This essay has been submitted by a student. This is not an example of the work written by our professional essay writers.
Spintronics, or the functional control and manipulating the electron spin degree of freedom into electronic devices, is a new and promising research field that attract the scientific and engineering interest . The possibility of performing electronics with the spin of the electron has been fundamental for potential prospects devices with high processing speed and integration, low power consumption, non-volatility, multifunctionality and their suitability for quantum computing . In the last decades, the explosive growth of digital data storage based on spintronics has been occurred. Currently, hard-disk drives read-heads are the maximum exponent . In fact, these devices are scientifically based simply on the tunnel magnetoresistance effect TMR which is defined as the change in electrical resistance of a device in presence of an external magnetic field . A tunnel magnetoresistive vertical spin valve is the principle spintronics device and composes of two ferromagnetic layers separated by a thin insulating layer. The two ferromagnetic electrodes exhibit two different switching fields, and the resistance can be switched between a parallel and antiparallel magnetization configuration upon the application of a magnetic field [1,3].
Recently, the use of organic materials has garnered much attention for advanced spintronics applications. Primarily owing to their ease, low-weight, small cost of processing and mechanical flexibility. Besides, the extremely long spin lifetimes found in organic materials propose considerable advantages comparing to other materials. This promising property is related to the small spin orbit interaction and nuclear hyperfine interaction . Utilizing thin layers of organic material have been performed successfully as spin tunnel junctions and significant magnetoresistance (MR) values have been obtained at low temperatures. Essentially, the room temperature is the basic operational requirement for most applications of spin transport in organic. Up to now, the studies on thin organic spin tunnel junction have exposed significant MR effect at room temperature. While the devices with organic layer thicker than 15nm indicates a decrease in MR below room temperature range . Then the study of organic spintronics extends to include C60 due to its considerable properties among other organic materials. C60 exhibit lower hyperfine coupling due to the absence of hydrogen in the molecular structure in addition to small natural abundance of the 13C nuclear spin [ 1,3,4]. The lowest unoccupied molecular orbital LUMO of C60 is well matched with the Fermi energy of common ferromagnetic materials. This matching allows easy current injection and maintains a moderate energy injection barrier at the same time .
Overview of current project
As suggested by the title, the main focus of my project is to understand the fundamental principles behind spin polarized tunneling devices. The tunneling in my study will be through molecules positioned between ferromagnetic electrodes. In details, C60 is inserted into ferromagnet/insulator/ferromagnet structures to form vertical spin transport devices. This allows an experimental study the tunneling and possibly inter-molecular transport rule of spin-polarized carriers in C60 in a vertical transport geometry based on magnetic tunnel junctions (MTJs). C60 is an attractive choice due its several properties that make it ideal for spintronics devices, presented in 3.1.3. This study accomplish with a Meservey-Tedrow technique, which ideally provides direct determination of spin-polarized current tunneling through an OS using a superconductor . In essence, it depends on the fact that a very thin superconducting aluminum films in an applied magnetic field, showing Zeeman splitting of the quasi-particle density of states into spin-up and spin-down parts. The resulting spin densities of states in the superconductor are then similar, but shifted from their original energy by ï‚±ïH, where u is corresponding to the magnetic moment of electron . Certainly, Meservey-Tedrow technique is straightforward to ascertain the quality of the C60 barrier junction and measuring the spin polarization of the ferromagnetic electrode. Indeed, the study involves samples fabrication and characterization using varies devices such as sputter deposition system, MR rig, helium flow cryostat and vibrating sample magnetometery (VSM).
This report shows scientific progresses have been made in tunneling magnetoresistance field using organic materials. Followed by deep explanation of experimental techniques. A brief review of the early stage obtained results is given subsequently. Finally, a summary of the study is provided and some suggestions for future work.
This section includes description of the studies conducted in the field of organic spintronics followed by studies accompanied on the use of C60 in particular. The corresponding analyses of experimental results are discussed clearly. Moreover, the main aspects and different factors that influence the magnetoresistance are presented.
Many scientific progresses have been made in tunnel magnetoresistance field. The system of FM/I/FM has been studied experimentally and theoretically. In 1975 Julliere  stated a model to estimate the magnitude this phenomenon. His model is a combination of Mott's current model in ferrmomagnetic metals besides the model proposed by Mesrevey-Tedrow of the effective density of states estimation using spin-polarized tunneling. Julliere model gave a good insight however it was lacking significant effects including temperature, voltage dependence and properties of the tunnel barrier such as the material, height and the width [ 8, 9]. This study altered by Slonczwski  to account the permeability of barriers that results in an overlap of the wave functions inside barriers. Nevertheless, the temperature and voltage dependence of the TMR ratio has not justified in this Slonczwski model. Two-step tunneling has been put forward to explain the temperature and voltage dependence in which the pervious model was unable to interpret . This followed by extensive studies whereby series of experiments were performed in this field.
