Pore induction of micro-scale, free-standing Germanium membranes by implantation of 4 MeV Bromine ions

Sammanfattning

The formation of pores in Germanium is a well known consequence of ion implantation, and porous Germanium is discussed as an exiting new material in a variety of applications. However, in spite of this the exact nature of the pore formation still remains largely unknown and open to discussion. By investigating Germanium layers as free-standing membranes fundamental mechanisms behind the formation of pores can be studied, and the behavior can be compared to when the Germanium layer is attached to a substrate. To these means micro-scale, nanometer thin membranes, have been fabricated using a combination of various micromachining processes. With the infrastructure of the Tandem Laboratory at Uppsala University the membranes were exposed to ion irradiation using 4 MeV Br2+ ions of various dosages with the goal of inducing pores in the membranes, and surrounding bulk material, to study the pore development as a function of the fluence. Various experimental techniques have been used to study the pores. Microbeam Rutherford backscattering spectrometry (μ-RBS) was employed to study the depth profile of the pores by using atomic layer deposition to coat the inside of the porous region with Hafnium. Scanning electron microscope was used to take pictures of both sides of the membranes and the cross-sections of the membranes and surrounding material. The pictures were used for measuring the sizes of the observed pores and nano-scale structures. Furthermore optical profilometry was utilized to measure the change in surface area after irradiation to study the strains introduced by the pore growth. Only two significantly different doses were successfully implanted, 2.1×1016 and 3.36×1016 ions/cm2. However, both doses proved successful at inducing pores 8–30 nm in size for the lowest implanted dose and 16–70 nm for the highest dose. The pores displayed a unique behavior, not previously observed, where the pores observed in the cross-sections were spherical in shape and randomly distributed throughout the depth of the membranes. These observations are in line with previous theories on pore formation where pores nucleate through microexplosions and grow in size by vacancy clustering. As the majority of the pores were situated below the surface they failed to be detected using μ-RBS. Additionally, the lack of pores in a region near the substrate provides further evidence for the role of the substrate as a pore inhibitor and increase in the membranes surface area after irradiation was observed which is linked to the volume expansion as the material transitions to the porous phase. Furthermore significant transmission sputtering was observed for the membranes, something that could be explained by near-exit microexplosions expelling material in chunks, something that fails to be accurately accounted for in theoretical predictions.