- Email: [email protected]
Etching behaviors of tunneling magnetoresistive (TMR) materials by ion beam etching system
Available online at www.sciencedirect.com
ScienceDirect Materials Today: Proceedings 5 (2018) 15186–15191
www.materialstoday.com/proceedings
ICAPMA_2017
Etching behaviors of tunneling magnetoresistive (TMR) materials by ion beam etching system N. Wongpanita, P. O. Å. Perssonb, S. Tungasmitaa,* a
b
Department of Physics, Faculty of Science, Chulalongkorn University, Bangkok 10330, Thailand Department of Physics, Chemistry and Biology, Linköping University, S-581 83 Linköping, Sweden
Abstract In data storage manufacturing, ion etching process is commonly chosen for the preparation of surface patterns, for example, under the read-write head of the hard disk drive (HDD) to form an air bearing surface (ABS). This read-write head consists of a multilayer thin film of different materials to form a tunneling magnetoresistive (TMR) device. In this work, we applied the Monte Carlo-based simulation package to calculate the etching yield of major materials in the HDD head’s structure. The plasma characteristic in the industrial-size ion beam etching (IBE) system has been studied by the special plasma diagnostic probes. Effects of input parameters in IBE system such as incident beam angle, ion beam current and gas flow in the system were considered. Plasma characteristic can be obtained from plasma I-V curve characterization. The sputtering yields of materials were found maximum when an incident angle is about 70° to the normal surface. The floating potential of plasma in the system with the ion/electron compensation from PBN is calculated. © 2018 Elsevier Ltd. All rights reserved. Selection and/or Peer-review under responsibility of 3rd International Conference on Applied Physics and Materials Applications. Keywords: Plasma characterization; Ion beam etching; Multilayer thin film process
1. Introduction In hard disk drive (HDD), the data reading and writing functions are operated by heads, called slider, which are made from multilayer thin films of different materials. During operation of HDDs, the slider is flying above the data storage disk surface by a few nanometers [1,2]. The modern read/write head uses the tunneling magnetoresistive (TMR) effects, which change the resistance of a material in the presence of magnetic field [3]. The HDD slider
* Corresponding author. Tel.: +66-(0)89-0414998; fax: +66-(0)2-2531150. E-mail address: [email protected] 2214-7853 © 2018 Elsevier Ltd. All rights reserved. Selection and/or Peer-review under responsibility of 3rd International Conference on Applied Physics and Materials Applications.
N. Wongpanit et al./ Materials Today: Proceedings 5 (2018) 15186–15191
15187
fabrication processes play an important role to the quality and reliability of HDDs. The ion beam etching (IBE) is one of the common process to fabricate a special surface pattern on the face-down side of the slider called air bearing surface (ABS) [4]. In this work, we applied the Monte Carlo-based simulation package – SRIM (the stopping and range of ions in matter) to calculate the etching yields of important material layers within that head structure [5-6]. The etching yields of the different magnetic materials in TMR device, such as CoFe, and NiFe, as illustrated in Fig.1, were calculated [3, 7-8]. For the fabrication process, the plasma characterizations of an industrial-size ion beam etching (IBE) system were measured and studied by using the special plasma diagnostic method. The essential plasma parameters - plasma ion types, ion energies, angles of ion beam and materials - were input into the calculations. The effects of those parameters were focused to the obtained etching yields along with the energy uses/losses in ion-surface phenomena during an etching process [5]. The plasma characteristics such as a floating potential, a plasma potential, and an ion saturation current in the system with the ion/electron compensation can be extracted from the plasma I-V characteristic curve [9]. These plasma characteristics can indicate the plasma behaviors of an industrial-size IBE system that used in the actual production. 2. Simulations on plasma etching mechanism As surface of any materials is knocked by an energetic ion or a neutral atom with a mass Mi, and its energy is more than a surface binding energy Eb, the surface atoms can leave. This phenomenon is called ion bombardment or sputtering or ion etching which a momentum and energy can be exchanged. The simplest definition of an etching process thus is a process whereby atoms are ejected from a solid target material due to a bombardment of the target by energetic particles [10]. In our calculations, about 100,000 molecules of argon (Ar) gas were accelerated toward the different magnetic target materials. The mean numbers of atom removed from the target per an incident ion is called etching yield that can be calculated by the following expression: S=
3α
(4M i M t )
4π 2 (M i + M t ) 2
E1 cos 2 θ Eb
(1)
where Mt is the mass of a target atom, α is a monotonic increasing function of Mi/Mt and E1 is the energy of incident ions (< 1 keV) [10].
