SearcharxivSearch

arXiv subjects

Pernille Undrum Fathi

Publications and source records attributed to Pernille Undrum Fathi.

3 recordsLinked to original sources

Surface Modification for III-V Selective Area Molecular Beam Epitaxy of Non-Selective Mask Materials

Selective-area embedded regrowth of III-V semiconductors by molecular beam epitaxy enables the seamless integration of metals and dielectrics into crystalline material for novel design of optoelectronic devices. However, traditional masks like $SiO_2$ and $Si_{3}N_{4}$ limit the design of high-contrast photonics in the infrared due to their high extinction coefficients at technologically relevant wavelengths. Consequently, there is a need to explore alternative mask materials to expand the selective area molecular beam epitaxy capabilities beyond those traditionally used. This study evaluates the deposition selectivity of the alternative materials $Al_{2}O_{3}$, $TiO_2$, and $HfO_2$, films with preferable spectral responses but higher surface reactivity. It was found that $Al_{2}O_{3}$ exhibits promising selective growth characteristics within typical GaAs growth temperatures, $HfO_2$ demonstrated a high non-selectivity dominated by Ga adsorption on the mask at temperatures up to 650 $^\circ$C, and $TiO_2$ proved reactive during deposition. To achieve selective growth of highly non-selective and even reactive mask materials, a surface modification technique was employed to improve the selective growth characteristics of any given film. Selective growth of $Si_{3}N_{4}$ and $TiO_2$ films was achieved with the application of a thin $SiO_2$ capping layer utilizing growth conditions typical of the GaAs/$SiO_2$ system. The relationship between the thickness of $SiO_2$ caps and growth selectivity was examined, revealing that sub-1 nm capping layers can significantly influence the mask surface chemistry, indicating that by depositing a thin layer of $SiO_2$, $SiO_2$-like selectivity for any mask material can be realized without degrading its optical response.

cond-mat.mtrl-sci

Quantum-Well-Metasurface to Maximize Nonlinear Polarization

Nonlinear frequency conversion unlocks technologies ranging from telecommunications to quantum computation; however, weak nonlinearities and architectures that resist miniaturization currently limit devices. Here, we combine a bandstructure-engineered GaAs/AlGaAs heterostructure with a high quality factor dielectric metasurface to simultaneously tailor the intrinsic nonlinear susceptibility and optimize the electromagnetic field within the heterostructure. By engineering a resonant interband transition, we realize a large second-order nonlinear tensor element, 1.6 nm/V at 1.57 um wavelength. We then make it free-space-accessible and boost the effective nonlinearity to ~ 14 nm/V using a metasurface patterned on the material. Our proof-of-concept experiment establishes that interband transition engineering and metasurfaces accessing otherwise unusable nonlinear tensor elements enable giant effective nonlinearities in the near-infrared to visible spectrum. This addresses material and device-level constraints in nonlinear photonics, providing a scalable route to compact, efficient devices.

physics.optics

GenAI for Systems: Recurring Challenges and Design Principles from Software to Silicon

Generative AI is reshaping how computing systems are designed, optimized, and built, yet research remains fragmented across software, architecture, and chip design communities. This paper takes a cross-stack perspective, examining how generative models are being applied from code generation and distributed runtimes through hardware design space exploration to RTL synthesis, physical layout, and verification. Rather than reviewing each layer in isolation, we analyze how the same structural difficulties and effective responses recur across the stack. Our central finding is one of convergence. Despite the diversity of domains and tools, the field keeps encountering five recurring challenges (the feedback loop crisis, the tacit knowledge problem, trust and validation, co-design across boundaries, and the shift from determinism to dynamism) and keeps arriving at five design principles that independently emerge as effective responses (embracing hybrid approaches, designing for continuous feedback, separating concerns by role, matching methods to problem structure, and building on decades of systems knowledge). We organize these into a challenge--principle map that serves as a diagnostic and design aid, showing which principles have proven effective for which challenges across layers. Through concrete cross-stack examples, we show how systems navigate this map as they mature, and argue that the field needs shared engineering methodology, including common vocabularies, cross-layer benchmarks, and systematic design practices, so that progress compounds across communities rather than being rediscovered in each one. Our analysis covers more than 275 papers spanning eleven application areas across three layers of the computing stack, and distills open research questions that become visible only from a cross-layer vantage point.

cs.SE