For use as the magnetic layer in perpendicular recording media, CoPtCr–SiO2 and/or –TiO2 granular films have been deposited by dc sputtering with a (Co)–(Pt)–(Cr)–(SiO2, –TiO2) composite target under high PAr from 4.0 to 8.0 Pa [Oikawa 2002, Uwazumi 2003, Ariake 2005, Mukai 2005, Girt 2006, Srinivasan 2008]. However, deposition under low PAr is desirable for improvement of the corrosion and impact resistance of the granular film [Johnson 2006, Tani 2008]. On the other hand, for Co-based metal–SiO2 granular films, a low PAr process resulted in the formation of strongly exchange-coupled media [Ariake 2005, Vokoun 2006] caused by the decrement of oxygen content in the films [Yamane 2005, Sasaki 2008, 2009]. This suggests that low PAr deposition of a granular film with a stoichiometric SiO2 phase is not possible using a conventional composite target with SiO2 powder as the only source material for Si and oxygen. Therefore, we propose a new type of sintered target consist of CoSi and CoO powders instead of SiO2 powder to adjust the O/Si ratio to 2 in a granular film under a low PAr process. In this paper, we report the successful fabrication of well magnetically decoupled CoPtCr–SiO2 granular media using a low PAr process.
At first, Co–Si–O ternary system was investigated in order to simplify the phase formation in a film. CoSi (melting point 1460 °C) and CoO (1935 °C) compounds were selected as raw materials for the target from the viewpoint of stability during the sintering process and independent control of Si and O content in a film. Mixed powders of (Co)–7.7 mol% (SiO2) (target A), (Co)–8.4 mol% (CoSi)–16.8 mol% (CoO) (target B), and (Co)–8.4 mol% (CoSi)–38.2 mol% (CoO) (target C) were prepared as sintered targets. The atomic composition of both the A and B targets was Co79.9Si6.7O13.4 (at%), whereas target C had a higher oxygen content than target A, but retained a relative atomic ratio of Co and Si at Co68.2Si5.7O26.1 (at%). Next, the system was expanded to Co–Pt–Cr–Si–O. The details of the quinary system are described in the latter section. The grain sizes of the (Co), (SiO2), (CoSi), and (CoO) raw materials was 0.5–2.0 μm, 0.5–1.0 μm, 0.5–3.0 μm, and 1.0–5.0 μm, respectively.
Fig. 1. SEM images of the surface of the target for A (Co)–7.7 (SiO2), B (Co)–8.4 (CoSi)–16.8 (CoO), and C (Co)–8.4 (CoSi)–38.2 (CoO) (mol%), respectively.
View All | Next
Fig. 2. Changes in (1) Co, (2) Si, and (3) O density, (4) ratio of Co to SiOx, n mol%, and (5) O/Si ratio x relative to the sputtering conditions of discharge power W and Ar gas pressure PAr for granular films made by sintered targets of A (Co)–7.7 (SiO2), B (Co)–8.4 (CoSi)–16.8 (CoO), and C (Co)–8.4 (CoSi)–38.2 (CoO) (mol%).
Previous | View All | Next
Granular films were prepared by dc magnetron sputtering on glass substrates at room temperature. The base pressure in the deposition chamber was 7 × 10−6 Pa. The sputtering conditions, i.e., discharge power (W) of 0.4–4.0 W/cm2 and PAr of 0.6–8.0 Pa, were varied in order to fabricate a film with a composition of Co79.9Si6.7O13.4 (at%) (the objective composition). The layered structure of the film was Ta (5 nm)/Pt (6 nm)/Ru (20 nm)/granular film (16 nm)/capping layer. The capping layer was deposited in order to avoid oxidation after fabrication [Sasaki 2008]. The total amount of each atomic element in the granular film was evaluated using X-ray fluorescence analysis (XRF) [Sasaki 2008, 2009]. The crystalline structure was examined by X-ray diffraction (XRD) using CuKα radiation and transmission electron microscopy (TEM). The physical grain diameter of granular film was estimated from the full-width at half-maximum (FWHM) of the in-plane XRD using Scherrer's equation.