Fig. 1. A TMR device contains many different materials such as NiFe, CoFe and Al2O3.
From the calculated results, Fig. 2 and Fig. 3 show the plots of the etching yield versus angle of incident for 4 different kinetic energies of argon ions. The etching yield increases with ion impact angle - and reach a maximum yield around 60° - 70° to the surface plane. The right y-axis of the graph is the number of backscattered ions as a function of incident angle. There is the highest percentage of incident ions that transformed to backscattered ions as the incident angle approaches 90°. Therefore, to decrease the number of backscattered ions, the calculated optimum yield can be obtained when the incident angle of bombarded ion is about 65°.
15188
N. Wongpanit et al./ Materials Today: Proceedings 5 (2018) 15186–15191
Fig. 2. Plots of the calculated etching yields and backscattered ions versus angle of the argon ions incident on CoFe material using SRIM.
Fig. 3. Plots of the calculated etching yields and backscattered ions versus angle of the argon ions incident on NiFe material using SRIM.
3. Experimental details The vacuum based pressure of an industrial-size IBE system, used in this study, has been pumped to 3.8 x 10-7 torr. To measure the plasma I-V curve, the plasma probes were installed at an actual specimen holder position inside the vacuum chamber, shown in Fig. 4(a). Normally, the flat probe is suitable for measuring the ions while the wire probe is more suitable to measure electrons. Both of them are made from metal and are connected by a copper cable with a ceramic sheath to a circuit which is illustrated in Fig. 4(b).
15189
N. Wongpanit et al./ Materials Today: Proceedings 5 (2018) 15186–15191
(a)
(b)
Fig. 4. (a) A schematic diagram of the experimental setup of plasma characterization in the IBE system and (b) the schematic circuit of the special flat/wire plasma probe’s circuit.
In the slider fabrication process, argon ions were accelerated - from a source through a set of ion-optic grids and eventually hit the surface of materials. In order to characterize ions that reach the target surface, the plasma probes were installed at the specimen holder position. The I-V curves in any desired conditions were plotted from the measured data – calculated current versus bias potential. The incident angles are fixed at 120°, 90° (normal to the surface) and 70°. Next parameter is the beam voltage in this system that is the potential of a screen grid aimed to extract the ions. The kinetic energy of bombarded ions is determined by this potential. Both the beam current and beam voltage were set as 400, 600, 800 mA and 400, 600, 800 volts, respectively. Other parameters causing a change in this system are the gas flow into a plasma bridge neutralizer (PBN) and into an ion source. The argon gas was fed into a PBN unit which is the electron source in an IBE system to generate free electrons. The rate flow of an argon gas into a PBN was varied to 4, 6 and 8 sccm. The last parameter is the argon gas that is flowed into an ion source that used to generate plasma. This rate flow was varied to 16, 20 and 24 sccm, respectively. 4. Results and discussion
(a)
(b)
Fig. 5. (a) Plasma I-V curves at different incident angles in the IBE system at grid potential of 800 V, beam current of 400 mA. (b) The linearly fitting of I-V curve to determine a plasma potential at incident angle of 90° (normal to the surface).
