Co-Cr layers for perpendicular magnetic recording

In this thesis the deposition of CO79Cr21 layers and the magnetic and structural investigations of their properties are described. These properties are correlated with the recording properties. In this way the possibilities of Co-Cr layers for high density video recording are investigated. Chapter 1 gives an introduction to the subjects of recording media, of perpendicular recording and of the use of Co-Cr layers as perpen- dicular recording media. The Co-Cr layers discussed in this thesis are prepared by RF diode sputtering. The control of the deposition process is described in chapter 2. Polyester and polyirnide flexible substrates are used. Special substrate holders are developed to be able to use these substrates. An important parameter in the deposition process is the substrate tem- perature because it influences the coercivity of the layer. With a thermostated substrate holder good control of the substrate temper- ature was obtained. The levelof impurity gases in the deposition sys- tem was sufficiently low to obtain layers with a perpendicular easy direction of magnetisation. The techniques used to investigate the magnetic and structural properties of the layers are described in chapter 3. The key parameters determined with these techniques to compare the layers with each other are the perpendicular coercivity (Hc.-L), the spread in c-axis ori- entation (L\ (Jso) and the effective anisotropy (KefT). The effective anisotropy is defined as the energy required or obtained when rotating the magnetisation in the layer frorn the in-plane direction to the per- pendicular direction. For a layer with a perpendicular easy direction ofmagnetisation the effective anisotropy is positive. In chapter 3 also the design of a vibrating sample magnetometer capable of detecting both the magnetisation in the direction of the applied field and per- pendicular to it is described. This is particularly useful for the re- cording simmulation experiments of chapter 7 and 9. The structural and magnetic properties of the Co-Cr layers depend strongly on the substrate material, as is shown in chapter 4. Layers with a perpendicular easy direction of magnetisation which are ho- mogeneous in the growth direction are obtained when a Ge underlayer of sufficient thickness is used. If no underlayer is used an initiallayer with an in-plane easy direction of magnetisation will be present. This initial layer is most pronounced in depositions directlyon the polyester substrate at elevated substrate temperatures. For polyimide substrates a much thinner underlayer is needed compared to the polyester substrates. During deposition of the layer more contam- inants emerge from the substrate when polyester is used. Ti, Si and Au underlayers also have a beneficial influence on the effective anisotropy of the Co-Cr layer . With increasing substrate temperature during deposition of the Co-Cr layers the perpendicular coercivity increases. In chapter 5 the possible origins of this increase have been studied by ferromagnetic resonance and X-ray microanalysis. The increase appears to be due to an increasing inhomogeneity of the Co-Cr layer in the lateral di- rection. However, large differences are observed between Co-Cr lay- ers deposited directlyon the polyester substrate or on Ti underlayers and the layers deposited on Ge underlayers. In the first case a Cr segregation towards the column boundaries is observed at high substrate temperatures, resulting in boundaries with a 6 at.% higher Cr concentration and which are most probably non-magnetic. For Co-Cr layers deposited on a Ge underlayer a Cr enrichment of the column center was observed at low substrate temperatures. Co-Cr layers deposited at high substrate temperatures show smaller struc- tures within the column which appear to be either enriched or de- pleted in Cr . In chapter 6 it is shown that the recording noise in Co-Cr layers depends on layer thickness and coercivity. Magnetic force microscopy shows that the high noise level for low coercive layers is caused by the influence of stripe domains on the recorded transitions, resulting in a meandering of the transitions. The width of the stripe domains present in these layers can also be inferred from the slope of hysteresis loops with the help of the Kooy and Enz model. The magnetisation inside the domain wall can account. for a large part for the in-plane remanent magnetisation. The observation of an enhanced remanence and coercivity in minor hysteresis loops compared to the major hysteresis loops can also be explained by domain wall motion. The decrease of the noi.&e level and of the enhancement effects with in- creasing coercivity can be explained by two mechanisms: a decrease in stripe domain period or a change to a more particulate-like layer in which the magnetisation reverses by incoherent rotation. Domain observations show that the domain period of a layer with a coercivity of 45 kAlm is smaller and the length of the domains is shorter com- pared to a layer with a coercivity of 15 kAlm. So most probably both mechanisms play a role. To get a high recording output the use of Co-Cr layers with a high coercivity and a high anisotropy is advantageous, as is shown in chapter 7. With such layers recording densities of 1 Mbitlmm2 are feasible. However, in the frequency response minima are observed, which are due to an overwrite interference caused by the bipolar na- ture of the perpendicular write field of a ring head, as we have shown with our vibrating sample magnetometer by simulating the recording process. In this simulation a maximum in the perpendicular magnetisation near a transition is observed. This maximum is re- sponsible for the minima in the frequency response. The in-plane field components play an important role, without these components such a maximum is not observed. The maximum decreases when the max- imum applied field is increased or when samples with a lower effective anisotropy are used. These observations are in good agreement with results from recording experiments. The simulation experiments also show that the use of layers with a tilted anisotropy axis, when re- corded in the right direction, results in a relatively lower maximum near the transition and in higher magnetisation values compared to perpendicular layers. These layers are more sensitive to one si de of the perpendicular write field of a ring head. In these layers columnar shape anisotropy is responsible for the tilted anisotropy. In chapter 8 the Co-Cr layers deposited on a soft magnetic under- layer are investigated. This combination is suitable for recording with a probe head. In this case the properties of the Co-Cr layer depend strongly on the structural properties of the underlayer and on the use of intermediate layers. In the case of Ni79Fe21 the combination of a Ti intermediate layer and bias sputtering of the Ni-Fe results in well oriented Co-Cr layers with a high perpendicular coercivity. In this case the in-plane coercivity of the Ni-Fe layer is low. The use of Ge between Co-Cr and Co86.~r4.8Nb8.8 improves the structure of the Co-Cr layer, but decreases its perpendicular coercivity to values below those when using Ti, because the crystallites are larger. The in-plane coercivity of the Co-Zr-Nb layer is low. Both kinds of underlayers with low in-plane coercivity cause spike noise in recording exper- iments with a ring head. The long-wavelength output of probe heads having a thin pole seems to be proportional to the perpendicular coercivity of the Co-Cr layer. However, for relatively thick Co-Cr layers the output starts to drop when the coercivity is increased be- yond an optimum value, probably caused by an insufficient capability of the probe head to write through the whole layer thickness. In the recording experiments with a probe head no interference effects, which cause additional minima in the frequency response, are observed. Finally, the recording properties of a Co-Cr layer with a high ef- fective anisotropy and a coercivity of about 70 kA/m are compared with the recording properties of a commercially available Co-Ni-O tape for longitudinal recording in chapter 9. The obtainable recording densities are about the same for both types of media. However, higher densities are expected for the Co-Cr layer when the overwrite inter- ference effect can be avoided, for example by using layers with a tilted anisotropy.