ALPIDE Bending Tests Related to ALICE ITS3 WP4

ALICE is an experiment at the LHC that studies heavy-ion collisions, and it relies on an Inner Tracking System (ITS) to reconstruct particle trajectories near the interaction point with high precision. Tracking performance depends strongly on how close the innermost sensor layer sits to the interaction point and how little material the particles have to pass through. ITS3 is an upgrade that aims to rework this innermost region by bending a single, wafer-scale monolithic active pixel sensor (MAPS) into a half-cylinder that wraps around the beam pipe, instead of mechanically assembling many modules. Doing so reduces the supports and coolant and lowers the material budget that particles cross close to the interaction point. For this design to work, however, one has to confirm whether a silicon sensor still operates correctly and keeps its characteristics once it is actually bent — and the bending fixture, readout setup, and standard scans documented on this page are the local unit within that verification effort.

More specifically, ITS3 replaces the three innermost layers of ITS2 with thin, wafer-scale silicon sensors and uses lightweight supports together with air cooling. The three layers have nominal bending radii of about 18, 24, and 30 mm, and the fixture covered on this page targets the innermost of them, roughly 18 mm.

As an undergraduate researcher in the HIPEx group at Pusan National University, I worked on the fixture and measurement procedure for bending and testing a thinned ALPIDE-and-FPC assembly. These included dummy-chip mounting trials, design of the PNU Guide & Frame, 3D profile measurement, readout setup, and execution of standard scans in flat, bent, and reverse-bent configurations.

The early work used an ALPIDE dummy chip (about 30 × 15 × 0.05 mm) together with Kapton tape to examine different mounting positions and contact areas. With a deliberately small taping area, the dummy chip stayed fixed on an 18 mm cylinder for weeks, showing how little adhesive the approach needed. I then designed and printed a two-part Guide and Frame so that the FPC could be positioned on the support while minimizing direct handling of the sensor surface and wire bonds.

The Frame used in this work had a nominal bending radius of 18 mm, and this page limits its fixture and test description to that condition.

The test used a 50 μm-thick ALPIDE Bronze sample. Because the DACMONV/I pads were not connected in this assembly, the ALPIDE on-chip ADC was used for the DAC Scan. The standard scan types recorded in the presentations were:

• Power Scan
• Analog Scan
• FIFO Scan
• DAC Scan
• Digital Scan
• FHR (Fake Hit Rate) Scan
• Threshold Scan

After basic checks in the flat configuration, the sample was mounted in the PNU Guide & Frame for the bent configuration. The same types of scans were later run in a reverse-bent configuration. Measurements were taken in a dark room while keeping the settings and sequence as consistent as practical after each configuration change.

The three configurations are named so as to avoid confusion with another group's usage: flat is the unbent chip; bent (forward) has the chip curving outward, the direction another group had been using first; and reverse-bent matches the direction in which ITS3 actually wraps its sensors around the beam pipe. The same fixture was reused to bend several chips in turn, so the procedure could be repeated without handling each sensor directly.