改變轟擊電壓與氧含量對於MgO/NiO/Ni3Fe與SiO2/NiO/Ni3Fe微結構與交換偏壓的影響

Abstract

This thesis explains the exchange bias and microstructure on both MgO(110) single crystal and SiO2 substrate, using different bombardment voltage and different O2/Ar ratio to deposit antiferromagnetic NiO and ferromagnetic Ni3Fe. We deposit 16nm underlayer (antiferromagnetic NiO) and toplayer(ferromagnetic Ni3Fe) by dual ion beam deposition system. The substrates are 90nm SiO2 grew out from Si(100) and MgO(110) single crystal. We use different bombardment voltages of RF ion source (VRF=100 and 200), and we control two variables: the oxygen flow is 0.3sccm and the argon flow is 2.5sccm. In another experiment, we use different oxygen flows (0.2, 0.3 and 0.4sccm), and we control two variables: the bombardment voltage VRF is 100 and the argon flow is 2.5sccm. All of these depositions are done under normal temperature. From the HRTEM(high resolution transmission electron microscope analysis), the main preferred orientations of NiO and Ni3Fe are (111), (200) and (220) on the SiO2 substrate. The result from HRTEM is consistent to that from XRD. From the areas of XRD peaks, we find that with 10.71% O2, I220/ I111 ratio is 0.42 when the bombardment voltage is 100V; however, when the bombardment voltage rises to 200V, I220/ I111 ratio becomes 0.68. This result shows that as bombardment voltage rises, its preferred orientation changes from (111) to (220). Besides, we also find obvious changes on the structure from HRTEM. When the bombardment voltage is 100V, the easy axis tends to be more perpendicular to interface (the angle between easy axis and parallel interface is 55.95。) ; however, when the bombardment voltage rises to 200V, the easy axis shifts away (the angle between easy axis and parallel interface becomes 44.59。). Moreover, we also calculate the interface roughness between NiO and Ni3Fe with 10.71% O2. On the SiO2 substrate, when the bombardment voltage is 100V, the roughness is 9.1 A; when the bombardment voltage is 200V, the roughness is 7.3 A. Nevertheless, on the MgO substrate, when the bombardment voltage is 100V, the roughness is 3.9 A; when the bombardment voltage is 200V, the roughness is 2.6 A. This result shows that as the bombardment voltage rises, the interface roughness decreases. In the experiment with 7.41% O2, we find that on the SiO2 substrate, there is ferromagnetic Ni in the antiferromagnetic layer (NiO). However, on the MgO substrate, this phenomenon doesn’t happen. In the MgO substrate, we deposit 100V 10.71 %O2, 200V 10.71% O2 and 100V 7.41 %O2. From the result of HRTEM, we find that the zone axes of them are [001] and their growing directions are perpendicular to the interface with (110) orientations, which are epitaxial growth. Therefore, we confine the preferred orientations. We measure the exchange biases of 100V 10.71% O2 and 200V 10.71% O2 on both SiO2 and MgO(110) through VSM (vibrating sample magnetometer) and SQUID(superconducting quantum interference device). We find that on the SiO2 substrate, because the changes on the structure makes the easy axis shift, exchange bias increases when the bombardment voltage rises. However, on the MgO substrate, because there is no change on the structure, we conclude it is the roughness that matters. The roughness makes the exchange bias decreases when the bombardment voltage rises.