2.4.2 Scanning Tunneling Microscope Lithography

The initial LON was performed on a hydrogen-passivated silicon surface using a STM in an air ambient with a positive tip bias voltage [38]. However, due to the poor reliability of the STM tip during the nano-oxidation process, very few LON studies have been performed with this technique in the mid 1990s [54], [55], [108], [134], [201], [227]. Since the STM tip cannot identify the grown oxide from the air ambient, it views a generated nanopattern as a depression on the surface. Therefore, in order for the STM to maintain its current constant, the feedback loop pushes the tip towards the surface, risking contact with the oxide [198], [200]. Many systems which use a STM in order to generate oxide nanopatterns use a subsequent AFM imaging step in order to be able to visualize the pattern which has been generated [55]. With pure STM systems, the apparent depth is considered as a measure of oxide height. It has been shown that increasing the tunneling current results in an increase in apparent depth [68]. In order for the STM to function at heights required not to contact the oxide surface, it must have an apparent depth of several nanometers [198], meaning that a nanometer-sized water meniscus forms in the area between the tip and the surface. The presence of water suggests that an electrochemistry process is responsible for the nanooxidation process and not current tunneling [52]. As the tunneling current increases, the distance between the tip and the surface decreases causing an increased electric field in the area, thus causing the oxide depth to increase. A logarithmic decay of the oxide depth is also noted when the tip speed is increased [198]. Due to the close interactions between the STM tip and the surface, tip damaging is often observed, leading to poor reliability for nanooxidation. AFM has much finer control of the tip-surface distance and is therefore seen as an improved tool for nanooxidation; AFM is explored further in the section to follow.