The angular-dependent plasma I-V plotted at the impact angle of 70°, 90° and 120° are shown in Fig. 5(a). The saturation part of those curves depends on an incident angle of ion beam. Floating potentials (Vf) for all cases is
15190
N. Wongpanit et al./ Materials Today: Proceedings 5 (2018) 15186–15191
about 3 V. As the ions from a source collide with the probe, the kinetic energy is transferred together with their charges and momentums, which create the ion current flows into the probe. Similar to electrons, they create an electron current into the probe circuit during positive bias applied. At the normal angle to a surface, energetic ions can transfer their energy to a probe more than at other angles. In case of the incident ion is normal to the surface (90°), an ion current saturates maximally about 4.2 mA, while the other angles saturate at about 4.0 mA. At the angle of 80° and 100°, it can be implied as the same angle. The difference of them is just the direction of an incident beam which ions come from the top and bottom. Hence, the ion saturation current is almost the same for angles of 70° and 120°. The plasma potential (Vp) can be extracted from the I-V plot as shown in the Fig. 5(b). It shows the plasma potential is about 2.5 V. The potential of the acceleration grid or the beam voltage can be used to estimate the maximum kinetic energy of the argon ions. If the negative bias is applied to the probe, only ions that have enough energy can pass through an operating chamber. However, as the grid potential increased, the energy and the number of ions are not increased linearly. The energy and the number of ions do not only depend on a voltage, but also depend on other plasma factors such as beam current, as shown in Fig. 6(a). To study the effect of the grid beam voltages, the plasma I-V curves at different beam voltages were measured with fixed beam current at 400 mA. The numbers of ions generated in the plasma source were considered constant. On the other hands, as the beam voltages were set at 400 V and 800 V. The plasma I-V curve, shown in Fig. 6(b), indicates the beam voltage does not result in a change of ion saturation current. Even the grid voltage was increased twice, from 400 V up to 800 V, the ion saturation currents of both cases were still remained the same. This can be due to the plasma current was fixed by the beam current via the grid potential. On the other hand, electron saturation current increased when a voltage biased to a screen grid was low. This phenomenon may be caused by a small electric field between the first two grids, confined electrons in an ion source part.
(a)
(b)
(c)
Fig. 6. (a) Plasma I-V curves at different beam currents, (b) plasma I-V curves at different beam voltages and (c) combinations of beam voltage and current, simultaneously. The ion impact angle was at normal incident angle for all cases.
Thus, the beam current is an important parameter to set and tune the energy of plasma species in an ion source. In order to generate the plasma, a current is biased to accelerate free electrons. Free electrons ionize argon gas in a source and the plasma occurs, eventually. As the current is increased with the fixed beam voltage, the amount of plasma species starts to rise up. This causes the numbers of energetic ions that though from grids to a chamber increase. Hence, ions in the chamber are much if the beam current increases. As seen in Fig. 6(a), the ion current saturated at about 4 mA with the beam current fixed to 400 mA and increased to about 5.9 mA as the beam current increased to 800 mA. Even if the ion saturation current was different, a floating potential and an electron saturation current of both cases were almost equal. Although the number of electrons that was generated in a source is much, they cannot through a set of grids into an operating chamber. The numbers of electrons in a chamber were not affected by beam current due to the PBN is working as the electron source to balance the charges near the sample surface during the etching. Hence, the plasma etching efficiency does not only depend on a beam voltage, but also depends on a beam current. A comparison between
N. Wongpanit et al./ Materials Today: Proceedings 5 (2018) 15186–15191
15191
