Hyperdoping Germanium for SWIR Photodetection and High Donor Activation

Abstract

The goal of this thesis is to create new Germanium (Ge) materials that can enhance and empower the rising Ge-on-Si material platform for integrated electronic and photonic applications. We fabricate these materials through a non-equilibrium hyperdoping method consisting of ion-implantation followed by nanosecond pulsed laser melting and rapid solidification. In chapter 1, we discuss the increasing interest in use of Ge for semiconductor devices, and provide background on the specific limitations with current Ge materials this thesis seeks to address. In chapters 2 and 3 we investigate Ge hyperdoped with Gold (Au) and chalcogens (Selenium and Tellurium; Se or Te), respectively, to extend room-temperature photodetection of Ge into the sub-bandgap Short-wavelength-infrared (SWIR; 1.4–3.0 µm) region of the electromagnetic spectrum. SWIR photodetection is important for various applications, including industrial and medical imaging, astronomy, chemical sensing, and surveillance. Today, standard SWIR photodetectors are made out III-V and II-VI narrow-bandgap semiconductors, which are silicon incompatible and often expensive or toxic. These material limitations have prevented applications involving SWIR photodetection from realizing their full potential. An alternative to using III-V or II-VI materials is inducing a low-cost, silicon-compatible material such as Ge to detect SWIR light. In chapters 2 and 3 we present a scalable non-equilibrium method for hyperdoping Ge with Au or the chalcogens (Se or Te), respectively, for dopant-mediated SWIR photodetection. Using ion implantation followed by nanosecond pulsed laser melting, we obtain single-crystal materials with peak dopant concentrations several orders of magnitude above Ge-dopant solubility limits. These materials exhibit sub-bandgap absorption of light up to wavelengths of at least 3 µm, with a sub-bandgap optical absorption coefficient comparable to that of commercial SWIR photodetection materials. Previous studies of Ge-based photodetectors have reported sub-bandgap optoelectronic response only at low temperature. We show that Ge hyperdoped with Au or Se can be used for sub-bandgap SWIR photodetection at room temperature. These new materials are a potential pathway to low-cost, room-temperature, silicon-compatible SWIR photodetection. In chapter 4 we obtain high level active n+ carrier concentrations in Ge, which has been a significant challenge to further development of Ge devices. We use ion implantation of phosphorus (P) and fluorine (F) into Ge and restore crystallinity using Nd:YAG nanosecond pulsed laser melting (PLM) we demonstrate 1020 cm–3 n+ carrier concentration in tensile-strained epitaxial Ge-on-silicon. Scanning electron microscopy shows that after laser treatment, samples implanted with P have an ablated surface, whereas P+F co-implanted samples have good crystallinity and a smooth surface topography. We characterize P and F concentration depth profiles using secondary ion mass spectrometry and spreading resistance profiling. The peak carrier concentration, 1020 cm–3 at 80 nm below the surface, coincides with the peak F concentration, illustrating the key role of F in increasing donor activation. Cross sectional transmission electron microscopy of the co-implanted sample shows that the Ge epilayer region damaged during implantation is single crystal after PLM. High-resolution X-ray diffraction and Raman spectroscopy measurements both indicate that the as-grown epitaxial layer strain is preserved after PLM. These results demonstrate that co-implantation and PLM can achieve the combination of n+ carrier concentration and strain in Ge epilayers necessary for next-generation, high-performance Ge-on-Si devices. Finally, in chapter 5 we conclude by summarizing the results of this thesis and discuss opportunities for future work.