At beginning of this century, the field of spintronics has been expanding by introducing organic semiconductor (OSC) materials leading to a new research area called organic spintronics. Different types of OSCs such as polymer and small molecule have been applied as spin transport layers in organic spintronics devices. OSCs have long spin diffusion time due to week hyperfine interaction and low spin orbit coupling. Also, it presents as a good candidate for spin injection unlike inorganic semiconductor spintronics where this feature considers as one of the major challenges. The later can be explained in term of mismatch conductivity where spin injection from ferromagnetic electrodes directly to inorganic semiconductor is inefficient in contrast to OSCs .
The first demonstration of an organic spintronic device was in 2002 by Dediu et al. . The authors designed a lateral device having two La0.7Sr0.3MnO3 (LSMO) electrodes patterned by electron beam lithography. These electrodes were separated by 70-500 nm and bridged by a narrow channel of sexithienyl (T6) (Fig 1a). A strong magnetoresistance (MR) response up to 30% was observed at room temperature for 100-200 nm T6 channel lengths (Fig. 1b). Beside that, an estimation of a spin diffusion length of 200 nm in the organic layer room temperature was reported. Nevertheless, this work did not provide a straightforward demonstration that the observed MR was related to the magnetization of the electrodes and consequently to its spin polarization. This is because the antiparallel configuration of two LSMO electrodes magnetization could not be set. Instead, a resistance change was measured between zero filed and a perpendicular magnetic field of 3.4 kOe, corresponding to a random and parallel magnetization alignment respectively .
Figure 1: (a) Schematic view of hybrid junction LSMO/T6/LSMO and dc four-probe electrical scheme and (b) Magnetoresistance (H = 0.3 T, where H is the magnetic field) of the hybrid junction device depicted in (a) as a function of the channel length (, reproduced in ).
A couple of years later, Xiong et al. reported the first vertical organic spin valve device consisting of LSMO and cobalt electrodes, with a thick (130-250 nm) layer of Alq3 in between (Fig. 2a). A sizeable inverse MR of 40% was obtained in an LSMO/Alq3/Co spin valve device with a 130 nm thick layer at 11 K. The MR exhibited a rapid decrease when the Alq3 film thickness was increased or the sample temperature was increased (Fig.2b). Subsequently, the MR vanishes when the temperature above 200K. Following this encouraging report, the same group performed a detailed study using various organic materials as a spacer layer between LSMO and varies metallic electrodes in 2005 . Their devices exhibit large negative high-field MR responses. They attributed this high-field MR to the anomalous field-dependent Fermi level shift in LSMO that leads to enhance carrier injection at the LSMO-organic interface. Since then, different groups carried out many studies on the devices with the same architecture using both Alq3 and other OSC martial as spacer layer. However, the observed MR is varied in size and sign even with the same LSMO/Alq3/Co structure. All these differences give clear evidence on the effect of ferromagnetic electrode/OSC interface properties. The latter can be induced by interface interactions, intrinsic properties of the material in addition to different fabrication conditions .
Figure 2: (a) Schematic representation of a typical LSMO/Alq3/Co device (b) GMR loop of a LSMO (100 nm)/Alq3 (130 nm)/Co (3.5 nm) spin-valve device measured at 11 K. The insets show the anti-parallel (AP) and parallel (P) configurations of the FM magnetization orientations at low and high H, respectively 
An key step forward for organic spin tunnelling was the fabrication of devices by direct in situ UHV organic vapour deposition with shadow masking in 2007 by Santos et al. . Moreover, a thin Al2O3 layer of 0.6nm was introduced at interface between the Co electrode and the Alq3 barrier by plasma oxidation to form the vertical tunneling device of Co/Al2O3/Alq3/NiFe. In fact, using a hybrid inorganic-organic ultrathin barrier produced positive (non-inverse) TMR when the electrode magnetizations were switched from parallel to antiparallel. This kind of device showed a few percent TMR at room temperature. Essentially, the study demonstrated the important role of inserting an ultrathin Al2O3 tunnel barrier to prevent penetration and chemical reaction  of the Co atoms into the Alq3. In details, Al2O3 interfacial layer suppresses the formation of trap states that are concentrated mainly at the metal-organic interface. As a result, higher spin polarization can be obtained hence improves spin injection efficiency a cross the FM/OSC interface. Recently, giant positive TMR up to 300% was detected at 2 K in Alq3-based nanojunction utilizing a LSMO and Co electrodes. However, the observed value undergoes a sharp decrease in temperature and vanished below the noise level at 180 K. Barraud et al. proposed a new mechanism to explain different sign and amplitude of MR in devices with thin and thick Alq3 tunnel barrier . The mechanism attributed to the formation of spin- hybridization-induced polarized states in the first monolayer at the electrode interface which can lead to a complete change in organic valve magnetoresistance. This in turn means an increase of the effective spin polarization and a change in sign of the electrodes as pointed above. Generally, although many studies took place in this new uprising field, the mechanism underlying the spin injection into the OSCs are still to be unraveled and remain one of the key challenges.
