Are you fascinated by electromagnetic modeling and numerical problem solving? Do you want to contribute to the development of state-of-the-art metrology for integrated-circuit production?
InformationIntegrated circuits are crucial for the functioning of electronic devices, providing access and storage of information (e.g. business transactions, medical diagnostics, entertainment) on which our data-driven society relies. The production of next-gen integrated circuits, with ever smaller feature sizes, introduces a whole host of challenges. One of these challenges is the effective and fast monitoring of produced wafers for compliance with the design requirements and the absence of production errors. This is the domain of wafer metrology. By illuminating a wafer and measuring the scattered fields, the scattering geometry can be reconstructed mathematically (this is called inverse scattering). This requires both sophisticated mathematical models and efficiently implemented algorithms. In the case of wafer metrology, we are dealing with layered media that potentially contain many scattering objects spread out over multiple layers.
In the computational Electromagnetics research group (the EMPMC lab), we have been pioneering the “spatial spectral method”. In this method, scattering problems, formulated as integral equations in the spatial and spectral domain, are solved by using efficient Fourier transformations. The current state of the art for this solver has shown excellent performance both timewise and in precision, but it is only applicable to a part of the challenges encountered in wafer metrology. As such, we are looking for a PhD candidate to take up the challenge of extending the capabilities of the spatial spectral methodology to more advanced wafer metrology applications.
As a PhD candidate, you will extend this method to be used in forward and inverse scattering problems applicable to wafer metrology. This will involve developing the solver to be used for complex scattering geometries, elaborate illumination profiles, and large computational domains surpassing several thousands cubic wavelengths. Furthermore, you will contribute to adapting the solver for massive parallel processing, as well as develop new algorithms for inverse scattering. These extensions will be employed in a reconstruction process on actual measured data from real-world metrology, in line with the long-term vision presented in the
IRDS roadmap for metrology.
The project will involve mathematical modeling, numerical simulations and, potentially, measurements. You will mainly do your programming work in a mixed programming environment, i.e. combining both high-level and low-level programming languages such as Python for the former and Fortran/C++ for the latter. You will be part of the
EMPMC lab, headed by prof. van Beurden, embedded within the broader Electromagnetics group.