This thesis describes the results of a research project in the field of high bit-density data-storage media. More specifically, the material aspects of the novel recording technique using patterned media have been studied. The aim of the work was the design, realization and characterization of such a patterned medium. Chapter 1 provides a general introduction to the field of data storage. The incredible progress in bit density of today’s data storage devices is highlighted. Moreover, an outlook to future developments is given. It is shown that patterned media, due to their discrete and single domain nature, have superior recording properties and allow much higher bit densities than conventional hard disk media. In Chapter 2 the material requirements for a prototype patterned medium are given. A proper patterned medium should consist of magnetic dots which have a strong intergranular exchange coupling, large uniaxial magnetic anisotropy and low switching field distribution. Using these guidelines several candidate materials are proposed: bariumferrite, Co or Fe based alloys with L10 phase, amorphous rare earth – transition metal alloys and Co based multilayers. All four materials have a large intrinsic uniaxial magnetic anisotropy, which guarantees a sufficiently large switching field and a long-term thermal stability. Moreover, the dots of the patterned medium can be shaped in such a way that magnetostatic interactions are suppressed. In this respect, single-element Co, Ni or Fe patterned media are disadvantageous because they suffer from too large magnetostatic interactions in a densely packed 2D dot-array and therefore will limit the ultimate bit density. With respect to the amorphous alloys some caution is required, because the origin of their magnetic anisotropy is not undisputed. In addition to these qualitative considerations, the single- to two-domain transition and the switching field of Co50Ni50/Pt multilayer dots have been predicted by analytical and micromagnetic calculations: dcrit » 70 nm and Hsw ³ 500 kA/m. However, it is also shown that the latter is strongly dependent on dot shape and can be considerably smaller. In Chapter 3 the patterning technology for our studies on submicron patterned magnetic thin films is motivated and discussed. A review of existing submicron patterning technologies shows that laser interference lithography uniquely combines simplicity, large areas and sub 100 nm dimensions. With the present process sub-100 nm resist structures at a period of 200 nm can be prepared. However, at these dimensions the process latitude is limited. With a higher contrast resist in combination with an antireflective coating considerable improvement is achieved. Although further progress may be limited by instabilities of the present exposure setup, lasers with smaller wavelength and improved (substrate) stability allows the succesfull patterning of dot sizes smaller than 50 nm with our technique. In Chapter 4 an extensive study into the relation between deposition conditions, microstructure and magnetic anisotropy of sputtered Co50Ni50/Pt multilayers is presented. The degree of texture, as determined with X-Ray Diffraction, appeared to be strongly dependent on the total layer geometry and deposition pressure. At low deposition pressure the multilayers have a low atomic roughness and a strong preferred (111) texture, while at high deposition pressure the multilayers have a large atomic roughness and an additional (200) texture is present. Moreover, if deposited without underlayer, the first several nm’s of the multilayer stack have a lack of texturing, even at low deposition pressure. The magnetic anisotropy consists of interface, shape and magneto-elastic anisotropy. The magneto-elastic anisotropy contributes to the perpendicular direction, but is only 10-20% of the shape anisotropy. Therefore, interface anisotropy is the most significant contribution to the effective perpendicular magnetic anisotropy. Thin multilayers have a lower average interface anisotropy because of the ‘initial layer effects’. Besides, due to an increase in interfacial mixing by high-energy bombardment, the interface anistropy decreases towards lower deposition pressure. On the other hand, due to the increase of atomic roughness, the effective area and herewith the effective interface anisotropy decreases towards higher deposition pressure. Therefore, in order to optimize perpendicular magnetic anisotropy, the multilayers should be as thick as possible and be deposited at an intermediate pressure. Finally, based on these results and the constraints of the patterning process, the deposition conditions of the Co50Ni50/Pt multilayers for the research on patterned media were selected. This multilayer has 26 bilayers of Co50Ni50 (6 Ã…) / Pt (6 Ã…) and is deposited at 12 mbar. It has a large perpendicular magnetic anisotropy (HK,eff = (4.7±0.2)·102 kA/m) and a strong intergranular exchange coupling. Therefore, the basic requirements for a patterned storage medium have been fullfilled. In Chapter 5 the magnetization reversal in submicron and micron-sized Co50Ni50/Pt multilayer dots is described. It was checked that the patterning process (deposition-resist spinning and patterning-ion beam etching) does not deteriorate the magnetic properties of the multilayers. Instead, an improvement of the perpendicular magnetic anisotropy was found. The magnetization reversal of Co50Ni50/Pt multilayer dots is strongly dependent on dot size (i.e. diameter). In fact, over the studied range of sizes, i.e. 60 nm – 215 mm, three main types of reversal can be distinguished. For dots larger than 20 mm the reversal is dominated by small structural defects in the as deposited multilayer. These dots have a broad nucleation field distribution and domains are formed during reversal. Dots with a size between 140 nm and 20 mm are actually multi domain as well, but no domains are formed during the perpendicular (easy-axis) reversal. Finally, dots smaller than 140 nm can not be demagnetized and appear as single domain dots. Their mode of reversal is incoherent rotation. The most important conclusion of Chapter 5 is that with the preparation of 70 nm Co50Ni50/Pt multilayer dots over a large area (order of cm2 and 200 nm center-to-center spacing) a prototype of patterned storage medium has been realized. This 16 Gbit/in2 medium consists of truly single domain dots with large uniaxial magnetic anisotropy. The thermal stability is not sufficient for long-term stability, but by an optimization of the patterning process this can be enormously improved without the need for another material. Due to the lack of magnetostatic interactions between dots and the presence of a strong intrinsic perpendicular magnetic anisotropy, this patterned medium has superior storage properties compared to single-element (Co or Ni) patterned media reported in literature so far. In Chapter 6 the experimental values of properties such as critical dot diameter and switching field, as presented in Chapter 5, are compared with the (theoretical) predictions made in Chapter 2. In particular, the switching field seems to be extremely sensitive to effects of shape, etching damage and/or quality of layer growth. This leads to a switching field distribution, which is hard to control. More in general, Chapter 6 shows that in this relatively new field of research many challenges are still present. This applies both for the development of the application of magnetic dots as patterned media and for the improvement of present models of the magnetization reversal of single domain particles