A remarkable study on C60-based vertical spin valves has been demonstrated by Gobbi et al. . The devices were fabricated in-situ in a UHV dual chamber evaporator. Through deposition process using shadow masks a structure of Co/AlOx/C60/Py was obtained. Where AlOx layer was achieved by depositing a thin Al layer (0.9 nm) followed by plasma oxidation. Additionally, a C60 layer was thermally evaporated through shadow mask with thickness between 5 and 28 nm. The devices were designed in such way that the fullerene only covered some of the bottom lines whereas the remaining electrodes were left only with the thin AlOx layer and are used as reference junctions. This allows observing the tunnelling effects due to the C60 layer purely. A sketch of the devices is shown in figure 3a. It is important to notice that the layer of AlOx was a "leaky" film instead of a full functional tunnel barrier. The idea of this choice is to ensure that the measured electronic properties in the hybrid ferromagnetic-organic junctions are corresponding to the C60 molecules. Therefore eliminating the effect of inorganic from that of the organic barrier. Nevertheless, the AlOx layer is significantly required to prevent the C60 reacting with the Co bottom electrode, since Co is very reactive with varies organic molecules as proved by many studies . A significant room temperature magnetoresistance in excess of 5% was achieved for different thickness of the C60 interlayer (from 5 nm to 28 nm). Explanation of both electronic and spin coherence transport was based in term of a multistep tunnelling model. In this model the electrons undergo a number of tunneling steps from molecule to molecule through the interface barrier until they reach the second electrode .
Figure 3: (a) Schematic of a C60 based tunnel junction with 3 of the Co bottom electrodes covered with C60 leaving 2 electrodes with only AlOx as reference junctions. (b) A cross-section representation for the Co/AlOx/C60/Py device .
Lin et al. have been accomplished a study using C60-based vertical spin valves. In the devices, a C60 layer was thermal evaporation as the organic semiconductor spacer layer on LSMO, covered by a 15 nm thick Co film on top . All fabrication steps were performed inside the vacuum evaporation chamber. Basically, the purpose of this study was to potentially overcome the limitations imposed by hyperfine coupling. Since the later paly an important role limit the spin-transport length in organic semiconductors. C60 has been proposed because it has orders of magnitude lower hyperfine coupling than most other organic materials . The lowest hyperfine coupling of C60 turned out because it does not contain hydrogen atoms, which considered as the origin of such kind of coupling (or hyperfine coupling originate from these atoms). In contrary most other organic materials exhibit strong hyperfine coupling due to the presence of hydrogen in the molecular structure [21, 22]. Despite the above-mentioned fact, Lin et al. study did not reveal a significant improvement of the spin-diffusion length. As a consequence, three possible interpretations have been suggested to reflect this observation. Firstly, a mechanism other than hyperfine coupling is responsible for the loss of spin-polarization. Secondly, the observed magnetoresistance is corresponding to tunneling magnetoresistance which vanishes beyond the tunneling range for C60 thicknesses. Finally, for thick devices the mismatch in conductivity limits spin-injection.
Fabrication and characterization
This section will describe the experimental techniques used throughout this study. The experimental techniques were carried out in two main steps. The first step is the sample fabrication using sputter deposition technique, plasma oxidation and thermal evaporation. The second step is deal with the characterization and electrical measurements used during this work. This includes structural and magnetic characterization beside magnetoresistance measurements at room temperature and low temperature conditions. Also, a general account of electrical contacts configurations has been briefly considered before establishing the measurements.
Additionally, this chapter outlines some of the basic principles behind these techniques. This is to provide an understanding of the results and limitations in theoretical and practical aspects of the analysis. A brief description of the materials used is given also at the beginning.
Ferromagnetic materials: Cobalt (Co) and Cobalt-Iron-Boron (CoFeB)
A sizable tunnel magnetoresistance (TMR) values were attained at room temperature since 1995, the point has attracted extensive attention on spin dependent materials. In principle these materials provides two key advantages comparing to other magnetoresistive materials, a high field sensitivity and orders-of-magnitude higher resistance which can be easily achievable even with small sizes. Since that a special attractive has been taken toward these materials for low field/low power device applications due to the sophisticated properties previously mentioned. In applications point of view and specifically in term of devices, such as MRAM, read heads and magnetic sensors, a high TMR value is required because it leads to larger signal level, lower power consumption and greater speed in addition to flexible design margin.  Thus, a considerable interest has been half metallic ferromagnetic electrodes, such as the manganite perovskites and CrO2, due to higher the tunnelling spin polarization (TSP) hence higher the TMR they have. However, it is important to notice that the highest TMR values in tunnel junctions with electrodes formed from the perovskite manganites have been obtained only at low temperatures. Room temperature measurements contributed very small effects. This has been interpreted in term of the low Curie temperature of these ferromagnets. On other hand, ferromagnetic metals formed from Fe, Co and Ni have much higher Curie temperatures which are above room temperature  For this reason, most studies are based on these kind of materials.
In the present work, ferromagnetic electrode of CoFeB is chosen in tunnel junctions due to a high tunnel magnetoresistance (TMR), at room temperature, associated with high spin polarization. In addition, it has low coercivity (Hc) which is desirable for sensor applications [26, 27, 28]. Most interestingly, amorphous CoFeB electrodes gain important advantage of being homogenous in the very small scales, which cannot achieve with polycrystalline electrodes. The one is needed for high density recording media . Moreover, TMR values of over 55% have been obtained using amorphous CoFeB as a magnetically soft electrode in MTJs. It have been pointed out through varies studies that CoFeB / AlOx interfaces are characterized by a high microstructural quality to which the large TM attribution can be considered .
