In a silicon detector, a particle leaves a trace—a tiny electrical signal that must be captured and precisely localized. But as pixels shrink and their number grows, the signal starts to spread: like water spilled on a table, it seeps into neighboring cells. Engineers from the RD50 collaboration have proposed a solution—a maze with partitions that keeps the signal from getting lost.
Resistive LGADs are silicon diodes with internal signal amplification and resistive readout, which distributes the signal across the surface. In previous AC-coupled versions, which only pass fast signals, about 30% of the charge leaked into adjacent pixels. It's as if a third of a hammer's blow went into the floor instead of the nail. The new design uses DC coupling—a galvanic connection that passes the entire signal. The key to confinement lies in small circular electrodes and oxide trenches: grooves of silicon oxide that isolate regions and interrupt the resistive layer.
TCAD simulations showed that the trenches must run from electrode to electrode; otherwise, the signal leaks through. With high contact resistance, the signal bypasses the electrodes and escapes into neighboring pixels, so low contact resistance is critical. After irradiation with fluences up to a few times $10^{15}$ neutron equivalents per square centimeter, the fraction of charge collected by each electrode remains nearly constant. This means the resistive layer is only weakly affected by radiation, thanks to its high peak doping concentration and shallow depth.
The first batch of DC-RSD was produced by Fondazione Bruno Kessler in November 2024, and its characterization has already begun: laboratory tests and a test beam at DESY in December 2024, with another planned for March 2025. If the tests confirm the simulation predictions, this technology could become the basis for trackers at future colliders such as the HL-LHC, where detectors must operate for years under extreme radiation. The maze with partitions may turn out to be not just a fitting metaphor, but a working tool that helps physicists navigate the particle flow without getting lost.
A spatial resolution of 15 μm with a pixel pitch of 450 μm means the detector determines a particle's position 30 times more precisely than the pixel size itself. That's like being able to pinpoint where a ball lands on a tennis court using only a grid with cells the size of a car.