different ion beam voltages and currents is shown in Fig. 6(c). It clearly indicates that as a beam voltage and current increase, ions passing out grids are obviously increased. The flat probe can detect more ion saturation current when these two factors are increased. 5. Conclusion The plasma etching behaviors on magnetic materials for read-write heads of HDDs were calculated using the Monte Carlo-based simulation package at different ion impact angles and incident ion energies. The calculated results from the simulation showed that the maximum etching yield was optimized at the incident angle between 60° and 70°. The PBN’s function is to balance the charge in a beam. Thus, these I-V curves are different from the ideal. The increase of yield depends on the ion energy and composition of the magnetic materials. Increase of the ion beam voltage and current can cause energetic ions travel plentifully through grids from an ion source which influence more ion saturation current and the etching rate of the system. Acknowledgements The authors would like to acknowledge the Department of Physics, Faculty of Science, Chulalongkorn University and the Science Achievement Scholarship of Thailand (SAST) for all the supports. References [1] C. F. Yong, E. Y. K. Ng, W. D. Zhou, W. K. Ng, Scientific Research (2010) 841-854. [2] B. Liu, M. Zhang, S. Yu, L. Gonzaga, H. Y. Sim, IEEE (2002). [3] Jian-Gang (Jimmy) Zhu, Materialstoda, Elsevier (2003) 22-30. [4] N. Fukushima, T. Sato, T. Wada, Y. Horiike, IEEE Trans. Magn. Vol. 32, No. 5 (1996) 3786-3788. [5] J. P. Greene, J. Nemanich, G. E. Thomas, S. L. Schiel, Nucl. Instr. and Meth. in Phys. Res. A 397 (1997) 91-98. [6] J. F. Ziegler, M. D. Ziegler, J. P. Biersack, Nucl. Instr. And Meth. in Phys. Res. B 268 (2010) 1818-1823. [7] T. Miyajima, R. Ito, K. Honda, M. Tsukada, FUJITSU Sci. Tech. J., Vol. 46, No. 3 (2010) 273-279. [8] J. R. Childress, M. J. Carey, S. I. Kiselev, J. A. Katine, S. Maat, N. Smith, J. Appl. Phys. 99 (2006). [9] R. L. Merlino, Am. J. Phys. 75(12) (2007) 1078-1085. [10] Krishna Seshan, Handbook of thin-film deposition processes and techniques, 2nd ed., Noyes Publication / William Andrew Publishing, Norwich, New York, U.S.A., 2002. .
ScienceDirect Materials Today: Proceedings 5 (2018) 15186–15191
www.materialstoday.com/proceedings
ICAPMA_2017
Etching behaviors of tunneling magnetoresistive (TMR) materials by ion beam etching system N. Wongpanita, P. O. Å. Perssonb, S. Tungasmitaa,* a
b
Department of Physics, Faculty of Science, Chulalongkorn University, Bangkok 10330, Thailand Department of Physics, Chemistry and Biology, Linköping University, S-581 83 Linköping, Sweden
Abstract In data storage manufacturing, ion etching process is commonly chosen for the preparation of surface patterns, for example, under the read-write head of the hard disk drive (HDD) to form an air bearing surface (ABS). This read-write head consists of a multilayer thin film of different materials to form a tunneling magnetoresistive (TMR) device. In this work, we applied the Monte Carlo-based simulation package to calculate the etching yield of major materials in the HDD head’s structure. The plasma characteristic in the industrial-size ion beam etching (IBE) system has been studied by the special plasma diagnostic probes. Effects of input parameters in IBE system such as incident beam angle, ion beam current and gas flow in the system were considered. Plasma characteristic can be obtained from plasma I-V curve characterization. The sputtering yields of materials were found maximum when an incident angle is about 70° to the normal surface. The floating potential of plasma in the system with the ion/electron compensation from PBN is calculated. © 2018 Elsevier Ltd. All rights reserved. Selection and/or Peer-review under responsibility of 3rd International Conference on Applied Physics and Materials Applications. Keywords: Plasma characterization; Ion beam etching; Multilayer thin film process
1. Introduction In hard disk drive (HDD), the data reading and writing functions are operated by heads, called slider, which are made from multilayer thin films of different materials. During operation of HDDs, the slider is flying above the data storage disk surface by a few nanometers [1,2]. The modern read/write head uses the tunneling magnetoresistive (TMR) effects, which change the resistance of a material in the presence of magnetic field [3]. The HDD slider
* Corresponding author. Tel.: +66-(0)89-0414998; fax: +66-(0)2-2531150. E-mail address: [email protected] 2214-7853 © 2018 Elsevier Ltd. All rights reserved. Selection and/or Peer-review under responsibility of 3rd International Conference on Applied Physics and Materials Applications.
