DATA STORAGE APPLICATIONS

Effects of Cr, CrV, and CrTi underlayers on magnetic, structural and materials reliability properties of CoPt thin films
- as reprinted from Journl of Applied Physics, Volume 85, Number 8, 1999

E.W. Singleton, and P.B. Narayan
MKE-Quantum Components, Louisville, CO 80027

Wei Xiong, Ramas Raman, and Hung-Lee Hoo
Materials Research Corp., Orangeburg, NY 10962

Co80Pt20 thin films on Cr, Cr80V20, or Cr90Ti10 underlayers were studied to improve the magnetic properties for permanent magnetic stabilization layer application in AMR/GMR devices. A significant improvement of nearly 15% in coercivity was achieved for CoPt films on CrTi underlayer, when compared to either Cr or CrV underlayer. XRD pattern analysis showed that CrTi underlayer increased the intensity of CoPt (10.0) indicating enhanced crystallographic in-plane texture. With CrTi underlayer, CoPt coercivity increased linearly as Mrt was reduced from 4.3 to as low as 0.4 memu/cm2. The squareness and S* values started to drop sharply when the Mrt of CoPt film decreased to below 1.0 memu/cm2. In addition to 15% higher coercivity, another important property of CrTi/CoPt was that for CrTi underlayer thickness between 3 and 20 nm very little change in the coercivity, squareness, and S* of CoPt layer was observed. The CoPt films showed excellent corrosion resistance towards high humidity and good thermal stability up to 300C. The CoPt films with the underlayer of CrTi and CrV showed slightly less corrosion resistance towards industrial pollutants than those with no underlayer and Cr underlayer.

I. INTRODUCTION

For permanent magnet stabilization layers used in anisotropic magnetoresistance/giant magnetoresistance (AMR/GMR) structures, it is essential to achieve high coercivity at a certain remanent magnetization-thickness product (Mrt) value 1-3. Additionally, for thin films used in AMR/GMR sensors for the Information Storage Industry, the materials reliability properties such as the thermal stability and corrosion resistance need to be understood 4,5. The photolithography techniques used in the manufacture of read/write heads include exposing the thin film devices to temperatures around 250C, and the device fabrication processes also include exposure of the films to water-based cleaning solutions, high humidity atmospheres and industrial pollutants such as chlorine and sulfur compounds. Thus, materials choice for read/write heads must be optimized for magnetic and reliability characteristics.

Currently, Co-based magnetic alloys such as CoPt and CoCrPt are commonly used for stabilization layers 1,2. The effort to improve their magnetic properties and other performance characteristics has become a subject of interest. One method of improving the hard magnetic properties of Co-based thin films is to choose an appropriate underlayer on which the magnetic layer is deposited. The most commonly used underlayer material has been Cr, but recent investigations on CoCrPt media films have shown that Cr alloys of Cr-V, Cr-Ti, Cr-Mo can enhance hard magnetic properties of Co-based media 6-8. Additionally, NiAl underlayers have been investigated for longitudinal recording media 9. These approaches take advantage of the closer lattice match between CoCrPt and Cr-V, Cr-Ti, Cr-Mo or NiAl. High coercivity in CoCrPt was attributed to the enhanced in-plane orientation of CoCrPt (11.0) for CrV 6 and CoCrPt (10.0) for CrTi underlayer 7. In addition, elevated substrate temperature was found to reduce the coercivity of CoCrPt on CrTi underlayer 7. For Co80Pt20, Cr80V20 underlayer was reported to be unable to raise the coercivity of Co80Pt20 8. In fact, Cr80V20 thinner than 20 nm resulted in a dramatic drop in Co80Pt20 coercivity. Such a sharp drop in coercivity was ascribed to the dominant FCC phase found in Co80Pt20 in combination with reduced grain size. The effect of CrTi underlayer on the coercivity of Co80Pt20 has not been reported.

II. EXPERIMENT

Thin films of Co80Pt20, Co80Pt20/Cr, Co80Pt20/Cr80V20, and Co80Pt20/Cr90Ti10 (all in at.%) were deposited on oxidized silicon wafers by dc magnetron sputtering. Cast Cr, Cr80V20, Cr90Ti10 targets and a low-permeability Co80Pt20 target were used in this study. The base vacuum was better than 3.0 x 10 -7 torr and the sputtering Ar pressure was 5 mtorr. The deposition rate was 3 nm/s for CoPt layer and 2 nm/s for all the underlayers. Both CoPt and underlayer films were deposited without dc bias or substrate heating.

Film magnetic properties were measured using a vibrating sample magnetometer (VSM) with a maximum field of 10kOe. Film crystallographic structures were examined by X-ray diffraction (XRD) pattern analysis.

