Ion irradiation of germanium foils and germanium nanowires

In the work presented in this thesis, ion irradiations were carried in situ within a transmission electron microscope (TEM) allowing the consequences of radiation damage on germanium to be investigated. In general, the study of radiation damage on semiconductors is of utmost importance as the use of ion beams during the processing of semiconductors is now a standard technique. Furthermore, as germanium materials in general, and germanium nanowires especially, are currently being considered as replacements for bulk silicon in future microelectronic devices, this thesis will address the effects of radiation damage on both germanium thin foils and germanium nanowires.

Concerning, the use of ion beam on nanowires, a remarkable, but yet to be fully explained consequence of radiation damage, is investigated in this work: the ion induced bending (IIB) effect. In the literature, it has been reported that during ion irradiation of nanostructures they may bend towards or away from the ion beam. However, the mechanisms invoked to explain IIB are various and still debated.

Following 30 and 70 keV xenon ion irradiation experiments, it is shown in this thesis that out of the proposed mechanisms only those based on dynamical rearrangement of the damage can explain the bending of the irradiated germanium nanowires towards the ion beam. In contrast, it is demonstrated that the mechanisms based on the accumulation of point defects or on the presence of an amorphous phase cannot explain the bending of the germanium nanowires irradiated in the current work. In a set of experiments where germanium nanowires were irradiated by 30 keV xenon ions at 400°C, bending was observed even though the accumulation of point defects and the collapse of the crystalline phase into an amorphous one is prevented by the relatively elevated temperature (i.e. 400°C). Similarly, in another set of experiments performed at room temperature it is shown via Monte Carlo calculations that there is a discrepancy between the distribution of the damage within the nanowires and that which would be required in order for the mechanisms based on damage accumulation to operate.

Furthermore, the work in this thesis also solves several issues regarding the use of IIB as a potential technique in industrial processing of nanowires. Whilst IIB can be considered as an unwanted side effects occurring during the ion beam doping of nanostructures, it also represents a powerful nanomanipulation technique. However, to make full use of IIB as such a technique, the bending direction must be controllable. For this reason, using an in-house MATLAB code combined with Monte Carlo calculations it was determined that the damage depth normalised with respect to the diameter of the nanowire could be used to forecast the bending direction. Lastly, as germanium nanowires became amorphous or partially amorphous during IIB, annealing experiments were performed. However, it is shown in this work that the shape modification obtained via ion beam irradiation can be unstable during recrystallisation. Consequently, irradiating the germanium nanowires at elevated temperature (e.g. 400°C in this work) is proposed as a novel way to use the IIB effect as it allows the nanowires to maintain their single-crystalline character during the nanomanipulation.

As stated above, ion beams are routinely used to process semiconductors. However, unless the irradiation is performed at elevated temperature, the damage accumulation may induce amorphisation. To investigate the currently debated mechanisms behind amorphisation and the effect of the ion mass on the amorphization rate, germanium foils were irradiated in situ within the TEM using either 300 keV xenon, 200 keV krypton, 100 keV argon, 80 keV neon or 70 keV helium ions. By monitoring the diffraction patterns during the in situ ion irradiations and modelling the collision cascades, it was shown that amorphisation must occur gradually via a heterogeneous damage accumulation mechanism where each ion induces an amorphous region at the core of the cascade surrounded by a highly defective crystalline shell. It was also revealed that the threshold displacement per atom (dpa) for amorphisation was not always lower for heavier ions as may be expected. This feature of the threshold dpa trend was shown to depend on the spatial distribution of the point defects in a collision cascade. Furthermore, the correlation between the experiments and the modelling of 1000 collision cascades induced by helium ions has shown that, in the case of helium, the amorphisation mechanism can be understood only when taking into account the stochastic nature of collision cascades. Indeed, it is revealed that the average collision cascade induced by helium ions could not induce amorphisation. On the other hand, it is shown that it is only occasional collision cascades involving a relatively larger number of defects which are responsible for the amorphisation process.