Basically, D. wang et al. investigated spin dependent tunneling (SDT) junctions with a stack structure of Si/Si3N4/Ru/CoFeB/Al2O3/CoFeB/CrMnPt. The SDT wafers were deposited utilizing dc magnetron sputtering and photolithography techniques were used for the purpose pattering and connection with one layer of metal lines. It has been revealed by high-resolution transmission microscopy that the CoFeB has an amorphous structure in addition to a smooth interface with the Al2O3 tunnel barrier. High TMR value of 70.2% has been obtained with a spin polarization of 51% for CoFeB at room temperature condition. Further study has been demonstrated one year later by, D. Djayaprawira et al. where magnetoresistance ratio up 230% at room temperatures was observed using MgO tunnel barrier sandwiched with amorphous CoFeB ferromagnetic electrodes prepared by magnetron sputtering . In most recent MgO- based MTJs, a very large tunnel magnetoresistance (TMR) ratio of 604% at 300K was observed . These studies provide interesting results about amorphous CoFeB electrodes and allow further investigation in present work.
Tunnel barrier - Aluminum Oxide (Al2O3)
Insulators such as AlN, Al2O3, and MgO have been proposed as a barrier for many magnetic tunnel junction devices. Of these, Al2O3 has several excellent properties in addition to insulating properties, make it a desirable material for such kind of devices. Aluminium oxide is transparency, high abrasive and corrosion resistance, as well as good optical properties. It has outstanding properties in terms of chemical inertness, mechanical strength and hardness . The most common form of aluminium oxide is the amorphous form. Crystalline aluminium oxide exists in different arrangements and is formed when aluminium is oxidized at very high temperatures exceed 4008 C .
A high quality barrier is required to improve junction properties according to its stability and the bias dependence. The barrier thickness is a critical parameter should be considered since the electrical resistivity of the barrier depends exponentially on the thickness of the barrier. In particular, a thin barrier is preferred to reduce the resistance of the junction . The present technique to form Al2O3 barrier is by oxidized a thin Al layer deposited on the ferromagnetic material. The oxidizing process involves sputtering aluminum in an argon atmosphere which contains oxygen where a homogeneous and stoichiometric Al2O3 layer can be obtained. Koski et al. gave a comprehensive study about this technique . Extensive studies have been done to optimize Al film thickness required for the barrier formation. Depending on the type of FM electrodes, a uniform coverage of Al film thickness varied between 7 A to 18 A. In general, the Al layer should be thick enough to avoid oxidizing the FM surface during the barrier formation. On the other hand, it should not be very thick so that an excess of Al metal will be left behind unoxidized. The presence of nonmagnetic metal at the interface leads to a reduction in polarization and the JMR consquentely .
Buckminsterfullerene - C60
The C60 molecule is a fullerene cage consists of 12 pentagons and 20 hexagons, arranged as in the polyhedron known as the truncated icosahedron, with a carbon atom at each vertex. C60 was discovered accidentally by H.W. Kroto, R. E. Smalley and co-workers in 1985. In fact, they evaporated graphite by laser irradiation and a remarkable stable cluster consisting of C60 has been produced. Their discovery leads to a Nobel Prize in 1996. The C60 molecule known as Buckminster-fullerenes, it was named for R. Buckminster Fuller, an American architect noted who created the geodesic dome [35, 36]. C60 a symmetric structure, sublime at moderate temperatures of 350oC. The electrical resistivity of undoped solid film of C60/70 has been found to be ï² ï‚»1014 (ï- cm)-1 at room temperature . For a highly crystalline C60 thin film, the room temperature resistivity reaches ï² ï‚»1010 (ï- cm)-1 similar to C60 single crystal measurement [38, 39]. A series of different techniques mainly the spectroscopic were utilized to establish the bandgap of C60 films. A range of 1.5eV to 2.3 eV has been indicated for the fundamental gap under these investigations, which essentially gives qualitative agreement with theoretical predictions . The electronic structure of polycrystalline C60 thin films has been shown to exhibit n-type semiconducting behavior based conductivity measurements and surface photovoltage spectroscopy (SPS) .
Due to several properties that C60 attain, it is considered as ideal choice for organic spintronic devices. Firstly, C60 molecules can be grown by sublimation under ultra-high vacuum (UHV) and are very robust. Accordingly they can be cleanly between ferromagnetic metallic thin films and sustain without being damaged [Gobbi]. Secondly, C60 exhibit lower hyperfine coupling due to the absence of hydrogen in the molecular structure in addition to small natural abudance of the 13C nuclear spin [1, 3, 4]. Finally, the lowest unoccupied molecular orbital LUMO of C60 is well matched with the Fermi energy of common ferromagnetic materials. This matching allows easy current injection and maintains a moderate energy injection barrier at the same time .
