How gamma rays change the surface and core of ATLAS18 silicon sensors — signs of surface damage saturation near a few Mrad
This paper reports how gamma rays affect the new ATLAS18 silicon strip sensors that will be used in the ATLAS Inner Tracker for the High-Luminosity LHC. The authors exposed small version sensors (“minis”) and simple diode test structures to cobalt-60 gamma radiation and then measured electrical changes. They separated currents coming from the silicon bulk (the detector interior) and currents coming from the surface and edges. The study covers low doses important for the tracker’s early life and extends to much higher doses to explore where surface effects level off.
The team irradiated samples inside a charge-particle-equilibrium box with a 60Co source. They covered a dose range from 0.5 to 100 krad (kilorad), which is most relevant for early operation, and performed additional tests up to a few megarad (Mrad). Some diodes from earlier work were also included at ultra-high doses of several hundred Mrad. Dose rates were 1.6–8.5 krad/min with an uncertainty below 5%. During irradiation the samples stayed below 35 °C, and after irradiation they were stored below −20 °C to avoid uncontrolled heating effects.
To tell surface and bulk effects apart, the team used two device types. MD8 diodes have a contactable guard ring that lets the experimenters measure bulk current and surface current separately. For minis, which have 104 strips and no contactable guard ring, they measured the current through a single strip and multiplied by the number of strips to estimate the bulk contribution. They also used standard current–voltage and capacitance–voltage tests, temperature scans, and transient current technique measurements.
A clear experimental result is that even at the lowest tested doses the total leakage current increases. This rise is dominated by surface-related current rather than bulk current. In the MD8 diodes studied at a bias of −300 V, the surface current grew with dose but showed signs of saturating at around 2 Mrad. By contrast, the bulk current stayed nearly constant up to about 2 Mrad and then increased only at higher doses. These observations are consistent with the idea that ionizing radiation mainly creates charge and traps at the silicon–oxide interface (surface damage), while bulk damage requires displacement of silicon atoms and appears at higher deposited energies or doses.