⚛️ high-energy experiments

The study of bulk damage of high gamma-irradiated $n^+$ -in- $p$ silicon diodes

This study investigates the bulk damage in high-oxygen and varying-resistivity $n^+$ -in- $p$ silicon diodes subjected to high-dose gamma irradiation, revealing a linear increase in leakage current, a non-monotonic evolution of full depletion voltage driven by acceptor removal, and a notable absence of annealing effects on electrical characteristics.

Original authors: M. Mikestikova, P. Federicova, P. Gallus, R. Jirasek, J. Kozakova, J. Kroll, J. Kvasnicka, V. Latonova, I. Mandic, K. Masek, P. Novotny, R. Privara, P. Tuma, I. Zatocilova

Published 2026-07-17

📖 4 min read🧠 Deep dive

CC BY 4.0

Original authors: M. Mikestikova, P. Federicova, P. Gallus, R. Jirasek, J. Kozakova, J. Kroll, J. Kvasnicka, V. Latonova, I. Mandic, K. Masek, P. Novotny, R. Privara, P. Tuma, I. Zatocilova

Original paper licensed under CC BY 4.0 (http://creativecommons.org/licenses/by/4.0/). ✨ This is an AI-generated explanation of the paper below. It is not written or endorsed by the authors. For technical accuracy, refer to the original paper. Read full disclaimer

Imagine you are building a giant, ultra-sensitive camera to take pictures of the universe's most elusive particles. To do this, scientists use tiny silicon chips, similar to the ones in your phone, but supercharged to detect individual particles zooming through space. However, these chips have a problem: when they are bombarded by high-energy radiation (like the kind found near powerful particle accelerators), the silicon gets "bruised." Think of the silicon crystal as a perfect grid of bricks; radiation knocks some bricks loose, creating tiny holes and bumps that mess up the chip's ability to work. Scientists have long known that hitting these chips with neutrons or protons causes this damage, but they were curious about what happens when the chips are hit by gamma rays—a different kind of radiation that is everywhere in space and in nuclear facilities. The big question was: does gamma radiation bruise these silicon chips in the same way, and can we predict exactly how much damage it will do so we can build better detectors?

This paper is like a stress test for those silicon chips. The researchers took three different types of silicon diodes (tiny electronic switches) and blasted them with gamma rays from a Cobalt-60 source, delivering doses as high as 8.28 MGy (that's a massive amount of radiation). They wanted to see how the chips' "leakage current" (a bit of electricity that leaks through when it shouldn't) and their "depletion voltage" (the power needed to turn the chip fully on) changed as the radiation dose increased. They also checked if baking the chips in an oven at 60°C for 80 minutes (a process called annealing) could heal the damage, just like how a bruise might fade with time.

Here is what they found. First, the leakage current went up in a perfectly straight line as the radiation dose increased. It's like a leaky faucet where the more you turn the handle (add radiation), the more water (current) pours out. However, the speed of this leak depended on the chip's starting recipe: chips with higher electrical resistance (lower boron doping) leaked faster than those with lower resistance. The researchers calculated a "damage coefficient" for each type, showing exactly how sensitive each chip was to the gamma rays.

The most dramatic change happened with the "full depletion voltage." Imagine the silicon chip as a sponge that needs to be fully squeezed to work. At first, as the radiation hit the chips, it became much easier to squeeze them; the voltage needed dropped significantly. This is because the radiation was effectively removing the "active" ingredients (boron atoms) that made the silicon p-type. But then, something surprising happened: after a certain point, the voltage started to climb back up. For the chip with the highest initial resistance, this turnaround happened at 4.14 MGy, while the others held out until about 6 MGy. This suggests that the radiation didn't just remove the old ingredients; it eventually started creating new, opposite ingredients, flipping the chip's nature from p-type to n-type.

The researchers used a special laser technique called TCT to peek inside the chips and confirm this flip. For the low-dose chips, the laser showed they were still the original p-type. But for the high-dose chips, the laser revealed that the electrical field had flipped, confirming the bulk of the silicon had turned into an n-type material. This is a crucial finding because it means the chips don't just get "broken"; they actually transform into a different kind of device under extreme radiation.

Finally, the team tested if the "baking" (annealing) could fix anything. They heated the chips to 60°C for 80 minutes, expecting the heat to help the silicon heal itself. The result? Nothing. The leakage current and the voltage stayed exactly the same before and after the oven. This tells us that the damage caused by gamma rays is "locked in"—the defects are so stubborn that a simple warm-up won't fix them. This is very different from what happens with other types of radiation, where heat can sometimes help. The study concludes that while gamma rays do cause significant damage and even flip the chip's identity, this damage is permanent and doesn't heal with mild heat, a vital piece of information for anyone designing detectors for future space or particle physics experiments.

Technical Summary: Bulk Damage of High Gamma-Irradiated n+-in-p Silicon Diodes

Problem Statement
Silicon detectors used in the inner tracking regions of future high-luminosity particle accelerator experiments face extreme radiation environments. While extensive research exists on hadron-induced damage in p-type silicon, there is a need for comprehensive data regarding gamma-ray induced displacement damage in p-type bulk silicon, particularly for the n+-in-p technology adopted for High Luminosity LHC experiments. Previous studies on p-type samples irradiated with gamma rays were limited to total ionizing doses (TID) of approximately 3 MGy. This study aims to extend the investigation of radiation-induced displacement damage up to 8.28 MGy, characterizing the evolution of full depletion voltage ( $V_{FD}$ ), effective doping concentration ( $N_{eff}$ ), and leakage current, while determining the relationship between gamma TID and 1 MeV neutron equivalent fluence.