Direct Fabrication "Sputter deposition"
Sputter deposition is a physical vapour deposition process that is extensively used to deposit high quality film of atoms on semiconductor wafer, on head surface and magnetic media. Specifically, sputtering is carried out by means of plasmas, which in turn generate charged particles that can be accelerated towards a surface electrically  In principle, sputter deposition process is basically involved removing the surface atoms by energetic particles, such as accelerated ions, and subsequently accumulating of these atoms onto a substrate to form a thin film [43, 44]. There are several sputtering methods including DC diode, RF diode, and magnetron sputtering.
DC sputtering system is considered to be the simplest model among other sputtering systems. This system consists of a pair of electrodes: cathode and anode. The cathode is biased at negative 2~5 kV and covered with target materials in the front surface to be deposited later. On the anode, the substrates are placed and usually grounded. The sputtering chamber is filled with an inert gas (typically Argon gas), which is introduced at low pressure. In sputter deposition, positive argon ions are created in plasma under the application of dc voltage between the two electrodes. Then argon ions accelerated towards the surface of the target since it is negatively biased. Due to the impact, atoms of the target material are ejected as a consequence of the emerging collision cascades. Subsequently, the sputtered atoms condense on the surface of the substrate and form the growing thin film [44, 45, 46].
In magnetron sputtering, a permanent magnet is located behind the target surface in order to increase the rate of ionization. Principally, the resulting magnetic field forms a closed loop path that acts as an electron trap so that the ejected electrons from the target move in cycloid curves immediately above the target. Magnetic field traps the electrons nearby the target and triggers them to spiral around in the parallel plane right above the target. This electron trapping effect strongly increases the probability of ionization of the sputtering gas (i.e. enhances the collision rate between the electrons and sputtering gas molecules and create ions) that is finally increases deposition rate effectively. Furthermore, the path length travelled by electrons is increased and the magnetic field in this way acting as the gas pressure had been increased. The above process enables to lower the gas pressures used in magnetron sputtering down to 0.5 mTorr [42,43, 44, 47]. A schematic diagram illustrates the basic principle of sputter deposition of a typical dc magnetron sputtering system is shown in the figure below.
Figure 4: The basic principle of sputter deposition where ions are produced in plasma and then accelerated by applying dc voltage between the electrodes towards a target. Whereby, the target atoms will be removed due to collision between the electrons and sputtering gas molecules. Finally, the atoms of the target will be condensed at the substrate .
In this research, A sputter deposition system namely IVOR is used for samples fabrication. It consists of a total of eight sputtering guns(two magnetic and six nonmagnetic). The operational process of the sputter system is run under high vacuum conditions using roughing and cryo pumps. The first one is used to pump the chamber down from room temperature to about 25 mTorr. Then the cryo-pump can pump down the main chamber to a base pressure of about 10-6 mTorr. Indeed, the base pressure in the main chamber is maintained at about 10-8 mTorr through liquid nitrogen, which flows in tubing system located inside upper part of the chamber. This step lowers the pressure by encouraging absorption of molecules into the walls.
The substrates are cut to appropriate sizes and cleaned with acetone and isopropanol subsequently. Then, they are mounted on the sample wheel with a shutter system to control the growing of multilayers structure. After loading the sample in the main chamber and reducing the chamber pressure following the above-mentioned procedure, the fabrication process is carried out. Taking into account, all targets are always subjected to a pre-sputtering cleaning process before actual deposition by energetic argon ion bombardment to remove contaminated surface layers from the targets. The fabrication of tunnel junctions' devices is processed with three-shadow masks process. The masks are designed to produce a cross structure with a circular shape at crossed point (see figure 5). Initially, the bottom electrode is deposited through the first mask to obtain a specific alignment of tunnel junction. After that, a thin insulating barrier is achieved using plasma oxidation. Specifically, a thin layer of aluminum, in range of 1nm, is deposited and followed by plasma oxidation. During oxidation, both argon and oxygen are introduced to the chamber with flow rate of 16 sccm and 76 sccm respectively while the magnetron gun is kept ignited. After that, C60 layer is formed with thermal evaporation, which is considered the simplest of all the thin-film growth techniques. In detail, with a base pressure of 10âˆ’8 Torr, C60 is heated to its boiling point inside of a vacuum chamber from a powder source at significantly low power . Vapor from the source travels to the substrate, where some of it condenses in the form of a thin-film. In current experiment, current of 21.5 A is applied to heat a filament and turn it evaporates C60 molecules. The later condense on the substrate surface under the high vacuum circumstances. Finally, the top electrode is deposited following the same procedures as the top electrode with different target materials.
For protection purpose or on other word to reduce oxidation in ambient air, the complete stack was topped of, over whole structure, by a cover layer of aluminum of 1.5nm. This step showed a significant improvement to sample resistance. The whole multilayer layout of the samples is Si-substrate, SiO2, Ta (7.5), Cu(5), Co(3), Al(1) plus oxidation for 40sec, C60(5) , Co60Fe20B20(5), Cu(5), Al(5). The numbers in parentheses give the layer thickness in nm. For Mesevey-Tedrow samples the structure consists of SiO2, AlSi (6), oxidation for 20sec, Py(6), Ta(5). The entire process in the sputtering system is programmed so that multilayer deposition, oxidation steps and evaporation process are Substrate
Insulating layer/ C60
Figure 5: (a) A diagram illustrates the structure of tunnel junction samples in cross hair configuration. (b) The same structure in stack form, colours here to as indication of structure in a and b.