N. Wongpanit et al./ Materials Today: Proceedings 5 (2018) 15186–15191
15187
fabrication processes play an important role to the quality and reliability of HDDs. The ion beam etching (IBE) is one of the common process to fabricate a special surface pattern on the face-down side of the slider called air bearing surface (ABS) [4]. In this work, we applied the Monte Carlo-based simulation package – SRIM (the stopping and range of ions in matter) to calculate the etching yields of important material layers within that head structure [5-6]. The etching yields of the different magnetic materials in TMR device, such as CoFe, and NiFe, as illustrated in Fig.1, were calculated [3, 7-8]. For the fabrication process, the plasma characterizations of an industrial-size ion beam etching (IBE) system were measured and studied by using the special plasma diagnostic method. The essential plasma parameters - plasma ion types, ion energies, angles of ion beam and materials - were input into the calculations. The effects of those parameters were focused to the obtained etching yields along with the energy uses/losses in ion-surface phenomena during an etching process [5]. The plasma characteristics such as a floating potential, a plasma potential, and an ion saturation current in the system with the ion/electron compensation can be extracted from the plasma I-V characteristic curve [9]. These plasma characteristics can indicate the plasma behaviors of an industrial-size IBE system that used in the actual production. 2. Simulations on plasma etching mechanism As surface of any materials is knocked by an energetic ion or a neutral atom with a mass Mi, and its energy is more than a surface binding energy Eb, the surface atoms can leave. This phenomenon is called ion bombardment or sputtering or ion etching which a momentum and energy can be exchanged. The simplest definition of an etching process thus is a process whereby atoms are ejected from a solid target material due to a bombardment of the target by energetic particles [10]. In our calculations, about 100,000 molecules of argon (Ar) gas were accelerated toward the different magnetic target materials. The mean numbers of atom removed from the target per an incident ion is called etching yield that can be calculated by the following expression: S=
3α
(4M i M t )
4π 2 (M i + M t ) 2
E1 cos 2 θ Eb
(1)
where Mt is the mass of a target atom, α is a monotonic increasing function of Mi/Mt and E1 is the energy of incident ions (< 1 keV) [10].
Fig. 1. A TMR device contains many different materials such as NiFe, CoFe and Al2O3.
From the calculated results, Fig. 2 and Fig. 3 show the plots of the etching yield versus angle of incident for 4 different kinetic energies of argon ions. The etching yield increases with ion impact angle - and reach a maximum yield around 60° - 70° to the surface plane. The right y-axis of the graph is the number of backscattered ions as a function of incident angle. There is the highest percentage of incident ions that transformed to backscattered ions as the incident angle approaches 90°. Therefore, to decrease the number of backscattered ions, the calculated optimum yield can be obtained when the incident angle of bombarded ion is about 65°.
15188
N. Wongpanit et al./ Materials Today: Proceedings 5 (2018) 15186–15191
Fig. 2. Plots of the calculated etching yields and backscattered ions versus angle of the argon ions incident on CoFe material using SRIM.
Fig. 3. Plots of the calculated etching yields and backscattered ions versus angle of the argon ions incident on NiFe material using SRIM.
3. Experimental details The vacuum based pressure of an industrial-size IBE system, used in this study, has been pumped to 3.8 x 10-7 torr. To measure the plasma I-V curve, the plasma probes were installed at an actual specimen holder position inside the vacuum chamber, shown in Fig. 4(a). Normally, the flat probe is suitable for measuring the ions while the wire probe is more suitable to measure electrons. Both of them are made from metal and are connected by a copper cable with a ceramic sheath to a circuit which is illustrated in Fig. 4(b).