Methodology
The study utilized three types of standard float-zone n+-in-p silicon diodes (labeled A, B, and C) provided by different manufacturers. These diodes shared similar geometries (active area ~0.014 cm², thickness ~285–290 µm) but differed in initial bulk resistivity ( $\rho$ ) and oxygen concentration:

Diodes A & B: $\rho \approx 3.1–3.3$ k $\Omega\cdot$ cm.
Diode C: $\rho \approx 24.0$ k $\Omega\cdot$ cm (lower boron doping or significant compensation).

Oxygen concentrations were determined via Secondary-Ion Mass Spectrometry (SIMS) up to a depth of 14 µm. The diodes were irradiated by a $^{60}$ Co gamma source at UJP PRAHA a.s. to TIDs ranging from 0.50 to 8.28 MGy. Irradiations were conducted in a charge particle equilibrium (CPE) box to ensure uniform energy deposition, with temperature maintained below 35 °C. Post-irradiation, samples were stored below –20 °C to prevent uncontrolled annealing.

Characterization involved:

I–V and C–V Measurements: Performed at room temperature before and after a standard annealing step (80 minutes at 60 °C). Bulk leakage current was isolated from surface currents using guard ring configurations.
Transient Current Technique (TCT): Top-TCT using a red laser (660 nm) was employed to verify $V_{FD}$ , extract electric field distributions, and determine the sign of space charge ( $N_{eff}$ ) for diodes irradiated to the lowest and highest doses.

Key Results

Leakage Current: The bulk leakage current ( $\Delta I/V$ ) exhibited a linear increase with TID for all diode types. The damage coefficient ( $\alpha_\gamma$ ) was found to be dependent on initial resistivity and oxygen concentration:
- $\alpha_\gamma \approx 6.33 \times 10^{-6}$ A $\cdot$ cm $^{-3}\cdot$ MGy $^{-1}$ for Diode A ( $\rho \approx 3.1$ k $\Omega\cdot$ cm).
- $\alpha_\gamma \approx 6.49 \times 10^{-6}$ A $\cdot$ cm $^{-3}\cdot$ MGy $^{-1}$ for Diode B ( $\rho \approx 3.3$ k $\Omega\cdot$ cm).
- $\alpha_\gamma \approx 10.20 \times 10^{-6}$ A $\cdot$ cm $^{-3}\cdot$ MGy $^{-1}$ for Diode C ( $\rho \approx 24.0$ k $\Omega\cdot$ cm).
- Surface currents increased steeply at low TIDs and saturated, though Diode A showed a secondary increase at doses > 4 MGy.
- Annealing Effect: No change in bulk leakage current was observed after the 60 °C annealing step.
Full Depletion Voltage ( $V_{FD}$ ) and Effective Doping ( $N_{eff}$ ):
- $V_{FD}$ and $N_{eff}$ initially decreased significantly with increasing TID, reaching a minimum before increasing again at higher doses.
- The TID at which the minimum $V_{FD}$ occurred correlated with initial resistivity: Diode C (high resistivity) reached its minimum at ~4.14 MGy, while Diodes A and B (lower resistivity) reached minima near 6 MGy.
- For Diode A, $V_{FD}$ remained low (~10–20 V) above 6 MGy, though TCT measurements at 7.78 MGy indicated full depletion at ~70 V, suggesting complex space charge distributions (double junctions) prevented the geometric capacitance from being reached in C-V measurements.
Space Charge Sign Inversion (SCSI):
- TCT measurements on low-irradiated diodes confirmed the retention of the p-type bulk (n+-in-p structure).
- TCT measurements on highly irradiated diodes (A at 7.78 MGy, B at 8.28 MGy) revealed a sign inversion to n-type bulk. The electric field distribution indicated the formation of a double junction with high-field regions on both sides of the diode.
TID to Fluence Conversion:
Assuming the leakage current increase is due to displacement damage, the study estimated the conversion between TID and 1 MeV neutron equivalent fluence ( $\phi_{eq}$ ):
- For Diode C ( $\rho \approx 24$ k $\Omega\cdot$ cm): 1 MGy $\approx 2.6 \times 10^{11}$ neq/cm².
- For Diodes A and B ( $\rho \approx 3.1–3.3$ k $\Omega\cdot$ cm): 1 MGy $\approx 1.6 \times 10^{11}$ neq/cm².

Significance and Conclusions
The study characterizes the gamma-radiation induced displacement damage in p-type silicon diodes up to 8.28 MGy, a dose range relevant for future high-luminosity experiments. The primary findings indicate that:

Defect Stability: Unlike hadron-irradiated samples which exhibit beneficial and reverse annealing, gamma-irradiated n+-in-p silicon diodes show no evidence of annealing effects (neither in $V_{FD}$ nor leakage current) after 80 minutes at 60 °C. The authors attribute this to predominantly immobile defects.
Resistivity Dependence: The evolution of $V_{FD}$ is strongly dependent on initial resistivity; higher resistivity samples (lower boron concentration) reach the point of minimum $V_{FD}$ and subsequent type inversion at lower TIDs.
Type Inversion: High doses of gamma irradiation induce bulk type inversion from p-type to n-type, confirmed by TCT, similar to effects observed in hadron-irradiated samples.

The paper concludes that the initial decrease in effective doping is likely caused by acceptor removal (deactivation of boron dopants), and the subsequent increase in $V_{FD}$ at high doses is associated with the formation of a double junction and bulk type inversion.

More like this