Speaker
Description
The development of hybrid silicon pixel sensors for High-Luminosity LHC experiments is entering a critical prototyping phase. The INFN IGNITE project aims to deliver next-generation 4D tracking readout chips (ROCs) featuring an intrinsic time resolution of σ < 50 ps per pixel. Recent milestones include the debut of IGNITE-64 and IGNITE-32 ROCs, supporting 4096 and 1024 channels, respectively. To accelerate R&D and optimize production costs, the project is implementing Anisotropic Conductive Film (ACF) bonding. This work discusses the use of the ACF hybridization technique, leveraged from CERN DRD3 developments, into the IGNITE workflow to achieve high-performance, cost-effective detector assemblies.
Summary (500 words)
The IGNITE project is focused on the development of next-generation 4D-tracking readout chips (ROCs) designed for high-luminosity experiments, with a particular focus on the LHCb Phase II upgrade. The long-term architectural goal involves heterogeneous 3D integration, decoupling analog and digital electronics into separate ASIC layers bonded to timing-optimized 3D silicon sensors. To validate the core electronic blocks, two ASIC prototypes have been fabricated in a 28 nm CMOS process. The IGNITE-32 (32x32 matrix) and IGNITE-64 (64x64 matrix) (Fig. 1 right), both featuring a 55 um pixel pitch and designed for sub-50 ps time resolution per pixel channel.
A primary bottleneck in the R&D chain for hybrid pixel detectors is the high cost and long turnaround time associated with traditional fine-pitch solder bump bonding. This issue is especially critical for low-volume prototypes, which typically receive lower fabrication priority compared to large production runs. To bypass these constraints, the project has considered the use of Anisotropic Conductive Film (ACF) as a low-cost, high-throughput alternative. ACF is a conductive adhesive film containing a high density of conductive particles. When placed between the sensor and ASIC and subjected to thermocompression, the particles establish electrical contact.
The AIDAinnova 3D trench-type sensor was selected as the primary candidate due to its high radiation tolerance and a proven intrinsic time resolution below 15 ps, as demonstrated in previous beam tests. Geometries featuring both continuous and discontinuous bias trenches are being compared to evaluate bonding performance and efficiency on complex surface topographies. The specific surface topology, characterized by the presence of both bias and readout electrodes on a single side, increases the complexity of the hybridization process, as any misalignment or conductive particle overflow risks fatal short circuits and the total loss of the detector module.
The hybridization effort follows two parallel tracks:
- CERN/UniGE Track: Four IGNITE-64 ROCs were assembled using AIDAinnova-64 3D trench pixel sensors. This track leverages the ACF expertise developed within the CERN DRD3 and University of Geneva framework (Fig 1,3).
- INFN-ETIC Track: A newly established facility at INFN Cagliari, part of the INFN-ETIC project (Fig. 3).
The assembly process is being systematically optimized by varying critical thermocompression parameters and surface metallization, using standard under Bumb metal, electroless plating or no surface treatment at all.
Preliminary results demonstrate the viability of the ACF technique for high-density interconnects. On the first produced devices, high contact yield has been achieved, exceeding 90% for the best-performing assemblies. The devices underwent comprehensive characterization using TCT (Transient Current Technique) and radioactive source testing using Sr/Y90. Front-end testing indicates that while the bare chip achieves an intrinsic resolution of 15 ps, the addition of the sensor and ACF interconnection results in an average system-level time resolution below 35 ps per channel due to the complete capacitance (ACF and sensor) added to the channel. These results represent the first almost complete characterization of this specific device-interconnect combination, confirming at the moment that ACF meets the stringent timing requirements for future 4D tracking applications.