Powered by Indico
v3.1
The 5 Seed Money Proposals for the Call 2023-2 will be presented in 10 min talks followed by 10 min discussions with the Selection Committee and interested guests.
Hybrid event but please try to attend in-person
https://tum-conf.zoom.us/j/67089822453?pwd=a1hZRVFFMXl5bVZPbG9POW5RUnhxdz09
Meeting ID: 670 8982 2453
Passcode: 851602
The overarching goal of this project is to create transient lipid vesicles based on the acyl phosphate chemistry we recently discovered. Lipids spontaneously assemble into a membrane that forms a vesicle. Those vesicles serve as a compartment for even the simplest living systems. Those modern vesicles are relatively stable compartments. Their variations are also extensively used for the construction of protocell-like compartments in order to study the processes related to living systems. They can help to understand how Life originated and if it can be recreated from scratch. A problem with these studies is that these compartments are almost always close to equilibrium. Life cannot exist there. That is why we have studied compartmentalization regulated by non-equilibrium chemical reactions. Now, for the first time, we have the chemistry that could make lipid vesicle-based compartments also out of equilibrium. In this seed project, we would like to construct such out-of-equilibrium vesicles by making mixed acyl phosphate anhydrides of aliphatic carboxylic acids and alkyl phosphates at the expense of a fuel.
Our research group focuses on understanding the origins of life by studying the essential functions associated with it, particularly the emergence of spatially organized chemical patterns that lead to mechanical changes like cell motion and division. Due to the complexity of such systems, numerical simulations are necessary for a qualitative description. Phase-field methods are a widely used class of simulations that allow the modeling of systems across multiple length scales. Implementing such simulations can be daunting, as it often requires writing custom code or using software geared toward engineers and may not cater to the needs of physicists. With LavaLamp, we plan to develop an easy-to-use, high-performant open-source API. The software will enable researchers to implement phase-field simulations easily and to customize them to fit their needs. Therefore, LavaLamp will prove a powerful tool for studying complex biological systems and the origins of life.
Gaseous micropattern detectors offer excellent spatial and temporal resolution, also at large sizes. One reason for this is the reduction of their characteristic cell sizes into the sub-mm range. As a consequence, the amount of readout channels and thus cost and power consumption scales at two-dimensional. (2D) strip-readout by 2* W/P with pitch P and width W, for pixel detectors the number scales even by the square of W/P. As for 2D strip detectors the correct strip-by- strip combination of perpendicular strips responding in a particle event is complicated, pixel geometries are preferable for full position information, unfortunately with a quadratically increasing number of readout channels. The goal of this R&D study is to develop a micropattern gaseous pixel detector with relatively few readout pixels but still very high position information by use of charge spreading on stacked pixel layers. The first layer closest to the anode consists e.g. of 1x1mm2 large pixels. At each of the. consecutive layers, the pixels are a factor of 2 larger, 2x2, 4x4mm2, etc. To preserve unique position information the corner of a readout pixel on the “next” layer is situated above the middle of a pixel on the “previous” layer. A charge signal couples capacitively from a pixel on the layer closest to the anode to superjacent pixels on the other layers. On a three-layer pcb the signal is spread from 1 pixel on the first layer to one, two, or four pixels on the third layer with characteristic patterns of pulse height. Comparing the pulse-height distributions on the pixels of the last layer, read out by the electronics, should allow for unique and excellent position information. This scheme would allow reducing the number of readout channels by a factor of 4 when using a 2 layer printed circuit board, by 16 when using 3 layers and by 64 for 4 consecutive layers. The basic principle of capacitive coupling between stacked readout layers works in Micromegas detectors. This is one of the working principles of the many 10x10cm2 2D precision tracking muon reference detectors with two crossed strip-readout planes stacked in two layers underneath a resistive anode. Thus we expect the presented charge-sharing scheme to work.
Precise magnetic field reconstruction and identification of internal and external noise sources are essential for the next generation of fundamental symmetry tests, including electric dipole moment (EDM) and Ultra-light Dark Matter (DM) searches. The HeXe2 experiment, a new spin-clock measurement using polarized 3-He and 129-Xe, sensitive to both EDMs and DM, is currently being developed. It will be based in the new magnetically shielded facility at TUM, which is funded through a 2.6 MEUR DFG grant.
To ensure that also the new generation of this experiment provides the best sensitivity in this field of research, we would like to demonstrate our new idea of 3D reconstruction of magnetic spin precession signals through an array of magnetic field sensors and advanced signal modeling methods. This will enable (i) conceptually novel insights into systematic issues of the next generation of such experiments and (ii) allow for a new method to suppress external noise sources through numerical separation of external and internal sources of signal.
The high data quality from our new approach will enable the use of new methods like AI or advanced statistical modeling and thus a concept step in this field of research. We could identify a new data acquisition unit, which can simultaneously sample 42 channels at >16-bit depth, high speeds of >100 kHz, and a GB of onboard memory while being stabilized with an atomic clock. The system is based on an enhanced version of the hardware that our group is already experienced with. To perform the project, we thus want to purchase such a new system together with several fluxgate magnetic field sensors for R&D, which also serves for optimization of the measurement environment.
The research objective of this 1-year long project is to evaluate the ultimate performance of Deep Learning strategies applied to the Imaging Atmospheric Cherenkov Telescope (IACT) technique for very-high-energy gamma-ray astronomy. We aim to compare its performance to current standard analysis strategies, and investigate potential improvements which could be achieved through a new concept of IACT design optimized to maximally exploit the power of Deep Learning.