Thermal stability of the films were studied by annealing the bulk films at 250, 300, and 350C for 20 hours each in a flowing inert gas (nitrogen) atmosphere and then measuring the changes in magnetic properties with VSM. The thermal degradation mechanisms include solid state diffusion in the CoPt film and at the underlayer/CoPt interface, stress annealing and surface oxidation. To minimize the latter during the thermal stability testing, the CoPt films were covered with a 200-nm thick Rh layer and the effect of diffusion at the CoPt/Rh interface was assumed to be insignificant. The humidity testing was conducted by keeping the thin film samples in 85C and 80% relative humidity (RH) non-condensing atmosphere for 16 hours. The corrosive effect was evaluated by measuring the possible changes in magnetic properties by VSM and by surface topographic investigation by optical microscopy and SEM. The effect of industrial pollutants was studied by keeping the CoPt films in the MGA containing 4 parts per billion (ppb) of Cl2, 28 ppb of SO2, 36 parts of NO2 and 58 ppb of H2S at 70% RH and at 28C for six days and examining the surface with optical microscopy, SEM and EDXS.

III. RESULTS AND DISCUSSIONS

Table 1 listed magnetic properties of 40 nm thick Co80Pt20 films deposited on thermally oxidized Si wafer and on 10 nm thick Cr, Cr80V20, or Cr90Ti10 underlayer films. It is evident that the CoPt film without underlayer exhibited significantly lower coercivity, squareness and Mrt. All other CoPt films with underlayer showed improved coercivity, squareness and Mrt. CoPt films on CrV and pure Cr showedvery little change in coercivity. However, it should be noted that the CrV thickness of 10 nm was much less than the 20-nm limit reported8 and no sharp decline in CoPt coercivity was seen. A 15% increase in CoPt coercivity was observed for CrTi when compared to Cr or CrV underlayer.

Table 1:
Comparison of magnetic properties of 40 nm thick Co80Pt20 films on 10 nm thick Cr, CrV, and CrTi underlayer.

Mrt (memu/cm2)

Coercivity (Oe)

Squareness

Co80Pt20

3.3

1286

0.61

Co80Pt20/Cr

5.0

1371

0.85

Co80Pt20/Cr80V20

5.0

1370

0.85

Co80Pt20/Cr90Ti10

5.0

1540

0.85





a)


b)

c)

Fig. 1 XRD patterns of Co80Pt20 films on a) Cr90Ti10 , b) Cr80V20 and c) Cr underlayer, showing suppression of the out of plane (10.1) CoPt peak for CrV and Crti and also enhancement of CoPt (10.0) peak for CrTi.

Fig. 1 displayed XRD patterns of the same Co80Pt20 films on a) Cr90Ti10 , b) Cr80V20 and c) Cr. CoPt on Cr showed a weak shoulder at 45 degree, indicating the out-of-plane CoPt (10.1). This shoulder practically disappeared in both CoPt on CrV and CrTi. However, no evidence of the CoPt (11.0) peak as reported for CoCrPt was observed at ~ 73 degree in the CrV case, nor the dominant FCC phase as reported for CoPt on CrV thinner than 20 nm. These results were consistent with the similar coercivities obtained for CoPt films on CrV and pure Cr. Apparently, CrV under current conditions was unable to improve the coercivity of CoPt film. The effect of thicker CrV underlayer on CoPt coercivity will be studied in the future.

Like CoCrPt, CoPt film on CrTi showed an obvious enhancement in the CoPt (10.0) intensity relative to the CoPt (00.2) peak when compared to the CoPt on Cr or CrV (see Fig. 1). Therefore, the coercivity increase for CoPt on CrTi can be related to the favored growth of the in-plane CoPt (10.0) texture. In addition, Cr-Ti (110) was the only feature detected for CrTi underlayer. As indicated for CoCrPt 7, the preferential growth of Cr-Ti (110) due to the segregation of Ti could be the reason for enhanced CoPt (10.0).

Fig. 2 Magnetic properties of Co80Pt20 on 10 nm thick CrTi underlayer vs. Mrt, showing decrease of S and S* at lower Mrt.

For the high-coercivity Co80Pt20/Cr90Ti10 films, the role of CoPt or CrTi layer thickness was further evaluated. CoPt films ranging from 3 to 40 nm thick were deposited on 10 nm thick CrTi underlayer films. Fig. 2 shows a plot of the coercivity, squareness (S) and coercive squareness (S*) as functions of Mrt. The CoPt coercivity increased linearly as Mrt reduced from 4.3 to 0.4 memu/cm2. Within the same range, film squareness and S* changed very little until Mrt dropped below 1.0 memu/cm2. The rapid drop in both S and S* below 5 nm suggested the interruption of exchange coupling because of too small and isolated grains of CoPt.

Fig. 3 Coercivity, squareness and S* dependence of Co80Pt20 films on the thickness of CrTi underlayer, showing the neglifible effect of the CrTi underlayer thickness on the CoPt magnetic properties..

