GeSn as next-generation material for short-wave infrared single-photon detection
In recent years, the automotive industry has aspired to bring self-driving vehicles to the general
public and light detection and ranging (LiDAR) sensors have emerged as the preferred solution for
car vision systems. At present, LiDAR technologies employ expensive Indium-based III-V materials
for optimal performance. However, in view of future mass production of the technology, this
approach is not sustainable due to the reliance on In, a scarce element already extensively used in
the semiconductor industry.
In this context, this thesis explores the potential of GeSn as absorber material for single-photon
detection in the short-wave infrared wavelengths to replace the current commercial III-V technology
employed in LiDARs. Ge and Sn are more abundant elements compared to In, making them
a more sustainable option for single-photon avalanche photodiodes (SPADs). Furthermore, the
possibility of monolithic integration of GeSn thin films on Si platforms allows for the utilization of
lower amounts of these elements in contrast with III-V SPADs, where In constitutes the bulk of
the device. Nevertheless, the use of the GeSn semiconductor comes with fundamental material
science challenges related to the material metastability and electrically active defects arising from
the thin film growth process on Si substrates.
In this thesis, we propose to integrate a GeSn absorber on a Ge-buffered Si diode to achieve single photon
detection targeting the wavelength of 1.55 &m. We aimed to demonstrate an all-group-IV
SPAD device by epitaxially growing the Ge/GeSn absorber stack employing magnetron sputtering
as the deposition method preferred for high-volume semiconductor production.
The thesis starts with a review of the physics of SPAD devices, justifying the need of GeSn as
absorber material to access the wavelength of 1.55 &m in all-group-IV devices. Subsequently, I
present a detailed assessment of the understanding of the optoelectronic properties of Ge and
GeSn thin films in the literature, reviewing additionally the works on sputtered epitaxial Ge and
GeSn films. I then discuss the results of our scientific research in four chapters, each focused on a
different layer composing the SPAD device.
We first investigate the in situ p-type doping of GeSn by In, and show that In acts as a surfactant
during the epitaxial growth of GeSn, inducing phase separation via the formation of Sn-In liquid
droplets.
Next, we move to the bulk of the research of the thesis, which involved extensive characterization
of epitaxial Ge and GeSn films grown by the magnetron sputtering method. We demonstrate
successful epitaxy of both materials, evidencing the critical influence of the substrate lattice
mismatch in inducing defects in the film. We additionally provide characterization of the electrical
properties of GeSn, which showed to be promising but affected by high impurity levels in the films due to contamination in the employed sputtering tools.
In the third section, we demonstrate the viability of flash-lamp annealing of Ge buffers as CMOS-compatible
annealing process, shedding light on the influence of Si-Ge intermixing in determine
the final defect density.
Lastly, we present the design of a GeSn-on-Si SPAD structure and present results on their optoelectronic
characteristics with sputtered GeSn, correlating them with the material's electrical
properties.
EPFL_TH9967.pdf
n/a
restricted
copyright
66.91 MB
Adobe PDF
c73305494465176490f884762001c90a