Chromium–Tantalum Pentoxide Sputtering Targets and Thin Films for Seedlayer of Magnetic Recording Media
- as reprinted from Proceedings of Sony Research Forum, 1998

Hung-Lee Hoo, Wei Xiong, Peter H. McDonald, Ramas Raman, Paul S. Gilman
Materials Research Corporation, 542 Route 303, Orangeburg, New York 10962, USA
TEL: 914-398-8282, FAX: 914-398-8476, E-mail: [email protected]

Abstract --- To improve the coercivity and recording performance of cobalt alloy media on glass-ceramic disks, Cr-Ta2O5 sputtering targets and seedlayers were developed. The characteristics and performance of sputtering targets and films were studied. The impact of the different target manufacturing processes, the Ta2O5 concentration, the seedlayer thickness and the sputtering conditions is discussed. The fine and uniform grain structure of Cr-Ta2O5 films might be attributed to the less mobile Ta2O5 oxides, which prevents the high mobility Cr atoms from forming large grains.


To achieve recording density above 5 Gb/in2 and to meet the requirements especially for MR and GMR recording, media films with coercivities higher than 2500 Oe and lower noise are needed. Concurrently, glass/glass-ceramic substrates have become popular, especially for mobile computing. Besides the media and underlayer development [1-3], the search for new seedlayers used for the glass or glass-ceramic surface has been the focus of extensive studies [4]. In this paper, the preparation and characteristics of the newly developed Cr-Ta2O5 sputtering targets and thin films are reported. The deposition process of media films including the seedlayer films along with their characteristics are also discussed.


Three different manufacturing processes were developed to prepare the targets of Cr-Ta2O5 of various sizes with different concentration of Ta2O5. For relatively small quantities and large size targets, consolidating powder by hot pressing (HP) at elevated temperature in an inert atmosphere was applied. For large volume production and relatively smaller size targets, hot isostatic pressing (HIP) process is more suitable. However, for lower Ta2O5 concentration of less than 1.5 at%, and for small and medium size targets, vacuum melting (VM) can be applied to produce targets of high quality.

The HP process consolidates the pure Cr and Ta2O5 powders in a graphite die at a temperature above 1400 °C and a pressure higher than 150 kg/cm2. Target sizes up to 40 cm can be produced.

In the HIP method, pure Cr powder and Ta2O5 powder were uniformly blended and vacuum-sealed in a steel can before consolidation. During the HIP process, a peak pressure of 1055 kg/cm2 (15000 psi) and peak temperature of 1250°C or higher were applied to achieve full density.

VM is an induction melting process for preparing metal alloys. In this case, instead of preparing a metal alloy, this method has been modified to produce a composite of metal and oxide powder – the Cr-Ta2O5. Since these two ingredients have similar melting points and without a great difference between their specific weight, development of this vacuum melting process became feasible. The Ta2O5 was added into molten Cr at the later stage of induction melting. The magnetic field generated by the induction coil provided the needed stirring for the mixing of the molten Cr and the molten Ta2O5.

The microstructures of the sputtering targets prepared with different manufacturing processes were evaluated by optical microscopy. Oxides in bulk samples of Cr-Ta2O5 targets were mapped by Auger electron spectroscopy (AES) and scanning electron microscopy (SEM).

The Cr-Ta2O5 targets were evaluated using a multi-target DC magnetron sputtering system. Sputtering targets of Cr-Ta2O5 made by HP, HIP, and VM methods were tested. Recording media films of CoCrPtTa/Cr were deposited on glass ceramic substrates pre-sputtered with Cr-Ta2O5 seedlayer or NiP-plated Al platters. The base vacuum for the sputtering system was better than 3.0 x 10-7 torr. Sputter argon pressure was set at 5 mtorr for both Cr and CoCrPtTa layers and above 7.5 mtorr for Cr-Ta2O5 seedlayer. Substrate heating was applied before the deposition of Cr underlayer and Cr-Ta2O5 seedlayer. No DC bias was applied during the deposition of Cr-Ta2O5 seedlayer.

Film magnetic properties were measured on a vibrating sample magnetometer (VSM) with a maximum field of 10000 Oe. Recording performance of the disks was tested on a Guzik 1701 spinstand using thin film inductive heads. The testing track was near the inner diameter (radius of 0.92 inch)) at 5400 rpm. The data rate was 32.8 Mflux/s.

