Detector response simulation of the CBM Micro Vertex Detector

One of the major physics topics of the CBM experiment is the study of the production of open charm in nucleusnucleus collisions at FAIR energies. For this purpose, the CBM Collaboration is planning to use a high-performance Micro-Vertex Detector (MVD) based on Monolithic Active Pixel Sensors (MAPS) [1]. CBM running conditions call for a MVD with high granularity, radiation tolerance, and readout speed. The efficient optimisation of the detector design requires detailed simulations of the detector response. So far, the MVD detector response in the CBM simulation software (CBMRoot) [2] has been described by applying a Gaussian smearing of the hit position provided by GEANT. This simple approach ignores effects like generation of clusters of pixels from impinging particles and is therefore not suited for reproducing the high track-density environment of the MVD. In the present work, we implemented a more realistic MVD detector response simulation program based on a software package which was developed recently within the framework of the ILC in order to describe the response of MAPS sensors to the passage of charged particles [3]. This package employs a MAPS response model which takes into account the sensitive volume of the sensor, its pixel structure, and its intrinsic noise. The path of a charged particle traversing the thin sensitive layer of the detector is subdivided into n segments. The energy deposited by the particle in the silicon is translated into a signal charge using a conversion factor of 3.62 eV per electron/hole pair. The diffusion of the signal electrons in the sensitive layer of the sensor is modeled by a Gaussian distribution with width tuned to measured data. This allows to convert the hit associated to the particle trajectory into a charge distributed over several pixels. It is possible to simulate the readout with a discrimination threshold and a digitisation with up to 12 bit ADC. A cluster finder algorithm was also implemented in the CBM simulation framework in order to reconstruct the hit position on the readout plane. This is done by determining the center-of-gravity of the charge stored in the fired pixels. The model was validated with real data collected with a MAPS prototype, called MIMOSA-17 (30 μm pixel pitch, 14 μm thick epitaxial layer), exposed to a 120 GeV/c pion beam at the CERN-SPS. On the left of Fig. 1, the probability is shown that the charge collected by neighbours of an already identified seed pixel (index 0) exceeds the discrimination threshold of 75 electrons (5 times the noise of MIMOSA-17) for simulated and for real data. If restricted to this comparison, an excellent agreement is observed between experiment and simulation for particles with incident angles of 40-90 with respect to the detector’s plane; nevertheless, disagreements observed for very small incident angles (10-30), e.g. reproducing the cluster shape, remain to be investigated.