Characterization and Measurements
X-rays are relatively short wavelength, high-energy electromagnetic radiation. They characterized by an electric field vibrating perpendicular to the direction of movement at constant frequency. This variation of the electric field provides electrons a sinusoidal oscillation with time at the same frequency. Hence, X-rays are generated as a consequence of periodic acceleration and deceleration of the electron [50, 51]. The conventional method of producing X-rays in a laboratory is to utilize a vacuum tube. This tube contains a tungsten cathode filament, which is heated by an AC voltage, and then electrons are produced through thermionic emission. Electrons are accelerated in vacuum under high potentials ranging from 5 to 80 kilovolts toward a metal target, the anode, creating the X-rays .
When an X-ray beam is subjected onto a crystalline material whose atomic arrangement shows the long range periodicity, a physical phenomenon called diffraction is occurred [Yoshio]. It has been found remarkable characteristic patterns of reflected x-radiation on crystalline materials where intense peaks have been observed for certain sharply defined wavelengths of incident directions. This investigation has been found by W. L. Bragg and the peak has been named the Bragg peak in honor of the discoverer. In principle, a crystal made out of parallel planes of ions. The planes spaced by a distance d. The x-rays specularly reflected by ions arranged in one plane. A constructive interference results from the reflected rays from successive planes. The path difference between the two rays is 2dsinï± which obeys Bragg condition where the pervious term is representing an integral number of wavelengths. Take into account ï± is the angle of incidence [52, 53].
Based on the principle mentioned-above, the measurements are carried out by varying the angle of incidence ï± and the corresponding intensities of the resultant diffracted peaks produce a well defined pattern. The pattern is used to probe out the thickness of the samples where the Bragg's Law condition is satisfied. In fact, X-ray diffraction (XRD) is a versatile, non-destructive technique that reveals specific details of the sample studied. For instance, structural, physical and chemical information about the material investigated can be obtained. However, samples thickness and determination of growth rate is the main purpose of using XRD.
Magnetic measurement in the study was done with a vibrating sample magnetometer (VSM). The VSM is a simple but effective technique for characterizing properties of magnetic materials established by S. Foner (1956). It has straightforward design allows investigation of a common experimental technique for measuring magnetic material properties such as hysteresis, saturation and coercivity. The VSM relies on the detection of the emf induced in a coil of wire based on Faraday's law of induction [54, 55]. The principle of operation is comprised two components, the sample vibrator mechanism and the induction signal detection coils. The latter is pair of identical pick-up coils wound in opposite directions being situated in close proximity to the sample which is also located in an external magnetic field. It is required that the sample oscillates in a periodic and stable manner typically through the use of electromechanical or piezoelectric transducers. Principally, when the applied magnetic field, coil position, or sample position is changed; the voltage is induced in a detection coil by a flux change. The induced voltage in the pickup coil is proportional to the sample's magnetic moment and gives a direct measure of the magnetization. Many different experimental arrangements can be employed to suit particular investigations of the magnetic induction measurements involving amplification and lock-in detection of the signal. Whereby, the entire procedure is running on the PC and control through the VSM software application [55, 56, 57].
Unlike other techniques, The VSM has extremely high sensitivity whereas changes of 5 X 10-5 to 5 X 10-6 emu can be detected. It is also simple, in- expensive, and versatile and at the same time provides precise magnetic moment measurements can be carried out as a function of temperature, magnetizing field, and crystallographic orientation .
In this study, the samples under investigation are vibrated perpendicularly to the applied field (up to 6T) and the oscillating magnetic field of the vibrating sample induces a voltage in the stationary detection coils. From voltage measurements the magnetic properties of the sample are deduced in temperature of about 1.7 K which is crucial to the present study.
Electrical Transport Measurements
The tunnel junction samples were mainly characterized with magneto-current measurements at RT. Magneto-transport measurements were performed using the four-point probe technique which was originally proposed to measure earth resistivity by Wenner in 1916. Clearly, it is provides considerable practical method to measure material resistivity therefore the magnetoresistance of the system in a simple and accurate way. Whereby, the influence of the lead resistances was removed from the measurement. In principle, the current is applied thorough one electrode to the other whereas the voltage drop is measured. From this, the resistance can be found using Ohm's law [58, 59, 60, 61, 62]. Essentially, this method enables to measure the resistance change as a result of varying magnetic field in term of magnetoresistance measurements. Important point regarding this method is that the measurements can be carried out for each electrode separately. Thus, the electric transport can be measured especially for superconducting electrode in case of Meservey-Tedrow samples as will be explained later in this report.
Throughout this project, the room temperature measurement of magnetic junction samples is carried out in MR Rig in which sample is placed in the center of two magnetic poles. A field ranging from -600 Oe to 600 Oe was varied in order to switch the two electrodes of the MTJs between the parallel and antiparallel configurations. Correspondingly, the current-voltage curves can be measured at fixed values of magnetic field.