The crystalline orientations of media films were determined by X-ray diffraction (XRD). High-resolution field-emission scanning microscopy (FE-SEM) was used to study the film microstructure and textures. The forms of oxide in Cr-Ta2O5 seedlayer film were determined by X-ray photoelectron spectroscopy (XPS).


Fig. 1 Typical microstructures of the sputtering targets prepared by a), HP, b), HIP, and c), VM.

Although high purity targets of 99.999% or above are routinely produced in MRC, the current purity request for magnetic recording application is only >99.9%. Therefore raw materials were selected such that the typical purity of targets produced was ~ 99.95%. Typical impurities detected were Fe 80 - 130 ppm, and Si 150 ppm (for HP & HIP product) or Si 500 ppm (for VM product).

Full density was obtained for targets manufactured by HIP or VM process, and 95% density or better was achieved for target manufactured by HP method due to a relatively lower pressing force applied.

Microstructural evaluation indicated a two phase structure – Cr matrix and Ta2O5 particles. Typical microstructures of the sputtering targets prepared by different manufacturing processes were studied (see Fig. 1 a, b, and c).

The target made by the HP process showed the original irregular shape and relative large size of Ta2O5 particles. In the target made by HIP process, most of the original Ta2O5 particles have been deformed and squeezed into Cr grain boundaries (see SEM image in Fig.1). This is a confirmation of the full density of the target. The microstructure of targets made by VM process was different. The Ta2O5 particle shape was spherical due to its melting and re-solidification.

AES of the Cr-Ta2O5 target samples made by the three different processes are shown in Fig.2 a), b), and c), (for Cr-1.5Ta2O5 target made by HP, HIP and VM methods, respectively ). The image of oxygen and tantalum were overlapped perfectly for all the three samples. The overlapping of Ta and O2 images indicates that the oxygen and Ta in the targets are bonded together.



Fig. 2, AES mapping of Ta, O, Cr and SEM images of sputtering targets of Ta2O5 prepared by a), HP, b), HIP, and c), VM method.

Fig. 3 Effect of Cr-2.5Ta2O5 thickness on the coercivity of CoCrPtTa/Cr films on glass substrates.
Fig. 4 XRD patterns of CoCrPtTa/Cr films without and with Cr-Ta2O5 seedlayers at different thickness.
Fig. 3 displays the effect of Cr-Ta2O5 seedlayer thickness on the coercivity of CoCrPtTa/Cr films on glass substrates. For both CoCrPtTa alloys investigated in this study, a significant increase in media coercivity was achieved with Cr-Ta2O5 seedlayer thicker than 40 nm for the low-Pt alloy and 60 nm for the high-Pt alloy.

The XRD patterns shown in Fig. 4 indicated that thicker Cr-Ta2O5 seedlayer promoted the Cr (110) and in-plane Co (10.0) textures.

The FE-SEM images in Fig. 5 show the microstructure of Cr underlayer and CoCrPtTa magnetic layers with and without Cr-Ta2O5 seedlayer. It is evident that pure Cr deposited directly on glass formed large size clusters of ~ 100 nm. In comparison, Cr on Cr-Ta2O5 seedlayer showed uniform textures with fine grains of ~ 25 nm.

Fig. 5, SEM images of 80-nm thick Cr films a), directely deposited on glass, and b) on 50-nm thick Cr-Ta2O5.

Due to the epitaxial growth, the microstructure of CoCrPtTa magnetic films basically replicated that of Cr underlayer (see Fig. 6). Large clusters with wide spread cluster size were observed in CoCrPtTa film without seedlayer. In addition, the grain boundaries of CoCrPtTa without seedlayer were blurred. In contrast, CoCrPtTa film using seedlayer displayed fine grains of ~ 25nm with very narrow spread in grain size. Its grain boundaries were also well defined.

The formation of Ta2O5 in Cr-Ta2O5 seedlayer film was confirmed by seedlayer XPS results. In Fig. 7, the shift of the Ta-4f (7/2) peak in Cr-Ta2O5 film matched that for Ta2O5. On the other hand, the Cr-2p peaks were found aligned perfectly with the peaks of pure Cr, indicating no observable Cr2O3 in the seedlayer film.

