Impact of 10 MeV Proton Irradiation on the Electrical Characteristics of Diamond Schottky and Heterojunction Diodes
Paper details
Section titled “Paper details”| Field | Value |
|---|---|
| Journal | ECS Journal of Solid State Science and Technology |
| Published | 2026-03-01 |
| Authors | Hsiao-Hsuan Wan, Katharina V. Loske, Joonyup Bae, Gyeonghwan Kim, Jihyun Kim |
| DOI | 10.1149/2162-8777/ae4f25 |
| Citations | 2 |
At 1.6 × 10¹⁴ cm⁻², the diamond Schottky diode lost two orders of forward current density and its on-resistance rose by 2.5 orders of magnitude. The carrier removal rate in the diamond drift layer reached 43 cm⁻¹ at the same fluence. The ITO/diamond heterojunction degraded faster, so device interfaces, rather than diamond alone, require separate qualification.
Radiation response and device architectures
Section titled “Radiation response and device architectures”The study compared vertical Ti/Au/p-diamond Schottky barrier diodes with vertical ITO/p-diamond heterojunction diodes. The Schottky structure places metal directly on the boron-doped diamond. The heterojunction inserts a 25 nm ITO layer before the upper metal contact. Both used 100 µm diameter front contacts and the same p-type drift layer on a p+ substrate.
The forward response was milder in the Schottky diode. Reverse leakage increased in both device types at every fluence. These results separate bulk damage in diamond from damage in the ITO layer and heterojunction interface. That distinction matters because the added oxide need not share diamond’s resistance to displacement damage.
Diamond material and contact preparation
Section titled “Diamond material and contact preparation”The starting material was single-crystal diamond supplied by EDP Corporation and synthesized by chemical vapor deposition. The substrates measured 5 × 5 mm² and were cut along the (100) orientation with a 3° miscut. X-ray rocking curves had a full width at half maximum below 50 arcsec.
The heavily boron-doped p+ substrate had a boron concentration of approximately 4 × 10²⁰ cm⁻³. A 10 µm epitaxial drift layer carried a hole concentration of 2.9 × 10¹⁶ cm⁻³. The complete diamond substrate was 260 µm thick. Mechanical polishing produced a root-mean-square roughness of 0.33 nm over a 5 × 5 µm² scan.
Electron-beam evaporation formed the full-area backside Ohmic Ti/Au contact with a thickness of 20/80 nm. Before front-side metallization, the surface received 30 minutes of ultraviolet and ozone exposure. XPS found no discernible sp2 carbon peak after this treatment. The authors used the treatment to produce a partially oxygen-terminated surface. It was selected to reduce surface conductivity relative to hydrogen-terminated diamond.
The front Ti/Au contact was deposited directly on diamond for the Schottky devices. For heterojunction devices, the same metal contact sat on top of sputtered ITO.
Proton exposure and electrical extraction
Section titled “Proton exposure and electrical extraction”A 30 MeV proton beam from a Scanditronix MC 50 cyclotron reached the samples at 10 MeV after aluminum degradation. The average beam current was 10 nA, monitored with a Faraday cup. Irradiation fluences were 1.0 × 10¹³, 6.0 × 10¹³, and 1.6 × 10¹⁴ cm⁻².
SRIM placed the ion distribution peak and maximum NIEL damage at about 400 µm inside diamond. This depth lay beyond the 260 µm substrate, so the samples did not contain the end-of-range damage peak. The calculated NIEL at 10 MeV was 5.62 × 10⁻⁵ MeV·cm²·g⁻¹.
Electrical measurements occurred in ambient air at 298 K. An Agilent 4156C analyzer recorded low-bias forward and reverse I-V data. Specific on-resistance came from the I-V slope using dV/dI. Carrier concentrations came from 1 MHz C-V measurements and linear fits to 1/C² against voltage.
Forward conduction and diode parameters
Section titled “Forward conduction and diode parameters”The Schottky diode lost about 60% of its forward current at 1.0 × 10¹³ cm⁻². Its on-resistance increased by about 60% at the same fluence. At 1.6 × 10¹⁴ cm⁻², forward current fell by two orders and on-resistance rose by 2.5 orders.
| Parameter | Reference | 1.0 × 10¹³ cm⁻² | 6.0 × 10¹³ cm⁻² | 1.6 × 10¹⁴ cm⁻² |
|---|---|---|---|---|
| Schottky ideality factor | 1.11 | 1.20 | 1.24 | 1.26 |
| Schottky barrier height (eV) | 1.34 | 1.27 | 1.26 | 1.26 |
| Heterojunction ideality factor | 1.47 | 2.16 | 3.02 | 3.51 |
The Schottky ideality factor rose from 1.11 to 1.26. The authors associate this shift with trap-assisted transport added to thermionic emission. The extracted Schottky barrier height fell from 1.34 eV to 1.26 eV. They interpret that reduction as irradiation-modified interface states or stronger Fermi-level pinning.
The heterojunction ideality factor increased from 1.47 to 3.51. This much larger change was consistent with greater disruption of the ITO and its interface. The reported ITO bandgap was 3.5 to 4.0 eV, below diamond’s approximately 5.5 eV. The authors suggest that an ITO phase or property shift could make the junction behave as a metal-oxide-semiconductor structure. They present this mechanism as an interpretation, not a direct structural measurement.
