Skip to content

Investigating ultra-thin 4H-SiC AC-LGADs for superior radiation-hard timing applications

FieldValue
JournalJournal of Instrumentation
Published2026-02-01
AuthorsJaideep Kalani, Saptarshi Datta, Ganesh J. Tambave, Prabhakar P. Palni
DOI10.1088/1748-0221/21/02/p02037

Section 4.4.2 reports a simulated time resolution below 20 ps for a 20 µm 4H-SiC AC-LGAD above 375 V. The case used 50 × 10¹⁴ neq cm⁻², 243 K, and 3.701 × 10¹⁶ cm⁻³ boron gain doping. The result came from WeightField2, not a fabricated 4H-SiC detector. The paper cites LGAD timing resolution at approximately 50 ps or lower for collider tracking.

WeightField2 build 6.0 modeled planar AC-LGADs with selectable thickness, strip geometry, bias, gain implant profile, and boron concentration. AC coupling capacitively links the readout to the gain layer through an insulating layer. The gain implant sat 0.5 to 1.0 µm below the top surface, while the Massey impact-ionization model supplied the gain mechanism.

4H-SiC MIPs generated 57 electron-hole pairs per micron under uniform ionization. Neutron damage included acceptor creation, exponential acceptor removal, and trapping coefficients of 4.9 × 10⁻¹⁶ cm²/ns for electrons and 6.2 × 10⁻¹⁶ cm²/ns for holes. Time resolution used the NA62 transimpedance amplifier model, a 20 ps TDC bin, and WF2 signal amplitude and rise time. Bias steps were 1 V, gain was reported to two decimals, and time was reported to 0.01 ps.

The framework reproduced published gain and rise-time trends for an FBK W6 silicon DC-LGAD. The physical sensor was 60 µm thick, while the simulation used a 55 µm active layer after thermal bonding. The comparison used an estimated 4.71 × 10¹⁶ cm⁻³ gain doping, 293 K before irradiation, and 253 K after irradiation. The estimate had the best agreement with published gain and rise-time data. FBK used B+C gain doping, whereas the material studies used boron. This validation concerned silicon. The reported 4H-SiC timing remained a simulation result.

A fixed-gain comparison used 20 µm sensors at 150 V and 243 K. Gain implant doping was adjusted separately for Si, 4H-SiC, and diamond before fluence increased. The 50 × 10¹⁴ neq cm⁻² results were:

Bulk materialTotal collected chargeTime resolution
Siliconabout 2 fCabout 23 ps
4H-SiCabout 2 fCbelow 18 ps
Diamondbelow 1 fCbelow 18 ps

Only SiC met both selection criteria: high collected charge and good timing across the fluence range. Diamond produced fewer electron-hole pairs than silicon, but fast transport and low noise kept its timing close to SiC. The paper treats the diamond result as conceptual, so it does not establish diamond as the preferred LGAD bulk.

The induced-current relation scales with G/d at fixed gain, saturation velocity, and generated charge. A thinner sensor therefore produces a larger current and slew rate under this model.

Reducing SiC thickness from 100 µm to 20 µm improved simulated time resolution by about 60% at constant gain. The study used 300 V, 243 K, five strips, and uniform 57-pair ionization. At gain 10, the corresponding improvement was 60% for SiC and 44% for silicon.

SiC gain rose more steeply with implant doping than silicon at fixed thickness and bias. Across all simulated thicknesses, SiC delivered higher gain for the same implant concentration. Required implant concentration increased with thickness, then changed little between 50 and 70 µm.

Gain fell as temperature rose from 243 K to 293 K at 320, 340, 350, and 360 V. The 20 µm SiC device used 2.301 × 10¹⁶ cm⁻³ boron gain doping. Higher bias offset part of the gain loss.

At 243 K and 360 V, SiC reached about 13 ps, compared with about 22 ps for silicon. From 293 K to 243 K, SiC’s time-resolution value fell about 33% at 320 V and 48% at 360 V. Silicon fell about 35% and 52% at the same biases. The authors attribute SiC’s stability to higher thermal conductivity and its wider bandgap, which limits leakage at elevated temperature.

Radiation dosage used the equivalent 1 MeV neutron fluence scale. Bulk defects trap and recombine carriers, while deactivated acceptors reduce gain and signal-to-noise ratio.

At 10 × 10¹⁴ neq cm⁻², a 20 µm SiC sensor needed 350 V to match the 150 V unirradiated gain. At 350 V, peak field fell from about 452 kV/cm at zero fluence to 201 kV/cm at 5 × 10¹⁵ neq cm⁻². Peak acceptor concentration fell from 3.7 × 10¹⁶ cm⁻³ to 2.4 × 10¹⁵ cm⁻³ over that interval. The multiplication region did not vanish; it widened as the doping peak declined.