15189
N. Wongpanit et al./ Materials Today: Proceedings 5 (2018) 15186–15191
(a)
(b)
Fig. 4. (a) A schematic diagram of the experimental setup of plasma characterization in the IBE system and (b) the schematic circuit of the special flat/wire plasma probe’s circuit.
In the slider fabrication process, argon ions were accelerated - from a source through a set of ion-optic grids and eventually hit the surface of materials. In order to characterize ions that reach the target surface, the plasma probes were installed at the specimen holder position. The I-V curves in any desired conditions were plotted from the measured data – calculated current versus bias potential. The incident angles are fixed at 120°, 90° (normal to the surface) and 70°. Next parameter is the beam voltage in this system that is the potential of a screen grid aimed to extract the ions. The kinetic energy of bombarded ions is determined by this potential. Both the beam current and beam voltage were set as 400, 600, 800 mA and 400, 600, 800 volts, respectively. Other parameters causing a change in this system are the gas flow into a plasma bridge neutralizer (PBN) and into an ion source. The argon gas was fed into a PBN unit which is the electron source in an IBE system to generate free electrons. The rate flow of an argon gas into a PBN was varied to 4, 6 and 8 sccm. The last parameter is the argon gas that is flowed into an ion source that used to generate plasma. This rate flow was varied to 16, 20 and 24 sccm, respectively. 4. Results and discussion
(a)
(b)
Fig. 5. (a) Plasma I-V curves at different incident angles in the IBE system at grid potential of 800 V, beam current of 400 mA. (b) The linearly fitting of I-V curve to determine a plasma potential at incident angle of 90° (normal to the surface).
The angular-dependent plasma I-V plotted at the impact angle of 70°, 90° and 120° are shown in Fig. 5(a). The saturation part of those curves depends on an incident angle of ion beam. Floating potentials (Vf) for all cases is
15190
N. Wongpanit et al./ Materials Today: Proceedings 5 (2018) 15186–15191
about 3 V. As the ions from a source collide with the probe, the kinetic energy is transferred together with their charges and momentums, which create the ion current flows into the probe. Similar to electrons, they create an electron current into the probe circuit during positive bias applied. At the normal angle to a surface, energetic ions can transfer their energy to a probe more than at other angles. In case of the incident ion is normal to the surface (90°), an ion current saturates maximally about 4.2 mA, while the other angles saturate at about 4.0 mA. At the angle of 80° and 100°, it can be implied as the same angle. The difference of them is just the direction of an incident beam which ions come from the top and bottom. Hence, the ion saturation current is almost the same for angles of 70° and 120°. The plasma potential (Vp) can be extracted from the I-V plot as shown in the Fig. 5(b). It shows the plasma potential is about 2.5 V. The potential of the acceleration grid or the beam voltage can be used to estimate the maximum kinetic energy of the argon ions. If the negative bias is applied to the probe, only ions that have enough energy can pass through an operating chamber. However, as the grid potential increased, the energy and the number of ions are not increased linearly. The energy and the number of ions do not only depend on a voltage, but also depend on other plasma factors such as beam current, as shown in Fig. 6(a). To study the effect of the grid beam voltages, the plasma I-V curves at different beam voltages were measured with fixed beam current at 400 mA. The numbers of ions generated in the plasma source were considered constant. On the other hands, as the beam voltages were set at 400 V and 800 V. The plasma I-V curve, shown in Fig. 6(b), indicates the beam voltage does not result in a change of ion saturation current. Even the grid voltage was increased twice, from 400 V up to 800 V, the ion saturation currents of both cases were still remained the same. This can be due to the plasma current was fixed by the beam current via the grid potential. On the other hand, electron saturation current increased when a voltage biased to a screen grid was low. This phenomenon may be caused by a small electric field between the first two grids, confined electrons in an ion source part.