(a) (b)
Fig. 6 FE-SEM images of 25-nm CoCrPtTa/ 80-nm Cr, a), directly on glass, and b), with 50-nm thick Cr-Ta2O5 seedlayer.

Cr-Ta2O5 targets with Ta2O5 concentration ranging from 0.2 at% to 3.5 at% had been sputter tested. Despite their dielectric Ta2O5 content, no sign of arcing was observed. In addition, high coercivity was achieved within a broad range of Ta2O5 concentration. Fig. 8 showed that high coercivity of CoCrPtTa/Cr films was insensitive to the variation of Ta2O5 content from 0.2 to 2.5 at.%.

Fig. 7 XPS spectra of Ta-4f (top) and Cr-2p (bottom).
Fig. 8 Coercivity of CoCr 12.5Pt12Ta1/Cr films on Cr-Ta2O5 seedlayer as a function of Ta2O5 content.

Moreover, media films of CoCrPtTa/Cr using Cr-Ta2O5 targets made by different manufacturing processes were investigated. Regardless of whether the seedlayer target was cast, hot pressed, or HIP’ed, the same high coercivity values were obtained in all cases. Therefore, target preparation methods had little impact on media coercivity.

Furthermore, high coercivity of the CoCrPtTa/Cr films was found independent of the deposition rate ranging from 1.5 to 8 nm and substrate heating from ambient temperature to 400oC. However, low coercivity was observed for sputtering argon pressure less than 7.5 mtorr.

As a result of increased coercivity and finer grains, CoCrPtTa/Cr films on glass ceramic with Cr-Ta2O5 seedlayer exhibited better signal-to-noise ratio (SNR), higher resolution (RES), and lower modulation (MOD) than that on NiP/Al substrates (see Table 1)


To prepare sputtering targets for Cr-Ta2O5 seedlayers, three different processes (HP, HIP and VM) were developed for various target production requirements. Although the microstructures of the targets made by the different processes were not identical and Ta2O5 concentrations were in a broad range, their sputtering performance was similar. AES analysis indicated that oxygen in Cr-Ta2O5 targets is bonded with Ta.

Table. 1
Magnetic and recording characteristics of CoCrPtTa/Cr films on NiP/Al and on Cr-Ta2O5 /glass.


Mrt S* PW50 (ns) SNR
NiP/Al 2711 1.1 0.92 35.7 25.4 87.2 11.3
Glass 2723 1.1 0.92 33.4 27.8 90.4 6.9

CoCrPtTa/Cr media films using a Cr-Ta2O5 seedlayer exhibited significant enhancement in coercivity and overall improvement in recording performance. These results were attributed to the observed fine grains, uniform film texture, and improved in-plane Co (100) orientation due to the Cr-Ta2O5 seedlayer. XPS spectrum confirmed the formation of Ta2O5 in the seedlayer film while showed no evidence of Cr oxide.

The high coercivity of media film was found independent of the manufacturing methods for the Cr-Ta2O5 targets and the Ta2O5 content from 0.2 to 2.5 at%.

Patents for the Cr-Ta2O5 targets and the above applications are pending in the USA and being filed in various other countries.


The authors acknowledge the manufacturing group of Materials Research Corporation’s New York facility for the preparation of Cr-Ta2O5 and other sputtering targets used in this development work.


  1. W.Xiong and H. L. Hoo, "Cobalt alloys and the search for 10 Gbit/in2 recording," Data Storage, 3(6), 47-50(1996).
  2. J.K. Howard, "Magnetic recording medium with a chromium alloy underlayer and a colbalt-based magnetic layer", U.S. Patent 4,652,499, (1987)
  3. Y. Matsuda, Y. Yahisa, J. Inagaki, E. Fujita, A. Ishikawa, Y. Hosoe, "Reduction of Co-Cr-Pt media noise by addition of Ti to Cr underlayer," J. Appl. Phys., 79(8), 5351-53, 1996
  4. W. Xiong and H.L. Hoo, "Cr-Ta2O5 Seedlayer for Recording Media on Alternative Substrates," IEEE Trans. Magn. 34, 1570-1573 (1998)