The magnetic properties were measured using a vibrating sample magnetometer (VSM) and torque magnetometer. Perpendicular magnetic anisotropy Ku ⊥ was derived from the saturated torque coefficient of the twofold component (L2θsat) corrected by the self-energy caused by the demagnetizing field [Saito 2002]. Activation volume Vact, which corresponds with the coherent magnetic reversal volume in a medium, was determined by irreversible susceptibility measurement using the following equations [Yamanaka 1995]: Vact = (kT)/(Ms Hf), Hf = [dHr(t)/d ln (t)], where Hf is the fluctuation field, T is the absolute temperature, and k is the Boltzmann's constant.
RESULTS AND DISCUSSION
Fig. 1 shows SEM images of sintered microstructure for targets A, B, and C. Electron probe X-ray microanalysis revealed that the bright and dark microstructures correspond to (Co) and (SiO2) phases in target A, while for targets B and C, three different microstructures with contrast from lighter to darker, corresponding to (Co), (CoSi), and (CoO) phases, were present. Therefore, raw materials were found to exist even after sintering. The densities of the targets A, B, and C were high at 98.2%, 97.8%, and 99.1%, respectively.
Fig. 3. Film composition plotted on Co–Si–O ternary map for granular films prepared under various sputtering conditions using targets A, B, and C. Each target composition is shown by open star (⋆: Co79.9Si6.7O13.4 (at%) for both targets A and B) and solid star (∗: Co68.2Si5.7O26.1 (at%) for target C). (1) Films sputtered at PAr = 0.6−8.0 Pa and W = 0.4−4.0 W/cm2 (▴: using target A and □: using target B). (2) Films sputtered at PAr = 0.6 Pa and W = 0.4−4.0 W/cm2 (□: using target B and ∘: using target C).
Previous | View All | Next
Fig. 2 shows the atomic volume density of (1) Co, (2) Si, and (3) O with respect to various sputtering conditions for granular films prepared using targets A, B, and C. In addition, (4) the ratio of Co to SiOx, n (mol%), and (5) the O/Si ratio x are also shown. In Fig. 2(4)– (5), dashed lines show the nominal target composition of each target. Fig. 2(1)– (3) for target A shows that the composition of the granular film was strongly dependent on PAr and W; the Co density was decreased (67–47 × 1021 atom/cm3), whereas that of oxygen was increased (4–10 × 1021 atom/cm3) when PAr was increased from 0.6 to 8.0 Pa and W decreased from 4.0 to 0.4 W/cm2. These tendencies are similar to those observed for targets B and C.
Fig. 3 shows film composition plotted on Co–Si–O ternary map for granular films prepared under various sputtering conditions (PAr = 0.6−8.0 Pa and W = 0.4−4.0 W/cm2) using targets A, B, and C. The area of the film composition distribution for target B (□) is almost the same as that for target A (▴), as shown in Fig. 3(1), which indicates that raw materials of the target do not have a significant influence on the film composition. This result agrees with our previous study using (Co)–(Si)–(CoO) or (Co)–(Si)–(Co3O4) sintered targets [Sasaki 2009]. Fig. 3(2) shows the composition of granular films sputtered under PAr = 0.6 Pa. As seen by (∘), a granular film with a composition in the vicinity of the objective composition (⋆) can be realized using target C, even under PAr = 0.6 Pa. In the following, the structure and magnetic properties of granular films with compositions close to the objective one is discussed (see Table 1).
Table 1 Sputtering Conditions (PAr, W), Film Composition, Saturation Magnetization (MsVSM), and Anisotropy Energy (L2θsat, Ku⊥) for Granular Films Fabricated Using Targets A, B, and C.
Fig. 4 shows (1) out-of-plane and (2) in-plane XRD profiles for granular films deposited using targets A, B, and C. For all samples, the diffraction peaks from Co grains appeared only around 44.6° in (1) and 41.6° in (2). These diffractions were identified as those from hcp (00.2), and (10.0) planes, respectively, which indicates that the sheet texture of Co grains in the granular films is the c-plane.
Fig. 4. XRD profiles for sputtered Co–Si–O films using targets A (Co)–(SiO2), B (Co)–(CoSi)–(CoO) under PAr = 4.0 Pa, and C (Co)–(Si)–enriched(CoO) under PAr = 0.6 Pa on a Ta/Pt/Ru underlayer: (1) out-of-plane and (2) in-plane.
