Silicon Simulation Code for Belle II and ILC

Abstract

Monte Carlo simulations in high-energy physics experiments face the non-trivial task of simulating realistically the response of individual detector components, while keeping the costs in terms of CPU time reasonably low. Such simulation procedures are called digitization and have to incorporate detector physics in as much detail as possible, while performing fast enough. Here, we present our approach to the simulation of the Belle II vertex detector (VXD) using the ILC software framework. We simulate the response of DEPFET (DEPleted Field Effect Transistor) pixel detectors (PXD) and double-sided silicon micro-strip detectors (SVD) in the presence of a magnetic field. To achieve sufficient performance, we use a combination of fast numerical techniques and reasonable simplification of in-detector physics. The simulation itself is divided into 3 steps: particle propagation through matter (using Geant4 with a detailed implementation of the VXD geometry), charge collection in the silicon detectors (digitization) and clustering. The second and third steps are performed in separate reconstruction modules -Marlin modules: SiPxlDigi (pixel detectors) and SiStripDigi (micro-strip detectors). The following physics processes are considered: continuous energy loss fluctuations, generation of e-h pairs, drift in electric field, diffusion, Lorentz shift, mutual crosstalk (strips), read-out/geometric pitch effect (strips), electronics/digital noise. The clustering procedure is based on the center-of-gravity and analog head-tail algorithms. All effects have been studied and compared to test beam data in order to validate the algorithms and to determine the relative importance of individual processes.

Development and construction of the Belle II DEPFET pixel detector

“…Both PXD sensors were operated in the low gain mode of the DCDB ASIC where the least significant bit had a width of 130 nA and a dynamic range of roughly 30 µA. The average in-pixel amplification for the whole sensitive area was estimated by comparing histograms for the simulated cluster charge obtained from the PXD detector simulation [19] with the measured cluster charge (right panel, figure 8). The PXD detector simulation takes into account the energy loss straggling, drift and diffusion of charge carriers in the fully depleted silicon sensor, the Lorentz shift of electrons in the magnetic field and the analog-to-digital conversion in the front-end electronics.…”

Improving the spatial resolution of NICA/MPD ECAL with new reconstruction methods

“…Charge center of gravity(CCOG) is a simple but effective method to estimate the position of the incident particles, and it is therefore extensively used in scientific position sensitive detectors [4][5][6], including the electromagnetic calorimeters [7]. Photons or electrons coming from the interaction point will produce showers in the medium of the ECal modules, leading to signal responses in a bunch of towers, which are regarded as clusters.…”

Physical Limitations to the Spatial Resolution of Solid-State Detectors

“…The result of a GEANT4 simulation of the setup with a detailed digitizer model [17], [25], [19] of the DEPFET detector response is superposed on the data curve in Figure 4. The simulation includes the effect of δ-electrons, but does not take into account imperfections in the read-out (such as non-uniformities in the gain) or measurement errors (such as an incorrect treatment of the finite telescope resolution or residual misalignment of the setup).…”

“…For a detailed description of the setup and analysis the reader is referred to Ref. [9] The resolution measurements based on these data are compared to GEANT4 [15], [16] simulations with a detailed description of the DEPFET response [17]. We note that even if the measurements correspond to a particular detector technology, the conclusions apply quite generally to position-sensitive devices based on silicon.…”