is applied (ZFC) as the temperature is slowly raised to 400 K. After that, a magnetic field (100 Oe) is applied (FC) as the sample is cooled down to 2 K again, and the second curve is plotted. From the ZFC-FC curves in Figure 3.3 (a) and (b), the difference between the each-two lines is positive and seems to indicate some kind of magnetic phase transition. There is, however, no abrupt decrease of moment above around 5K, which either indicates a very high Tc or is the feature of a spin glass state. The M-T curves in Figure 3.3 (b) show a crossing point at around 400 K. This crossing has been seen in Ⅱ-Ⅵ DMS before but at much lower temperature, usually around 10 to 20 K [3]. In those cases, this kind of crossing indicates a spin glass state. In the above M-T curves we can see a change of moment at around 60K, which is due to the residual oxygen. 3.3 Summary We have used a new technology to deposit ZnO based Co doped films in an attempt to get a room temperature ferromagnetic DMS. ALCVD technology enables us to precisely control the doping level and the fabrication of δ-doping structure. We can alter the doping level 37 in an atomic-layer precision, which should cause some interesting effects. We got successful deposition of ZnO films without and with Cobalt doping, but the doping concentration remains fairly small ([...]... plane of CrO2 thin film 84 (b) M-H curve: perpendicular to plane of CrO2 thin film 84 (c) MR curve of CrO2 thin film 85 M-H curve of double layer of CrO2 films 87 MR curve of double layer of CrO2 films 87 FIG 5.6 (a) MR curves of effect of Ion Milling to the CrO2 films: before Ion Milling 89 (b) MR curves of effect of Ion Milling to the CrO2 films: Ion Milling for 3 mins 89 (c) MR curves of effect of. .. lattice structure of (Zn, Co)O should be the same as that of ZnO Figure 2.1 is the lattice structure of ZnO ZnO is a wide band gap - compound semiconductor, which has a direct band gap at the Γ point in the energy band The s-orbital of the Zn2+ and the p-orbital of the O2- form the conduction band and the valence band Normally, the interstitial Zn atoms or O vacancies introduce a donor level of 0.05eV under... Resistivity, carrier density and Hall resistance of ZnO-CoFe (2 nm)-ZnO: Al 71 h (c) Resistivity, carrier density and Hall resistance of ZnO-CoFe (1 nm)-ZnO: Al 72 (d) Resistivity, carrier density and Hall resistance of ZnO-CoFe (0.5 nm)-ZnO: Al 73 FIG 5.1 Schematic illustration of CVD process to deposit CrO2 film 77 FIG 5.2 XRD patterns of CrO2 epitaxial films 79 FIG 5.3 AFM data of CrO2 thin films 81 M-H... [20] Some of them claimed that they have obtained ferromagnetic semiconductor, but some of them reported that they only got some materials of other magnetic phase [21] 1.2.2 Fabrication and Characterization of ZnO based DMS In growth of ZnO based DMS, Co and Mn were used as magnetic dopants [21] [22] [23] [24] [25] [16] [26] [27] Also other magnetic materials were used as dopants [18] Transport and magnetic... patterns of modulated doped ZnO thin films with different structures 61 FIG 4.9 Resistivity of modulated doped ZnO thin films FIG 4.10 M-H curves of ZnO-CoFe-ZnO: Al films 64 67 FIG 4.11 (a) XRD patterns of ZnO-CoFe-ZnO: Al films: without annealing 68 (b) XRD patterns of ZnO-CoFe-ZnO: Al films: with annealing at 300 ℃, 1.5 h 69 FIG 4.12 (a) Resistivity, carrier density and Hall resistance of ZnO-CoFe (4... challenge because of the inconsistent results of different groups and the lack of a mature theory to explain the source of ferromagnetism properties, many researchers reported successful growth of both Co and Mn doped ZnO DMS In addition, besides sputtering, we tried a new technology, ALCVD, which can grow high quality films and precisely control the doping level Moreover, the new δ-doping and modulated... dimension is one of the basic and main requirements Using both spin and charge of carriers can effectively shrink the size of devices, lower device power consumption and enhance running speed In addition, we can integrate computation and storage components together, which can be used in high-density data storage application 1.1.2 DMS and half metal DMSs and half metals will play important roles in spintronics. .. research on ZnO material growth using ALCVD and sputtering should be able to get some good results At the same time, the strategies of δ-doping and modulated doping should enable us to observe some new physical phenomena On the other hand, the special magnetic properties of CrO2 make it a good candidate to make new magnetic devices The reported successful low pressure CVD growth of CrO2 thin films makes... Wang and W Wang, Applied Physics Letters, 83, pp.518 2003 W K Park and R J Ortega-Hertogs, Journal of Applied Physics, 91, pp.8093 2002 W K Park and A T Hanbicki, Science, 295, pp.652 2002 J M D Coey and M Venkatesan, Journal of Applied Physics, 91, pp.8345 2002 K Sato and H Katayama-Yoshida, Phys Stat Sol (b), 229, pp.673 2002 K Ueda and H Tobata, Applied Physics Letters, 79, pp.988 2001 Z Jin and. .. Kim and H Kim, Journal of Applied Physics, 92, pp.6066 2002 D P Norton and S J Pearton, Applied Physics Letters, 82, pp.239 2003 S.-J Han and J W Song, Applied Physics Letters, 81, pp.4212 2002 H.-J Lee and S.-Y Jeong, Applied Physics Letters, 81, pp.4020 2002 J H Park and M G Kim, Applied Physics Letters, 84, pp.1338 2004 K Rode and A Anane, Journal of Applied Physics, 93, pp.7676 2003 S G Yang and ... Chapter CVD growth of CrO2 and fabrication of Devices 5.1 CVD growth of CrO2 5.2 Characterization of CrO2 thin films 5.2.1 XRD and surface roughness 5.2.2 Magnetic properties 5.2.3 Effect of Ion Milling... Application of DMS and (Zn, TM)O 1.4 CrO2 1.4.1 CrO2 1.4.2 Growth and characterization of CrO2 1.5 Application of CrO2 1.6 Motivation of researches Reference 1 1 4 7 8 11 Chapter Structures and Properties... GROWTH AND CHARACTERISATION OF SPINTRONICS MATERIALS LIU WEI (B Eng., Tianjin Univ., P R China) (M Eng., Tianjin Univ., P R China) A THESIS SUBMITTED FOR THE DEGREE OF MASTER OF ENGINEERING