Speaker
Description
Microlens arrays (MLAs) are widely employed for beam homogenisation and shaping, either as stand-alone optics or integrated with active components such as wafer-level optics (WLO). In image sensors, each microlens—commonly called an on-chip lens (OCL)—directs incoming light toward the active area of the pixel, improving photo detection efficiency (PDE) by boosting the effective fill factor. This is especially advantageous for front-illuminated sensors with inherently low fill factors.
We present recent progress in optimizing and integrating MLAs on high-performance single-photon avalanche diodes (SPADs) and silicon photomultipliers (SiPMs). In particular, the development of the latter is conducted in the context of CERN’s next-generation Large Hadron Collider beauty (LHCb) scintillating fiber tracker (SciFi Tracker) located in a high radiation environment. Improvements of the PDE, external crosstalk and single photon time resolution are reported thanks to the MLA.
Our work tackles major challenges such as substrate variability (bare dies, packaged chips, full wafers up to 200mm), wide optical transmission (NUV to NIR), lens geometries spanning micrometers to millimeters, as well as radiation tolerance of the MLAs and operation at cryogenic temperatures. Furthermore, our approach supports multi-project wafers (MPW), allowing multiple designs to be prototyped on a single wafer. This strategy significantly lowers development costs and accelerates time-to-market for research and industry partners.
These capabilities are offered through CSEM’s MLA foundry services, providing end-to-end support from optical design and simulation to mastering, tooling, and UV replication. Our thermal reflow and UV-curing techniques guarantee ultra-smooth surfaces and precise alignment, even for complex detector architectures.
This scalable and flexible integration approach boosts next-generation photon detector performance for applications ranging from life sciences to advanced detectors for the CERN Large Hadron Collider.