Abstract
Interfaces between individual layers in thin films and multilayers affect mechanical, optical, electric and magnetic properties of the films. When the layer thicknesses approach the nanometer or even atomic scale, imperfections at the each of the interfaces cannot be ignored. They may consist of layer roughness, interdiffused material between the layers, and structures resulting from unwanted chemical interactions. This thesis is focused on the development of two separate, advanced analytical techniques with a high sensitivity to such interface processes: Grazing Incidence X-Ray Reflectivity (GIXRR) and in-vacuo Low Energy Ion Scattering (LEIS).
The conventional approach to GIXRR analysis lacks flexibility in its representation of the interfaces, since a priori assumptions on the interface profile are required in the model. In this thesis we used a free-form, or a model-independent approach, where the entire structure, including interfaces, is divided into multiple sublayers. The composition of each sublayer is reconstructed independently. To direct the algorithm towards physically more correct smooth interface profiles, a regularization function was additionally introduced. This allows freedom of the interface profile and absence of unrealistic abrupt features in the profiles of optical constants. The sensitivity of our approach to atomic scale interface features was demonstrated experimentally.
Unlike XRR, in-vacuo Low Energy Ion Scattering (LEIS) analysis is much less established for the analysis of buried interfaces. In this thesis, the limits of its usage are explored and expanded. The surface atomic fraction at every moment of thin film growth is found to be affected by two simultaneous phenomena: intermixing at interfaces and surface segregation. To separate these two processes, a phenomenological model of surface evolution was developed. As a result, by combining information from LEIS surface peaks and signals originating from deeper layers, interface transition effects were separated from surface segregation effects, and interface profile width as well as two segregation parameters were obtained.
In addition, also matrix effects in LEIS, arising from specific material combinations in this thesis, were investigated. Three different types of matrix effects were found and described, with three different neutralization mechanisms responsible. Despite these matrix effects, a detailed quantification of surface composition was shown to be possible. The restrictions imposed by the matrix effects limit the accuracy of buried interface analysis by in vacuo LEIS, and more fundamental research on the behavior of matrix effects in LEIS is needed to further extend the limits of LEIS metrology. This thesis is a step in this direction.
The conventional approach to GIXRR analysis lacks flexibility in its representation of the interfaces, since a priori assumptions on the interface profile are required in the model. In this thesis we used a free-form, or a model-independent approach, where the entire structure, including interfaces, is divided into multiple sublayers. The composition of each sublayer is reconstructed independently. To direct the algorithm towards physically more correct smooth interface profiles, a regularization function was additionally introduced. This allows freedom of the interface profile and absence of unrealistic abrupt features in the profiles of optical constants. The sensitivity of our approach to atomic scale interface features was demonstrated experimentally.
Unlike XRR, in-vacuo Low Energy Ion Scattering (LEIS) analysis is much less established for the analysis of buried interfaces. In this thesis, the limits of its usage are explored and expanded. The surface atomic fraction at every moment of thin film growth is found to be affected by two simultaneous phenomena: intermixing at interfaces and surface segregation. To separate these two processes, a phenomenological model of surface evolution was developed. As a result, by combining information from LEIS surface peaks and signals originating from deeper layers, interface transition effects were separated from surface segregation effects, and interface profile width as well as two segregation parameters were obtained.
In addition, also matrix effects in LEIS, arising from specific material combinations in this thesis, were investigated. Three different types of matrix effects were found and described, with three different neutralization mechanisms responsible. Despite these matrix effects, a detailed quantification of surface composition was shown to be possible. The restrictions imposed by the matrix effects limit the accuracy of buried interface analysis by in vacuo LEIS, and more fundamental research on the behavior of matrix effects in LEIS is needed to further extend the limits of LEIS metrology. This thesis is a step in this direction.
Original language | English |
---|---|
Awarding Institution |
|
Supervisors/Advisors |
|
Award date | 14 Nov 2018 |
Place of Publication | Enschede |
Publisher | |
Print ISBNs | 978-90-365-4650-8 |
Electronic ISBNs | 978-90-365-4650-8 |
DOIs | |
Publication status | Published - 14 Nov 2018 |