The deposition process

Within the main sputter chamber, 6 targets are placed which make it possible to deposit multilayer materials. The sputtering is done in a continous current process which is shown in figure 3.1. To start the deposition, an Argon gas is inserted into the chamber to increase the pressure up to 10⁻² mbar. Next a large voltage is applied across the cathode and anode, creating an Argon plasma. The Argon ions are accelerated towards the (negatively charged) target where they will collide with the target atoms. Target atoms will be released from the target itself and will be deposited on the substrate. In order to increase the ionization rate of the Argon near the targets, magnets behind the targets are used to confine electrons close to the target. This increases the ionization rate of Argon near the target resulting in

34

Figure 3.1: Schematic representation of a target system for the magnetron sputter-ing deposition technique. An Argon ion is accelerated towards the target. When the ion hits the target, an atom is knocked of which condensates at the substrate surface.

more collisions between Argon ions and target atoms which increases the deposition rate. By calibrating the deposition rate of the different target materials, control over the thickness of all the layers in complex multilayers can be obtained. In order to grow wedged samples, a wedge mask is used. During the deposition process, the wedge mask is located close to the sample surface in order to prevent shadowing.

The mask protects part of the substrate from the incoming target atoms, preventing growth below the mask. By moving the mask gradually during the deposition pro-cess, a wedged sample with varying thickness can be created as shown in figure 3.2.

The measurement of this wedge with local technics such as MOKE (see section 3.2) provides a fast method to study the thickness dependent effects.

3.1.2 Plasma cleaning

Before depositioning the materials on a SiB substrate, the substrate is cleaned by means of plasma oxidation in the oxidation chamber. After the sample is inserted in the oxidation chamber, an O2 gas is inserted in the chamber, increasing the pressure to 1.0 · 10⁻¹ mbar. A large electric field is applied, ionizing the gas and accelerating the ions towards the substrate. The ions bind with the impurities on the substrate and evaporate, removing any impurities. Since the oxidation chamber is directly connected to the main sputter chamber, in-situ cleaning is possible and the substrates will remain under high vacuum when transported to the main sputter chamber.

Figure 3.2: Illustration of the fabrication of wedged samples. The position of the wedge mask and the deposited material (blue) is shown at a) t= 0, b) 0 < t < tf inal

to c) t = tf inal.

3.2 Magnetic characterization techniques

In this section, the experimental setups that are used to characterize the magnetic properties of our samples are discussed. In section 3.2.1 the MOKE setup is de-scribed. The MOKE setup is a quick and easy method to characterize the magnetic properties of a sample. However, in our setup measurements can only be done at room temperature and a maximum field of ∼0.4 T can be applied. For measurements at higher fields and different temperatures, the VSM-SQUID is used. This setup is explained in section 3.2.2.

3.2.1 The Magneto-optical Kerr effect setup

The MOKE setup uses the fact that the polarisation (ellipticity and rotation) of light changes when it is reflected on a magnetic surface. By analysing the change in either ellipticity or rotation it is possible to extract the magnetic properties of the measured samples qualitively. In this section the MOKE setup will be presented. A schematic overview of the MOKE principle is given in figure 3.3. Light is linearly polarized before hitting the sample. Linear light is a combination of left- and right-handed circularly polarized light . When light hits a magnetic material, the left-handed circularly polarized light travels with a different speed through the magnetic material and has a different absorption rate as compared to right-handed circularly polarized light. The difference in absorbtion rate results in the linear light becoming elliptic and the difference in speed causes a phase difference, resulting in a net rotation. By analysing either the difference in rotation or the ellipticity of the reflected light as compared to the light before hitting the magnetic material, qualitive information about the magnetization in the magnetic material can be obtained.

The MOKE principle is a very easy and fast method to measure hysteresis loops of magnetic samples. Another advantage is that due to the small laser spot size (v 300µm) the magnetization is measured locally, which is needed to analyse wedges.

Figure 3.3: Schematic representation of the MOKE principle. The change in ro-tation and ellipticity after linear polarized light reflects from the surface of magnetic material provides information on the magnetization in the sample.

3.2.2 The Vibrating Sample Magnetometer - Superconduct-ing QUantum Interference Device (VSM-SQUID)

In order to characterize the magnetic properties of magnetic samples at different tem-peratures and applied fields, the Vibrating Sample Magnetometer (VSM) - Supercon-ducting QUantum Interference Device (SQUID) is used. With this setup quantitive information of the magnetization is obtained, and saturation and coercive field values can be extracted. Figure 3.4 illustrates a VSM-SQUID setup used during this thesis to characterize magnetic samples. As shown in the left part of the figure, a magnetic sample is placed between four detection coils. Vibrating this magnetic sample leads to a gradiant in the spatial magnetic field produced by the sample, causing a voltage over the detection coils. The voltage produced in the coils is proportional to the magnetic moment in the sample. This is used to magnetically characterize the sam-ple. The VSM uses a multi-coil configuration, a so-called second-order gradiometer, which removes the zeroth and first spatial derivative of the magnetic field.