Low temperature measurements, especially for Meservey-Tedrow, are performed with variable temperature helium cryostat, which equipped with 8T magnet. The magnet consists of a number of coaxial solenoid sections wound using multifilamentary superconducting wire. A variable temperature insert is used to allow a continuous adjustment of sample temperature over a wide range. The sample temperature can typically be controlled over the range from 1.4 K to 300 K. This is done by balancing the liquid helium with the heater and the rotary pump. Temperatures below 4.2 K, the normal boiling point of helium, are obtained by reducing the vapour pressure of liquid helium in the sample space. The later can be filled with liquid helium continuously by controlling the needle valve. A constant liquid level can be maintained if the flow rate set correctly so this can just replace the evaporating liquid. The sample temperature can be set to approximately 1.4K, if a large enough rotary pump is used to reduce the vapour pressure of the liquid. The rotary pump is also used to extract air so it does not enter and freeze inside [63, 64]. Similar to room temperature measurement, four- point probe technique. The current is varied and the voltage drop cross the sample is recorded. In the same approach, conductivity-voltage measurements were built by using the lock-in amplifier output. For measuring the tunneling magneto resistance (TMR) the magnetic field was swept, between negative and positive fields, while the sample current was kept at a fixed. The entire system is controlled by computer allowing different kind of measurements such as current, temperature and field sweeps.
Results and discussion
To present, two different sets of samples have been prepared. The first one is magnetic tunnel junction samples with a structure consists of, from the substrate side, Ta (7.5)/Cu (5)/Co (3)/AlOx-C60(5) /Co60 Fe20 B20 (5)/Cu (5)/Al (5) (numbers are nominal thicknesses in nanometers). The second set is the Meservey-Tedrow samples with a superconducting electrode of AlSi.
Given the importance of determining the thickness of the layers in the samples, XRD measurements have been conducted for this purpose. The deposition rate for each material then can be set for sputtering process. Basically, calibration samples, thin films samples, are exposed to X-rays at an angle ï± with respect to the layer surface, and at an angle 2ï± the reflected x-rays are detected. As explained in section 3.3.1, constructive interference occurs when the Bragg condition is satisfied. These peaks provide information on the crystal structure, lattice constants, layer roughness's and layer thicknesses when proper analysis is done. .
Figure 6 shows the x-ray reflection intensity as a function of ï± for C60 film on Si/SiO2 sputtered for 200s using a current of 22A. By analyzing the periodicity of the interference fringes (Bragg peaks), a thickness of 238.8ï‚± 5.1nm is found, corresponding to the deposition rate of 12.0nm/s. The current used to grow C60 in present work is 21.5 A where the deposition rate of 9.0 nm/s has satisfied following the same procedure.
Figure 6: (a)The XRD pattern for a C60 sample (b) The data fit of Bragg peaks in (a) and the corresponding thickness of the sample.
A four-point measurement technique is used to measure the magnetic field dependence of the junction resistance. The current sourced through two contacts of the top and bottom electrodes while the voltage drop a cross the other two contacts was measured as a function of the external magnetic field. Following same technique, IV characteristic curves were measured at different magnetic fields. Accordingly, The electronic transport properties of the MTJs, in which C60 is deposited between the alumina barrier and the ferromagnetic electrode, have been investigated using Al2O3 and C60 barriers 5 and 1 nm thick respectively. The I/V characteristics of all devices are nonlinear and symmetric. Figure 7a shows a typical I/V curve for MTJ sample at 10K. The inset shows the conductance versus voltage V of the same junction. There are two criteria to highlight: first, the I/V is nonlinear and second parabolic shape of the differential conductance as a function of voltage are suggestive of tunneling as proposed by Rowell .
Figure 7: a) Current voltage data for MTJ sample at 10K. The inset the tunnel conductance G is plotted against the applied voltage across the junction at same temperature. b) TMR was measured at 1.4K as a function of magnetic field.
The change in the junction resistance (at 1uA) is plotted as an external magnetic field, which goes from 600 to -600 Oe and then reversed from -600 to 600 Oe with the samples at 1.4. 10, 20 and 50 K is shown in Fig 7b. The field dependence of resistance in FM/I/FM junction can be explained briefly as follow. At high fields, the magnetizations of the two FM electrodes are fully saturated and have alignment in the direction of the applied field. The tunneling probability and hence the current is high in this case. As the applied field decreases toward zero, changes signs and reach a value of Hc of one of the electrode, then the magnetization of the electrode with lower Hc reverses whereas for the electrode with higher Hc the magnetization remains the same. Two electrodes exhibits opposite magnetization, antiparallel to each other, in this field range thus the tunneling probability and the current is low. The configuration of both electrodes becomes parallel upon a further raising the field to reach Hc for the second electrode. The tunneling probability and the current again attain higher value. Principally, The difference of Hc in the FM electrodes causes the change in tunnel junction and tunnel resistance in consequence . Therefore, the tunnel magnetoresistive effect can be calculated using the relation:
Where Rp and Rap are the resistances of junction at parallel and antiparallel magnetization of two ferromagnetic layers, respectively .