Raising bias from 300 V to 385 V lowered time resolution about 77% at 20 × 10¹⁴ neq cm⁻². At 50 × 10¹⁴ neq cm⁻², the improvement was about 51% across a 40 V increase. At 365 V and 10 × 10¹⁴ neq cm⁻², SiC lowered time resolution by about 40% relative to silicon. At 50 × 10¹⁴ neq cm⁻², SiC’s time-resolution value was about 37% lower than silicon’s.

The SiC results assume a silicon-based gain-layer parametrization applied identically to Si, 4H-SiC, and diamond. The model treats gain and bulk regions as one semiconductor with different doping. SiC therefore tests bulk transport and signal shaping under an imposed gain profile, not a verified SiC gain layer. Diamond is more conceptual because its charge collection and timing use the same imposed gain mechanism.

4H-SiC trapping coefficients were unavailable, so the study assigned silicon values of 4.9 × 10⁻¹⁶ and 6.2 × 10⁻¹⁶ cm²/ns. The assumed irradiation response of the SiC gain layer remains unverified experimentally. Uniform MIP deposition excluded Landau fluctuations, making the unirradiated result below 10 ps idealized. WF2’s 20 ps TDC bin also limits interpretation of the reported timing values.

No fabricated 4H-SiC AC-LGAD was tested. Atomic displacement energy was cited as a likely additional benefit but was not included in the simulation. The authors therefore frame the SiC advantage as a benchmark for future devices, not measured hardware performance.

The paper gives different optimized values at 50 × 10¹⁴ neq cm⁻². Section 4.4.2 reports below 20 ps above 375 V, while the conclusion reports about 55 ps. This discrepancy remains unresolved in the text.

The results favor a 20 µm 4H-SiC bulk for high-radiation timing, with bias and temperature treated as design variables. For diamond, the paper reports competitive timing but lower charge collection than SiC. The modeled result does not support diamond as the preferred LGAD bulk.

The paper treats diamond only as a conceptual LGAD sensor bulk, so a 20 µm diamond plate or wafer would be needed for gain-layer integration. 6CCVD’s diamond materials catalog lists MPCVD-grown single-crystal and polycrystalline diamond with thicknesses from 0.1 to 500 micrometers. The engineering support page can be used to specify the plate geometry and contact requirements for a custom detector build.

View Original Abstract

Abstract The Low Gain Avalanche Diodes (LGADs) are promising particle detectors for timing resolution better than 50 ps under a high radiation environment. This study investigates n-in-p LGAD architecture, focusing on ultra-thin sensors of thickness less than 50 μm using the WeightField2 program. The capabilities of WeightField2 are demonstrated by comparing its results with irradiation measurements from an FBK LGAD wafer, showing good agreement across unirradiated and neutron-irradiated conditions. This paper presents device simulations in High Luminosity LHC conditions (lifetime integrated fluence 𝒪(10 14 ) n eq cm -2 ), temperature ≈ 243 K), and taking into account radiation damage, gain reduction due to fluence, and lattice defects. It is shown that a 20 μm thick sensor achieves the best timing performance. Among Silicon (Si), Diamond (C), and 4H-Silicon Carbide (4H-SiC), we found 4H-SiC to be the most promising: it provides the highest gain value for a fixed thickness and gain implant layer configuration, and best retains high charge collection value and timing capability under increasing fluence up to 50×10 14 n eq cm -2 . A time resolution less than 25 ps is reported with different gain implant concentrations for a 20 μm 4H-SiC sensor. This work presents the potential of SiC-based LGADs in high-radiation collider environments.

  1. 2016 - The High-Luminosity upgrade of the LHC: Physics and Technology Challenges for the Accelerator and the Experiments [Crossref]
  2. 2019 - FCC-ee: The Lepton Collider: Future Circular Collider Conceptual Design Report Volume 2 [Crossref]
  3. 2019 - FCC-hh: The Hadron Collider: Future Circular Collider Conceptual Design Report Volume 3 [Crossref]
  4. 2020 - Design of a Segmented LGAD Sensor for the Development of a 4-D Tracking Detector [Crossref]
  5. 2022 - Measurements of time and spatial resolution of AC-LGADs with different designs [Crossref]
  6. 2019 - A MIP Timing Detector for the CMS Phase-2 Upgrade
  7. **** - Technical Proposal: A High-Granularity Timing Detector for the ATLAS Phase-II Upgrade [Crossref]