(a)
(b)
(c)
Fig. 6. (a) Plasma I-V curves at different beam currents, (b) plasma I-V curves at different beam voltages and (c) combinations of beam voltage and current, simultaneously. The ion impact angle was at normal incident angle for all cases.
Thus, the beam current is an important parameter to set and tune the energy of plasma species in an ion source. In order to generate the plasma, a current is biased to accelerate free electrons. Free electrons ionize argon gas in a source and the plasma occurs, eventually. As the current is increased with the fixed beam voltage, the amount of plasma species starts to rise up. This causes the numbers of energetic ions that though from grids to a chamber increase. Hence, ions in the chamber are much if the beam current increases. As seen in Fig. 6(a), the ion current saturated at about 4 mA with the beam current fixed to 400 mA and increased to about 5.9 mA as the beam current increased to 800 mA. Even if the ion saturation current was different, a floating potential and an electron saturation current of both cases were almost equal. Although the number of electrons that was generated in a source is much, they cannot through a set of grids into an operating chamber. The numbers of electrons in a chamber were not affected by beam current due to the PBN is working as the electron source to balance the charges near the sample surface during the etching. Hence, the plasma etching efficiency does not only depend on a beam voltage, but also depends on a beam current. A comparison between
N. Wongpanit et al./ Materials Today: Proceedings 5 (2018) 15186–15191
15191
different ion beam voltages and currents is shown in Fig. 6(c). It clearly indicates that as a beam voltage and current increase, ions passing out grids are obviously increased. The flat probe can detect more ion saturation current when these two factors are increased. 5. Conclusion The plasma etching behaviors on magnetic materials for read-write heads of HDDs were calculated using the Monte Carlo-based simulation package at different ion impact angles and incident ion energies. The calculated results from the simulation showed that the maximum etching yield was optimized at the incident angle between 60° and 70°. The PBN’s function is to balance the charge in a beam. Thus, these I-V curves are different from the ideal. The increase of yield depends on the ion energy and composition of the magnetic materials. Increase of the ion beam voltage and current can cause energetic ions travel plentifully through grids from an ion source which influence more ion saturation current and the etching rate of the system. Acknowledgements The authors would like to acknowledge the Department of Physics, Faculty of Science, Chulalongkorn University and the Science Achievement Scholarship of Thailand (SAST) for all the supports. References [1] C. F. Yong, E. Y. K. Ng, W. D. Zhou, W. K. Ng, Scientific Research (2010) 841-854. [2] B. Liu, M. Zhang, S. Yu, L. Gonzaga, H. Y. Sim, IEEE (2002). [3] Jian-Gang (Jimmy) Zhu, Materialstoda, Elsevier (2003) 22-30. [4] N. Fukushima, T. Sato, T. Wada, Y. Horiike, IEEE Trans. Magn. Vol. 32, No. 5 (1996) 3786-3788. [5] J. P. Greene, J. Nemanich, G. E. Thomas, S. L. Schiel, Nucl. Instr. and Meth. in Phys. Res. A 397 (1997) 91-98. [6] J. F. Ziegler, M. D. Ziegler, J. P. Biersack, Nucl. Instr. And Meth. in Phys. Res. B 268 (2010) 1818-1823. [7] T. Miyajima, R. Ito, K. Honda, M. Tsukada, FUJITSU Sci. Tech. J., Vol. 46, No. 3 (2010) 273-279. [8] J. R. Childress, M. J. Carey, S. I. Kiselev, J. A. Katine, S. Maat, N. Smith, J. Appl. Phys. 99 (2006). [9] R. L. Merlino, Am. J. Phys. 75(12) (2007) 1078-1085. [10] Krishna Seshan, Handbook of thin-film deposition processes and techniques, 2nd ed., Noyes Publication / William Andrew Publishing, Norwich, New York, U.S.A., 2002. .