Lessons learned from Bright Pixels and the Internal Background of the EPIC pn-ccd Camera

Transcription

1 Lessons learned from Bright Pixels and the Internal Background of the EPIC pn-ccd Camera Elmar Pfeffermann, Norbert Meidinger, Lothar Strüder, Heinrich Bräuninger, Gisela Hartner Max-Planck-Institut für extraterrestrische Physik meeting 1

2 Overview 1. Generation of bright pixels during orbit #156 and the most probable explanation 2. The internal background of the pn-ccd and what can be done in future missions meeting 2

3 Facts about the bright pixel event during revolution #156 Simultaneous formation of more than 30 bright pixels Q2 Q1 Q3 Q0 Distance between bright pixels < 20 arcmin Leakage current of the hottest bright pixels increased up to A at the normal operating temperature of -90 O C No effect in the MOS1 and MOS2 cameras in rev. #156 Similar effects in MOS1 rev. #325 and in MOS2 rev. #108 meeting 3

4 Generation of bright pixels 1 Pixels are characterized as bright when the charge content exceeds the lower event threshold every 10 th readout frame Lattice defects in the silicon crystal are usually the source of bright pixels Lattice defects are either present from the beginning (production of Si-crystal), or can be generated in space environment by ionizing radiation or by a mechanical damage meeting 4

5 Generation of bright pixels 2 In the first moment the generation of lattice vacancies by ionizing radiation seems to be more likely The almost worst case ionizing event would be a 1GeV iron nucleus (just stopped in 250µm silicon) The leakage current of the lattice vacancies generated by such an event would be several orders of magnitude lower compared to the leakage current of the hottest bright pixels ( A) generated during revolution #156 in the pn-ccd meeting 5

6 Mechanical damage of the CCD by a micrometeoroid A micrometeoroid has no straight trajectory through the Wolter optics. It has to interact with the mirror surface and has several possibilities to generate a surface damage of the CCD. Disintegration of the particle during interaction with the mirror and forward ejection of fragments of the original particle. The particle gets stuck in the gold layer of the mirror surface ejecting gold particles. The particle scatter through the mirror shells and produces ejecta in the filter foil. meeting 6

7 Experimental investigation of particle interaction with a mirror surface under grazing incidence 1 At the dust accelerator facility of the MPI für Kernphysik charged iron particles of o,2-2 µm diameter can be accelerated in the electrical field of an Van-de-Graaff generator to velocities of km/s. A collimator system in front and behind the scattering mirror allows to analyze the particle interaction and to measure the scattering distribution. meeting 7

8 Experimental investigation of particle interaction with a mirror surface under grazing incidence 2 The particles scattered from the mirror surface were traced by three different methods. 1. The angular distribution of the scattered particles was measured with the collimator system and a copper target connected to a charge sensitive amplifier 2. The traces of impact of the scattered particles were registered on a polished Si-wafer later inspected with an electron microscope 3. The effect of scattered particles on an operative CCD was measured directly meeting 8

9 Experimental investigation of particle interaction with a mirror surface under grazing incidence 3 Scattering probability and geometry The scattering efficiency of the mirror for dust particles with grazing incidence angles between deg is above 80% The forward scattering angle is small in the order of 0.1 deg The scattering of the particles on the mirror does not reduce the particle speed only the charge of the particles meeting 9

10 Analysis of the scatter particle interaction with the CCD surface Electron microscope image of the impact area of a scatter particle on a silicon wafer EDX spectrum of the impact area of a scatter particle meeting 10

11 Analysis of scatter particle damage on an operative CCD The left part of the image shows the change of the dark pn-ccd image caused by the impact of the fragments of an iron particle scattered from the mirror. The right side of the image is a zoom of the impact area and only one readout frame at the moment of the impact was used for this image meeting 11

12 Background of the CCD The detector background in space is composed of several components 1. Events generated by minimum ionizing particles can be eliminated quite easily due to their high energy deposit and their pattern. 2. Charged particles generating signals in the accepted energy interval cannot be distinguished from real X-ray events. 3. Fluorescent X-rays generated by the cosmic rays in the material surrounding the CCD cannot be distinguished from cosmic X-rays focused by the mirror system on the detector. meeting 12

13 Internal background of the pn-ccd Printed circuit board mounted in a1mm gap distance behind the CCD Background image of the pn-ccd at 17.4 kev Mo-Kα meeting 13

14 What can be done? An active anticoincidence cannot be integrated in a CCD detector, without loosing too much life time It is not possible to avoid all materials producing fluorescent X-rays in the accepted energy band The only chance is to shield the sensitive areas of the CCD by appropriate shielding materials Graded-Z shielding provides an effective tool to shift the energy of locally generated X-rays to low energies, where low-z material rather produces Auger electrons than fluorescent X-rays Low energy Auger-electrons can easily be stopped in the passivation layer of the CCD, except in the area of the entrance window meeting 14

15 Combinations of shielding materials Due to the fact, that the CCD is operated at low temperatures, the shielding material should have a similar thermal expansion coefficient like silicon Material combinations like Al 2 O 3 /B 4 C, Si 3 N 4 /B 4 C or AlN/B 4 C meet the thermal requirement and the graded-z shield requirement X-RAY TRANSMISSION OF B 4 C AND Si 3 N transmission % mm B 4 C mm Si 3 N energy [kev] meeting 15

16 Conclusions 1 The internal fluorescent background can be reduced by an appropriate graded-z shielding. Shielding materials have to be carefully selected to avoid materials with high-z impurities or adjuvants meeting 16

17 Conclusions 2 The bright pixel event was most probably caused by a micrometeoroid impact meeting 17