In the fourth edition of the BIS PRELUDIUM, 248 applications were submitted, and 58 of them received funding. These funds can be used for scholarships, projects and internships of foreign doctoral students. The research team consists of a doctoral supervisor (who applies for funding and is the project supervisor) and a doctoral student.

Conductivity should be high and efficient

Agnieszka Łękawa-Raus, PhD, from the Center for Advanced Materials and Technology CEZAMAT will lead the development of scalable methods for the production of acid-free liquid crystals that can be further used for spinning carbon nanotube (CNT) fibers.

These nanotubes have unique properties: electrical, thermal and mechanical. Such accumulation is highly desirable when forming various types of electrical conductors.

– There are many possible ways of combining nanotubes into macroscopic conductors, but these unique properties are then typically lost – explains Agnieszka Łękawa-Raus, PhD – Recently, a theory has been put forward that properly structured CNT fibers can overcome this problem and transfer unique properties to the macroscale.

Such an effect can be achieved by spinning the fibers from the slurries of liquid crystal carbon nanotubes. This helps to obtain the best results. However, the method has a disadvantage – superacids are used here, which are dangerous and toxic. In her project, our researcher intends to give them up.

– I want to combine my unique experience in working with CNT fibers collected abroad and the methodology of manufacturing conductive pastes and inks for applications in printed electronics developed at the Warsaw University of Technology – says Agnieszka Łękawa-Raus, PhD.

The project will be implemented in cooperation with Karolina Milowska, PhD, (Ikerbasque Research Fellow – CIC nanoGUNE, Spain), who will co-supervise the doctoral student and carry out the theoretical part of the project.

The results of the research are intended to obtain high conductivity CNT fibers and contribute to improving the performance of less conductive but inexpensive and highly versatile printed CNT conductors.

There is a competition for a doctoral student who will carry out research as part of the project. Calls can be submitted until June 15, 2023. The advertisement is available at CEZAMAT.

Artificial intelligence to create 3D video

How to use deep neural networks for three-dimensional reconstruction of films in real time? The subject will be dealt with by Associate Professor Tomasz Trzciński and a student pursuing a doctorate under his supervision.

– Our goal in this project is to increase the effectiveness of training of neural radiation fields (NeRFs), which is to lead to their wider applicability in various real applications – says the researcher from the Faculty of Electronics and Information Technology. – Our preliminary results show that it is feasible.

Neural radiation fields are a type of deep learning algorithm that is used to generate realistic-looking images of three-dimensional objects. NeRFs are able to take into account the complex parameters of lighting and image shading, thanks to which they are used in many areas: from computer graphics, through virtual reality, to medical imaging. However, they also have limitations. This primarily involves a computational complexity, which may hinder their use in large-scale or real-time applications.

Obviously, there are ways to speed up this process, but they take up a lot of memory, and the quality of the film is not the highest.

– Please note that the films contain a lot of redundant information, as many subsequent frames are similar to each other – emphasizes Associate Professor Tomasz Trzciński. So we hypothesized that we could use redundancy to encode films as radiation fields in a computationally efficient way. We want to create such a method as part of this project.

The project is based on publications carried out in cooperation with Microsoft Research Cambridge, the main author of which is a professor’s current doctoral student, Kacper Kania. The new project will be carried out in collaborationthe Faculty of Electronics and Information Technology with excellent research centres, such as the University of British Columbia or Simon Frasier University), and the doctoral student will also be looked after by a graduate of WEiTI (the Faculty of Electronics and Information Technology), currently a researcher at Microsoft Research Cambridge – Marek Kowalski, PhD. Recruitment is ongoing – interested student are encouraged to send a CV and a cover letter to: tomasz.trzcinski@pw.edu.pl (deadline June 15, 2023).

Innovative research in optical metrology

Can the statistical inference method known as Bayesian inference be used in optical metrology? Such a project will be led by Professor Maciej Trusiak from the Faculty of Mechatronics. The topic is of particular importance for the semiconductor industry. The metrological methods used there are indispensable to ensure successful production and quality verification. Optical techniques are particularly useful as they provide non-invasive means for precise and rapid measurements with nanoscale accuracy. Particularly noteworthy here are methods without scanning (such as wide-field interferometry). They enable simultaneous measurements in all camera pixels in a wide field of view, without the need to combine information from multiple separately conducted measurements.

– Information on the geometry of the semiconductor element and the distribution of the refractive index is encoded in the phase component of the stripe image (interferogram, hologram) – explains Professor Maciej Trusiak – The central aspect of this technique is the demodulation of the phase map.

The most popular demodulation technique is also problematic, as it requires precise registration of many interferograms and is partly unreliable due to the transfer of noise from the field of intensity to the field of phase. This imperfection opens the field for further exploration.

– In the project, we will examine the innovative concept of algorithmic phase demodulation using a single striatum image – says Professor Maciej Trusiak – Previous single-frame phase demodulation algorithms used Fourier, Hilbert or waveform transforms and were limited by strong numerical errors. To the best of our knowledge, the use of Bayesian inference in optical metrology is an innovative idea on a global scale and will minimize noise while maximizing the precision of phase demodulation and measuring the geometry of semiconductor structures such as waveguides.

The grant plans include not only research in Poland, but also the doctoral student's six-month trip to The Arctic University of Norway to the laboratory of Professor Balpreet Ahluwalia – a pioneer in the development of optical fluorescence nanoscopy "on-chip" technology using semiconductor waveguide structures. There are plans to start collaboration with Maciek Wielgus, PhD, from the Max Planck Institute for Radio Astronomy, a member of the Black Hole Initiative – Event Horizon Telescope – an expert in the field of numerical analysis of interferometric data.

Recruitment for a PhD student is ongoing. Applications are accepted until June 14, 2023. Detailed information can be found at laboratorium.