Fig. 3 shows the effects of Cr90Ti10 thickness on the magnetic properties of Co80Pt20 films. For CrTi thickness between 3 and 20 nm, CoPt coercivity varied slightly. Within the same thickness range, the squareness and S* of CoPt films remained basically the same. In the case of Cr/CoPt, the coercivity decreases dramatically for Cr thickness below 5 nm 10 and as a result, CrTi/CoPt has an advantage in the magnetic sensor applications.

In the thermal stability testing, the CoPt/Cr and CoPt/CrTi films did not show any significant change in VSM measurements after annealing at 250C for 20 hours or at 300C for 20 hours (Table 2). These results indicate that the thermal stability characteristics are mostly independent of the Cr or Cr-Ti underlayer. CoPt films with no underlayer produced an 4.4 and 8.0% increase in coercivity and 8.5 and 13.0% decrease in saturation magnetization after annealing at 250C for 20 hours and at 300C for 20 hours, respectively (Table 2).

Table 2:
Comparison of magnetic properties of 40 nm thick Co80Pt20 films on 10 nm thick Cr and CrTi underlayer after annealing at 250, 300, and 350° C for 20 hours.

 

% Change In

Anneal @ temp (°C)

Coercivity

Ms

Squareness

for 20 hours

250

300

350

250

300

350

250

300

350

Co80Pt20

+4.4

+8.0

+28.0

-8.5

-13.0

-15.0

+2.2

+4.1

-25.0

Co80Pt20/Cr

-0.2

-0.5

+31.0

-1.5

-1.5

-12.0

-1.0

-2.2

-6.0

Co80Pt20/Cr90Ti10

+0.1

-0.1

+28.0

-0.5

-3.0

-11.0

0.0

0.0

-6.0

The CoPt, CoPt/Cr, CoPt/CrTi, CoPt/CrV, Cr, and CrTi films did not show any chemical reaction after the high temperature/high humidity test, when studied with optical microscopy and SEM. The CoPt films did not show any significant change in VSM measurements after the test, indicating good materials reliability towards high humidity atmospheres.


(a)
(b)

Fig. 4 SEM micrographs at 5000X magnification of a) CrTi/CoPt and b) Cr/CoPt films, showing a higher corrosion pit density on CrTi/CoPt.

In the MGA testing, after six days, the CoPt/CrTi and CoPt/CrV films showed more pitting and more corrosion reactivity than the CoPt/Cr and CoPt films. As a comparison, both Cr and CrTi films did not show any evidence of chemical reaction in six days in the MGA. Figure 4 shows the SEM micrographs at 5000X magnification of the a) CoPt/CrTi and b) CoPt/Cr, indicating the higher density of corrosion pits on the former. EDXS indicated the presence of primarily chlorine and a small amount of sulfur in the corrosion pit and the corrosion product. The amount of chlorine present (which is roughly proportional to the corrosivity) on the more reactive CoPt/CrTi and CoPt/CrV is about two times larger than the other two films. The higher reactivity of the two films indicate a possible combination of larger segregation cobalt-rich phase (which is more prone to corrosion) at the grain boundaries and probable higher tensile stress.

ACKNOWLEDGMENT

The authors wish to thank Materials Research Corporation at Orangeburg New York for supplying the sputtering targets used in this study.

REFERENCES

  1. H. Edelman, G.W. Brock, T. Carr, J. Freeman and N. Smith, IEEE Trans. Magn, 26, 3004 (1990).
  2. S.H. Liao, T. Torng, and T. Kobayashi, IEEE Trans. Magn. 30, 3855 (1994).
  3. M. Lederman, IEEE Trans. Magn. 34, 549 (1998)
  4. P.B. Narayan and Y.K. Kim, IEEE Trans. Magn. 30, 3921 (1994).
  5. P.B. Narayan, R.D.Silkensen, S. Bryant and, S.Dey, IEEE Trans. Magn. 28, 2934 (1992).
  6. M.A. Parker, J.K. Howard, R. Ahlert, and K.R. Coffey, J. Appl. Phys. 73, 5560 (1993).
  7. Y. Shiroishi, Y. Hosoe, A. Ishikawa, Y. Yahisa, Y. Sugita, H. Suzuki, T. Ohno, and M. Ohura, J. Appl. Phys. 73, 5569 (1993).
  8. Y. Matsuda, Y. Yahisa, J. Inagaki, E. Fujita, A. Ishikawa, and Y. Hosoe, J. Appl. Phys. 79, 5351 (1996).
  9. James K. Howard, US Patent 4,652,499 (1987).
  10. C.A. Ross, T.P. Nolan, R. Ranjan and D. Lu, J. Appl. Phys. 81, 7441 (1997).
  11. J.X. Shen and S.H. Liao, INTERMAG Conference, FP-08 (1997).