Previous | View All | Next
Fig. 5 shows an in-plane TEM image of the granular film deposited under PAr = 0.6 Pa using target C. Gray or black grains with 5–10 nm in diameter surrounded by white grain boundaries can be clearly observed in the image. Therefore, Co-based magnetic grains considered to be well isolated by SiO2-rich grain boundaries.
Fig. 5. In-plane TEM image of a Co–Si–O granular film sputtered under low PAr of 0.6 Pa using target C.
Previous | View All | Next
Table 2 Sputtering Conditions (PAr, W), Slope at Hc(α), Activation Volume (Vact), Physical Grain Diameter (GDgrain), Activation Volume (Vact), the Volume Ratio (Vact/Vgrain), where Vgrain is Physical Grain Volume of the Co–Pt–Cr–Si–O Granular Films Fabricated Using Conventional Target D and Proposed Target E.
Table 1 tabulates the film compositions, sputtering conditions (PAr, W), saturation magnetization (Ms), and anisotropy energy (L2θsat, Ku⊥) for granular films fabricated using each target. The magnitude of MsVSM is 767–820 emu/cm3, which is good correspondence with the estimated value derived by Co content on the assumption that whole Co in a film has Ms of 1422 emu/cm3 [Bozorth 1951]. The value of Ku⊥ is 2.1–2.6 × 106 erg/cm3, which is also close to the estimated value (3.3–3.7 × 106 erg/cm3), when considering the uniaxial magnetocrystalline anisotropy of 5.97 × 106 erg/cm3 for bulk hcp Co [Pauthenet 1962]. These results also support the idea that most Co grains in the granular film have c-plane orientation.
Finally, an attempt was made to apply the proposed low PAr process using a (Co)–(Pt)–(Cr)–(CoSi)–(CoO) sintered target. In Fig. 6, typical perpendicular hysteresis loops of the media using targets D and E [D: conventional target of (Co)–(Pt)–(Cr)–(SiO2) and E: proposed target of (Co)–(Pt)–(Cr)–(CoSi)–enriched(CoO)] under low PAr of 0.6 Pa are shown as D0.6 Pa and E0.6 Pa, respectively. The composition of target D was 68.1(Co)–9.2(Pt)–14.7(Cr)–8.0(SiO2), whereas, that of target E was 39.2(Co)–16.0(Pt)–10.0(Cr)–8.7(CoSi)–26.1(CoO) (mol%). The objective film composition was (Co74Pt16Cr10)–8 mol% (SiO2), which was the same as the nominal composition of target D. The hysteresis loop for a medium deposited under high PAr of 4.0 Pa using target D is also shown as D4.0 Pa for comparison. The hysteresis loop of medium D0.6 Pa has a pinched-shape. In contrast, the hysteresis loop of medium E0.6 Pa has larger coercivity (Hc) and a small slope at Hc(α). In Table 2, sputtering conditions (PAr, W), slope at Hc(α), activation volume (Vact), activation grain diameter (GDact), physical grain diameter (GDgrain), the ratio of Vact divided by physical volume (Vgrain), where Vgrain is calculated on the assumption that GDgrain of columnar grain was constant, of medium D0.6 Pa, D4.0 Pa, and E4.0 Pa are summarized. Comparing medium D0.6 Pa with E0.6 Pa, α for medium D0.6 Pa is much larger than that for medium E0.6 Pa. This fact means that medium D0.6 Pa has strong exchange coupling among magnetic grains. Furthermore, Vact and Vact/Vgrain for medium E0.6 Pa are smaller than those for medium D4.0 Pa. Considering that Vact/Vgrain means the average number of magnetically coupled grains, the present medium fabricated with the proposed target under low PAr process is quite effective on magnetic isolation in medium.
Fig. 6. Perpendicular hysteresis loops of CoPtCr–SiO2 films sputtered under low PAr of 0.6 Pa, using conventional target D (dotted line) and the proposed target E (Co)–(Pt)–(Cr)–(CoSi)–enriched(CoO) (solid line). The hysteresis loop of a film deposited under high PAr of 4.0 Pa using target D is also shown as D4.0 Pa (broken line) for comparison.
Previous | View All