PRIN 2022 CALL. DECREE NO. 104 OF 2 FEBRUARY 2022
TITOLO DEL PROGETTO: A pioneering Compton Camera for hadrontherapy with a 3D silicon Pixel Chamber
CUP CODE D53D23002810006
Budget: € 86,500
P.I. or U.R. Coordinator Prof. Giuseppe Eugenio Bruno
Other Research Units University of Cagliari, INFN
Brief project description
The project proposed the development of a new generation of high-performance Compton Cameras, based on the innovative use of a silicon Pixel Chamber as a scattering detector. The objective is to overcome the limitations of traditional Compton Cameras, which require a large number of photons to reconstruct the direction of the gamma source, preventing truly online use in critical applications such as hadron therapy.
The Pixel Chamber was conceived as a three-dimensional continuous-tracking detector, obtained by superimposing numerous ultra-thin monolithic sensors with extremely high granularity. This configuration makes it possible to reconstruct, with micrometre precision, the trajectory of the recoil electron produced in the Compton interaction. Information on the tracking of the Compton electron makes it possible to significantly narrow the volume of uncertainty regarding the direction of the incident photon, reducing the reconstruction from an entire conical surface to a small cone segment, with a potential increase in imaging efficiency of several orders of magnitude.
Purpose
The project demonstrated the possibility of assembling and operating the A9, the basic unit of the Pixel Chamber consisting of a stack of 9 ultra-thin silicon ALTAI or ALPIDE chips, as conceived in the original proposal. Studies on the mechanical and electrical integration of several A9 cells to obtain volumes suitable for achieving high efficiency values for Compton interaction in the scatterer of the Pixel Chamber revealed implementation challenges when using chips with dimensions on the order of ~ 1 cm2 (the typical “field of view” dimensions of the lithographic techniques used by the foundries producing monolithic chips, such as ALPIDE or ALTAI). The possibility, thanks to the new “stitching” technique, of producing chips with much larger dimensions, up to the limits imposed by the size of the silicon wafer itself (e.g. with a diameter of 300mm in the 68nm TPSCo CMOS technology, chips measuring 10x10 cm^2 can be readily obtained
and even up to 15x15 cm^2), with all the electrical interconnections for readout and power supply located at the edge of the chip itself, would resolve the implementation challenges. With the adoption of this solution, the only critical aspect remaining is the dissipation of the heat produced: it is therefore very important to reduce power consumption per unit of area, especially when aiming to produce a stack of approximately ten large chips. Accordingly, in addition to focusing on ALTAI chips, the chip characterisation campaign was extended to the development and characterisation of new test structures and the first prototypes for large chips in 68nm CMOS technology. The results obtained demonstrated the possibility of reducing power consumption to 35 mW / cm.
Expected results
An in-depth thermal analysis was conducted on the scatterer, consisting of a 4x3 array of Pixel Chambers equipped with interdigitated heat sinks, using COMSOL simulations. The study showed that forced-air ventilation is sufficient and that the addition of external heat sinks further improves heat dissipation, especially for ventilation speeds ≥ 2 m/s, making it possible to maintain the scatterer within the operating range of 40-45 °C.
Achieved results
As part of the project, a complete simulation of the Pixel Chamber was implemented in GEANT4, including the detailed geometry of the sensor stacks with their respective alignments and inter-chip gaps. The simulation used a 1 MeV photon source, an energy representative of the range of interest for hadron therapy applications. Events useful for reconstruction were selected by requiring a single Compton interaction within the active volume and at least 10 detected track points from the recoil electron. The electron direction was reconstructed using a linear fit to the Compton interaction point and the first 3 subsequent hits along the track; this choice was motivated by the need to limit the impact of multiple scattering, which is more pronounced in the distal regions. The results obtained show a mean value of the
angular deviation of 15.7° and an RMS of 14.1°, with a distribution consistent with a significant component of well-reconstructed low-angle events. The main limiting factors identified are multiple Coulomb scattering of the recoil electron, pixel granularity, and the limited track length available for the fit. These results provide an initial quantitative benchmark of the proposed geometry and offer concrete guidance for optimizing the Pixel Chamber design in preparation for future developments.