Figure 1b shows the magnetoresistance (MR) for junctions with FeCoB top electrodes plotted versus the external magnetic field. One can distinctly see the two stable resistance states as the applied field is varied. The junction magnetoresistance (JMR) seen in this case (defined with respect to the peak resistance) is 0.62% at 1.4K corresponding to 1.67Kï-. These values varied slightly with increasing the temperature. Moreover, the entire work shows a decrease in the resistance as the temperature increases from 1.4 to 50Kwhich represent a typical of tunnelling behavior (see figure 8a).
A common characteristic for CoFeB junctions is that the MR is not maximum at the lowest temperature (1.4 K), but increases with higher temperatures to reach a maximum around 10-20 K (as shown in figure 8b). This could be due to a change in the hopping process in C60 and the relationship between the thermal energy available with the activation energy. Below 10-20 K, this will result in shorter tunneling steps and higher spin flip scattering. At 10-20 K the temperature is sufficient to generate phonon assisted tunneling with longer hopping distances. For temperatures above 10-20 K, the MR is reduced due to the decrease in polarization of the magnet and the increased vibron/phonon density.
Figure 8: Temperature dependence of (a) the resistance and (b) the magnetoresistance.
One of the ideal probes to study spin transport is the spin-polarized tunneling utilizing the Meservey-Tedrow method of detecting spin-polarized tunnel current using a superconductor. Here samples with structure of AlSi(6)/ barrier/Py(6)/Ta(5) have been prepared with varying C60 thicknesses alumina barrier oxidation time. The critical temperature, current and field of the superconducting electrode has been characterized first then the junction measurements have been carried out. The following figures show the measurements of AlSi electrode of one of the sample.
Figure 9: (a) The I-V charatuerstic curve of AlSi electrode measured at 1.4K. (b) The resistance measurements as a function of the applied magnetic fields which revealed a critical field value between 2.2 T and 4.3 T.
However, the barrier measurements have not shown Zeeman splitting of the quasi-particle density of states under application of the magnetic field, as it is clear in figure 10. In essence, this phenomenon has no ideally optimized in our study in which the direct determination of spin-polarized current tunneling through an OS has not achieved. This can be interpreted by two reasons. The first one, may the sample need to cool down below 1.4 K so the effect can be clearly observed. The second one is the electronic since such kind of measurement for low voltages (below 150mV) very noisy for that an electronic circuit need to be satisfied.
Figure 10: (a) Tunnel conductance versus bias at 1.4 K with and without an applied magnetic fields for a junction consists of: 6 nm AlSi/barriers/ 6 nm Py/ 5 nm Ta with 1 nm Al2O3/ 2 nm C60 as barrier. (b) Normalized conductance of an Al-Al2O3-Ni junction measured as a function of voltage for different magnetic fields .
Spintronics is a new considerable research field has a great potential for use in a variety of technologies that involve highly developed magnetic field sensors, such as hard disk drives, read heads and digital data storage. Recently, an important step has been developed combining the potential of spintronics and organics electronics to reveal a promising field called molecular spintronics. It offers flexibility, chemical engineering and low production costs. These are particularly important essential advantages comparing to inorganic devices [19, 69].
In the light of the above advantages and practical applications, present study introduced a considerable experimental work to study tunneling magnetoreistance(TMR) of magnetic tunnel junctions (MTJ) with organic molecules as an important stage in the development of spintronics. Current study deals with two types of samples based on tunneling junction concept. The first set includes magnetic tunnel junctions in which and insulating layer of aluminum oxides is sandwiched between two ferromagnetic electrodes. A layer of organic material of C60 involves as spacer layer is considered to study spin transport through this type of organic material which provides excellent properties comparing to other organic materials. Electrodes material is selected based on specific characteristic as explained previously. The second set of samples comprises replacement of one of ferromagnetic electrode with superconductor electrode namely Meservey-Terdrow samples. It gives direct determination of spin-polarized current tunneling through an organic material. In this study, a well-defined experimental work has been followed including samples fabrication, characterisation and resistance and magnetoresistance measurements at both room temperature and low tempera.
ture up to 1.4 K.
In present work, TMR of 0.6% has been obtained from tunnel junction samples using Co and CoFeB electrodes. Further investigation need to be demonstrated with these samples.
For Meservey-Terdrow samples, characterizations superconducting characterizations such as critical current, field and temperature has been achieved. However, Zeeman splitting of the quasi-particle density of states under application of the magnetic field has not obtained yet.
The obtained results up to this stage can provide a good platform for more detailed analysis for future study of tunnel magnetoresistance effect. Basically, there are some suggested improvements for future work outlined as follow. A fabrication method needs to be improved whereby varies parameters could be improved to get good samples including pressure, power and oxidation process. Cobalt-Gadolinium could be used as alternative of CoFeB feromagntic electrode. Furthermore, Mesevey-Tedrow samples need to be optimized so the next step will be a much low temperature measurement in range of mK using 3He flow cryostat. Optimizing this point allow a suitable ferromagnetic material for magnetic tunnel junctions.
Indeed, more research is required to gain a complete understanding of spin transport through organic materials. A better insight into the transport through C60 needs to perform through good fabrication and characterization procedures. By the end of this project, optimized C60 based magnetic tunnel junctions need to attain.