Carrier removal and radiation hardness
Section titled “Carrier removal and radiation hardness”C-V extraction gave the following carrier losses and removal rates.
| Proton fluence (cm⁻²) | Carrier concentration change, Δn (cm⁻³) | Carrier removal rate, Δn/Φ (cm⁻¹) |
|---|---|---|
| 1.0 × 10¹³ | 1.7 × 10¹⁵ | 170 |
| 6.0 × 10¹³ | 3.9 × 10¹⁵ | 65 |
| 1.6 × 10¹⁴ | 6.8 × 10¹⁵ | 43 |
The absolute carrier loss increased monotonically with fluence. Δn reached 6.8 × 10¹⁵ cm⁻³ at the highest fluence. The nearly linear Δn-versus-fluence fit indicated continued accumulation in the linear damage regime.
The CRR fell from 170 cm⁻¹ at 1.0 × 10¹³ cm⁻² to 43 cm⁻¹ at 1.6 × 10¹⁴ cm⁻². The authors read this decline as a saturation effect in incremental carrier removal. Their nearly linear Δn fit indicates that total carrier removal remained in the linear damage regime.
For direct context, the paper cites approximately 200 to 300 cm⁻¹ for β-Ga₂O₃ under 10 MeV protons. The measured diamond rates were below that range and consistent with earlier proton irradiation studies of diamond. The authors link this behavior to diamond’s atomic displacement energy of about 43 eV, which limits lattice displacement.
Reverse leakage and damage mechanisms
Section titled “Reverse leakage and damage mechanisms”Reverse leakage current increased with fluence in both the Schottky and heterojunction devices. The authors attribute this response to displacement-induced deep levels in the diamond. Under reverse bias, those defects facilitate trap-assisted tunneling and generation-recombination. They can also distort the local electric field and lower the effective barrier.
The paper notes energetically possible (p,p) and (p,α) nuclear reactions between 5 and 10 MeV protons and carbon. Their cross-sections are lower than those for primary displacement events, but secondary particles may add lattice disorder. The greater heterojunction degradation suggests that the ITO layer and interface are more vulnerable than the diamond bulk.
Test limits and engineering implications
Section titled “Test limits and engineering implications”The experiment covered one proton energy, three fluences, and one device temperature. Electrical data were taken at 298 K in ambient air, while C-V measurements used a fixed 1 MHz frequency. The 260 µm substrate also stopped short of the simulated 400 µm damage maximum. These results therefore describe moderate-fluence, non-end-of-range damage rather than the full range of proton stopping conditions.
The device set was specific to p-type single-crystal diamond, boron levels, Ti/Au contacts, ozone-treated surfaces, and 25 nm ITO. No structural measurement in the paper directly confirms the proposed ITO phase or property shift. The paper also reports no annealing study or long-term irradiation reliability test.
For radiation-hard power electronics, the direct Schottky contact was the more stable architecture under the tested conditions. Diamond’s low carrier removal rate supports its use as a radiation-tolerant drift material. Interface qualification remains necessary because the heterojunction lost conduction more rapidly. The authors conclude that Schottky diodes retain favorable electrical performance at moderate proton fluences.
The highest tested fluence was 1.6 × 10¹⁴ cm⁻². The paper does not establish performance at higher fluences, other proton energies, or elevated temperatures. Device selection for space or accelerator use should therefore include the intended particle spectrum and operating temperature.
Materials and support
Section titled “Materials and support”For the paper’s vertical p-type rectifiers, 6CCVD’s diamond materials catalog includes boron-doped single-crystal diamond grown by MPCVD, with thickness from 0.1 to 500 µm and substrates up to 10 mm. The paper’s 260 µm substrate and 10 µm drift layer sit within that range, and Au is an available metallization option for its Ti/Au contacts. Custom fabrication support for nonstandard stacks is available through engineering support.
View Original Abstract
The radiation tolerance of single-crystal diamond devices was investigated under 10 MeV proton irradiation. Schottky barrier diodes (SBDs) and ITO/diamond heterojunction diodes were fabricated on boron-doped diamond substrates with a 10 μm lightly doped drift layer and exposed to proton fluences from 1.0 × 10 13 to 1.6 × 10 14 cm −2 . Irradiation induced increased on-resistance, reduced saturation current, and enhanced reverse leakage, with heterojunction devices showing greater degradation due to the vulnerability of the ITO layer and interface. Capacitance-voltage measurements revealed carrier removal rates of 171, 65, and 43 cm −1 for fluences of 1.0 × 10 13 , 6.0 × 10 13 , and 1.6 × 10 14 cm −2 , respectively, confirming diamond’s superior radiation hardness compared to Si, GaN, and Ga 2 O 3 . This resilience is attributed to diamond’s high atomic displacement energy, which limits lattice damage. These results demonstrate the potential of single-crystal diamond devices for radiation-hard power electronics and high-radiation environments.
Boron Doped Single Crystal Diamond Plate
View Specs →
Electronic Grade Intrinsic Diamond Epitaxial Wafer H Terminated Superior 100 Mosfet Ready Substrate
View Specs →
Diamond Epitaxial Wafer For Diodes 2
View Specs →References we cover
Section titled “References we cover”- High Current Density Diamond Photoconductive Semiconductor Switches With a Buried, Metallic Conductive Channel
- A study of the radiation tolerance of poly-crystalline and single-crystalline CVD diamond to 800 MeV and 24 GeV protons
Original source
Section titled “Original source”References
Section titled “References”- 1999 - Optically triggered diamond switches
- 2024 - High voltage and high-power diamond based junction-gate field effect transistor (JFET) switch with photo-controlled gate