Radiation hard monolithic CMOS sensors with small electrodes for High Luminosity LHC

The upgrade of the tracking detectors for the High Luminosity-LHC (HL-LHC) requires the development of novel radiation hard silicon sensors. The development of Depleted Monolithic Active Pixel Sensors targets the replacement of hybrid pixel detectors with radiation hard monolithic CMOS sensors. We designed, manufactured and tested radiation hard monolithic CMOS sensors in the TowerJazz 180 nm CMOS imaging technology with small electrodes pixel designs. These designs can achieve pixel pitches well below current hybrid pixel sensors (typically 50 × 50μm ) for improved spatial resolution. Monolithic sensors in our design allow to reduce multiple scattering by thinning to a total silicon thickness of only 50μm . Furthermore monolithic CMOS sensors can substantially reduce detector costs. These well-known advantages of CMOS sensor for performance and costs can only be exploited in pp-collisions at HL-LHC if the DMAPS sensors are designed to be radiation hard, capable of high hit rates and have a fast signal response to satisfy the 25 ns bunch crossing structure of LHC. Through the development of the MALTA and Mini-MALTA sensors we show the necessary steps to achieve radiation hardness at $10^{15}$ n$_{eq}$/cm$^2$ for DMAPS with small electrode designs. The sensors combine high granularity (pitch 36.4x 36.4μm$^2$), low detector capacitance (<5fF/pixel) of the charge collection electrode (3μm), low noise (ENC≈10e$^−$) and low power operation (1 μ W/pixel) with a fast signal response (25 ns bunch crossing). The sensors feature arrays of 512 × 512 (MALTA) and 16 × 64 (Mini-MALTA) pixels. To cope with high hit rates expected at HL-LHC (>200 MHz/cm$^2$) we have implemented a novel high-speed asynchronous readout architecture. The paper summarises the optimisation of the pixel design to achieve radiation hard pixel designs with full efficiency after irradiation at > 98% after $10^{15}$ n$_{eq}$/cm$^2$).